Hume's Enquiry, sections IV part 2, and V

---

Hume, Section IV part 2

• Our reasoning about matters of fact is based on cause and effect, and our reasoning about cause and effect is based upon experience - but then, what is the nature of all of our conclusions from experience?
• These conclusions "are not founded on reasoning, or any process of the understanding."
• Compare:
  1. I have found that such an object has always been attended with such an effect
  2. I foresee, that other objects, which are, in appearance, similar, will be attended with similar effects.
• Hume holds that: we do leap from 1 to 2 (we actually find that we are compelled to leap from 1 to 2), but we have no reasoning leading from 1 to 2.
• Experience shows us individual instances of experience right now.
• Reasoning about cause and effect draws conclusions based on kinds of experiences in the future.
• This is not given in reasoning about relations of ideas, since there is no contradiction in supposing a causal claim false.
• This is not given in reasoning about matters of fact.
• BUT! All knowledge from matters of fact is based upon experience, and all experience is based upon reasoning about cause and effect! So, we cannot appeal to experience to explain cause and effect.
• Thus, causal reasoning comes down to: we see some correlation, and we believe it will continue.
• "From causes, which appear similar, we expect similar effects. This is the sum of all our experimental conclusions."

Hume, Section V, part 1

• Philosophy can tend to lead to error, when we want our theories to give us good answers.
• We can avoid making errors about what is unclear by remaining skeptical when appropriate.
• Our skeptical problem (from Section IV) is: why assume that there is a special relation between causes and effects?
• Hume gives not a "straight" answer, which would describe such a special relation, but rather a skeptical solution, which asserts our lack of any solution but can live with this result.
• Suppose someone comes fresh into the world: for him all things are like random events.
• After experiencing the world, he comes to see some correlations. He begins to assume that one kind of thing always follows another.
• This is a custom or habit.
• Belief in cause and effect is not based on any feature in the world, but is rather just a feature of human nature.

  "after the constant conjunction of two objects, heat and flame, for instance, weight and solidity, we are determined by custom alone to expect the one from the appearance of the other" (page 28).
  
  "All inferences from experience, therefore, are effects of custom, not of reasoning" (p 28)

• This is the most compelling argument for the problem of induction ever given.

An aside about the Problem of Induction

• Hume's Enquiry Section IV gives our first statement of the problem of induction that we have encounted.
• Section V gives Hume's skeptical "solution" to the problem.
• (Section VII is where Hume completes his account of the problem of induction, by attacking the notion of a power or necessity underlying causation.)
• Hume's statement of the problem is the most famous, and after Hume a number of solutions have been tried.
• The problem of induction
  • Hume's formulation: we assume that A causes B and will do so in the future because we've seen that As are followed by Bs in our experience.
  • Bertrand Russell's restatement: we have no reason to believe in "the uniformity of nature."
  • It must be conceded, to begin with, that the fact that two things have been found often together and never apart does not, by itself, suffice to prove demonstratively that they will be found together in the next case we examine.
  • For example, Russell says, "The only reason for believing that the laws of motion will remain in operation is that they have operated hitherto."
• Kant offers one very influential solution: argues that experience requires cause and effect. Since there is experience, there is cause and effect.
• A contemporary solution to the Problem of Induction
  • Today, another solution is the most widely accepted: a pragmatic solution.
  • The pragmatic solution is based on two ideas
    1. There are events with dependent, converging probabilities
    2. Time is not a special variable in testing convergences
  • Point 1 says that: The aim of induction is to find series of events whose frequency of occurrence converges towards a limit. (Hans Reichenbach)
  • A frequency converges towards a limit if it has a given probability. An ideal coin toss converges towards 50% heads, for example.
  • Point 2 says that being in the future is not different than being in another place as far as testing for convergences is concerned.
  • Frederick Will's example of the island with regions called past and future is relevant.
  • Will accuses Russell and perhaps Hume of an implicit mistake: tomorrow, like the old song, is "always a day away."
  • This solution works, BUT it does mean that we still need to assume one thing: there are phenomena in the world that are interdependent to some probability that converges at the limit.
  • We have gotten rid of the problem of induction, if this argument is sound, but we still have a (perhaps slightly less troubling) "problem": we have identified something that we must assume, but cannot prove. We must assume that these converging relationships exist.
  • Epistemology may just come down to "you have to stand somewhere," and the goal then is finding somewhere reasonable to stand.
• Other solutions to the problem of induction, much less discussed or even well understood, include Solomonoff's proof that less complex assumptions are more probable.

Hume, Section V part 2

• Hume argues that belief is a kind of sentiment that we attach to impressions.
• This sentiment is more commonly attached to stronger impressions.
• Impressions are stronger when related to other (strong) impressions; all such relations are
  1. Resemblance
  2. Contiguity
  3. Causation
• When we believe something X, and another thought Y resembles, is continuous with, or "caused" by that thought X, then we are more inclined to feel the sentiment of belief for Y.
• Cause and effect then primarily serves to generate in us sentiments of belief about one thing, given that another thing we believe occurred before it typically in the past.