Notes for 3/4/2026

  

3/4/2026

[Philosophy Club every Monday, 4-5 pm, in the Buchtel College of Arts and Sciences room 436 ("The Cave")]

[Bioethics Club: Mondays from 5:30pm-6:30pm in Leigh Hall 408]

 

Are you at all superstitious?

 

 

Hume notes that cause-effect pairs are not analytically connected (we can’t know the effect by analyzing the cause or vice-versa).

 

If causes and effects are not analytically connected (logically necessary), then what OTHER type of connection can there be?

 

Can there be another kind of necessity than logical necessity? (Physical necessity?)

 

Where COULD the idea of physical necessity come from in experience?

 

How do we learn causal connections?

 

 

 

Ivan Pavlov

 

 

Hume says we learn cause-effect relations by “custom” (or habit).

 

 

Hume claims that the connection between cause and effect isn’t really necessary. It just FEELS necessary to us because our minds have been habituated to connect them.

 

There seems to be, for Hume, no difference between BEING causally necessary and FEELING causally necessary.

 

This seems to imply that causal beliefs are somehow not RATIONAL.

(This poses an obvious threat to science. (Major motivation for later philosophy (in particular, Immanuel Kant).))

 

 

Causal realism: Cause-effect relationships are really OUT THERE in the world.

 

Causal anti-realism (or causal skepticism): Cause-effect relationships are merely artifacts of our minds.

 

 

 

Exercise: Think of a case where someone believes in a causal connection that you don’t believe in (pseudoscience, superstition, etc.).

 

 

Correlation is not causation

 

Post hoc ergo propter hoc.

(After this therefore because of this.) 

 

 

A possible response to the apparent skepticism of Hume’s analysis of causation is to hold that causal beliefs are not DEDUCTIVELY rational, but they are INDUCTIVELY rational.

 

Deductive reasoning depends on logical structures that are defined in certain ways.

 

Consider the argument:

1. Either Wilma or Betty is the winner.

2. Wilma is not the winner.

3. Therefore, Betty is the winner.

 

If we ask how we know that deductive reasoning is reliable, the answer is that once we understand how deductive reasoning works, we understand that it HAS TO work.

 

Inductive reasoning is different, though.

Consider the argument:

1. All the ravens I’ve ever seen are black.

2. Therefore, all ravens are black.

 

The general form of this argument is:

1. All the x I’ve ever seen are y.

2. Therefore, all x are y.

 

Clearly, the form of this argument doesn’t make it so that “all x are y” has to be true.

 

It’s the nature of inductive reasoning that it is fallible. (Another point in favor of fallibilism.)

 

 

But the question remains as to why inductive reasoning is ever any good to begin with.

 

At the heart of inductive reasoning is this claim:

 

     The future will resemble the past.

 

Is this necessarily true? No.

 

So, it is only known by experience. How do we know it?

 

Answer: Because in the past, the future has turned out to resemble the past.

 

 

To put it another way:

 

There is no deductive argument that inductive reasoning is reliable.

 

The argument for the reliability of inductive reasoning relies on inductive reasoning. Which is circular reasoning.

The problem Hume identifies here has been called The Problem of (the justification of) Induction.

 

In brief: Inductive reasoning cannot be deductively justified and can’t be inductively justified without presupposing that inductive reasoning is justified.

 

Is circular reasoning bad?

(For purposes of explanation, yes. “X because X” doesn’t really explain anything.)

 

The problem of induction is puzzling, but not necessarily all that worrisome.

 

There are other cases where one needs/uses X in order to do X.

Here are two examples from computer science:

 

Self-hosting: It is possible for compilers or interpreters to be written in the same language they compile or interpret (for example, an implementation of Python can be written in Python).

 

Recursive functions: Some functions can call themselves as part of their procedure.

 

 

Example:

def factorial(n):

if n == 0 or n == 1:

return 1

else: return n * factorial(n - 1)

 

Another example is that it is impossible to give a definition of ‘language’ that isn’t given in language.

 

Another possibility:

Distinguish rational from pragmatic justification.