# I Information and Cardinality

1. Jan 7, 2017

### tzimie

To select an element from countably infinite set (list set of integers) you need to provide finite amount of information. To specify an element in continuum in general case you have to provide infinite amount of information: any real number is specified as countable-infinite number of digits. So here is a pattern:

$\aleph_0$ - finite information
$\aleph_{continuum}$ - countably infinite number of digits,

So the amount of information is 1 degree less than the cardinality of set.

Now, lets deny continuum hypotesis and lets assume continuum = $\aleph_2$, so there is 1 cardinality between countable and continuum (I've heard that Goedel believed in it), and that cardinality is obviously $\omega_1$

What information is required to completely specify an element in $\omega_1$ ?

2. Jan 7, 2017

### Staff: Mentor

Shouldn't it be $\aleph_1$ instead of $\omega_1$ by its definition? But anyway. Whether you put it in a pseudo information theory language, or leave it where it belongs to, namely ZFC, makes no difference. Since it is undecidable in ZFC, it stays as such in all other languages based upon ZFC.

3. Jan 7, 2017

### mathman

Rational numbers are real. Not all real numbers require an infinite number of digits.

4. Jan 8, 2017

### tzimie

yes, but in general case you have to provide an infinite number of digits
most of reals are random numbers.

5. Jan 8, 2017

### tzimie

Thank you for the correction.
But then question "what amount of information is required to specify a countable ordinal?" (which is an element of $\omega_1$ ) must be also undecidable, because answering it you also provide one or another "answer" to CH?

6. Jan 8, 2017

### Staff: Mentor

Yes, only that it can't be another answer. Assuming CH were decidable in some axiomatic system, then it would have to be different from ZFC. However, information theory isn't.

7. Jan 19, 2017

### Zafa Pi

If the cardinality of R is $\aleph_2$, then there is a subset A of R with cardinality $\aleph_1$. The elements of A can't be specified with a finite amount of information for that would imply A is countable. But the elements of A can be specified by a subset of those that specify R, so the answer is countable.

Given what everyone else has said I must have stepped in some pile of logical poop.