[tjgq]

> 2. If it's not prime, it has divisors. This proof claims it must include divisors that are prime but are not among those we started with. <-- THIS IS THE BIT I DON'T GET

This step of the proof invokes the Fundamental Theorem of Arithmetic [0]: the statement that every natural number can be expressed as a unique product of primes (uniqueness isn't the important bit here, just the fact that such a factorization exists).

So you need to accept the Fundamental Theorem of Arithmetic as true before you can fully understand this proof.

[0] http://en.wikipedia.org/wiki/Fundamental_theorem_of_arithmet...

[tjgq]

To expand a bit further: this is an example of the "rabbit-hole" nature of mathematics. All but the most trivial theorems depend on previous results, and in most cases you cannot realistically follow all the dependencies until you get to the first principles, also known as axioms (and even then, there's the question of which axioms you are willing to accept!)

In order to be able to understand and appreciate mathematical proofs, you have to develop the ability to accept the truthness of a result - and to realize the consequences of it being true - even though you do not yet understand why it is true. You have to learn to accept the ensuing confusion as a natural state of mind (read [0] if this idea intrigues you).

[0] http://j2kun.svbtle.com/mathematicians-are-chronically-lost-...

[lutusp]

> In order to be able to understand and appreciate mathematical proofs, you have to develop the ability to accept the truthness of a result - and to realize the consequences of it being true - even though you do not yet understand why it is true.

I have to disagree. A trained mathematician understands why a proof is or it not valid, based on a combination of axioms and a logical sequence predicated on axioms, but with no gaps or overlooked assumptions.

Without knowing and explaining why, it would not be possible to write a proof that would pass muster with other mathematicians, people who by instinct and training refuse to accept fuzzy explanations.

This is why Gödel's Incompleteness Theorems came as such a shock, at a time when many people expected to be able to systematize all of mathematics and predicate it on a handful of unassailable logical principles (as Russell and Whitehead attempted to do in the early 20th century).

The Incompleteness Theorems show the degree to which mathematicians expect to know why something is true, and if they cannot, why not.

[tjgq]

> I have to disagree. A trained mathematician understands why a proof is or it not valid, based on a combination of axioms and a logical sequence predicated on axioms, but with no gaps or overlooked assumptions.

First off, to dispel any misunderstanding, I was talking about the process of becoming a mathematician, which you necessarily go through before you can call yourself one.

But even for full-fledged mathematicians, what I'm saying is true to an extent. For instance, although ZFC set theory is usually accepted as the basis for all mathematics, most mathematicians (precisely: those who do not study formal logics or other areas of metamathematics) do not state their results in terms of set theory, but instead write informal proofs that appeal to other established results in their field of study.

That is absolutely not the same as saying that their thinking is fuzzy or sloppy. The fact that I personally do not state (or, even, understand) all the details behind an established theorem X has no bearing on the validity of my proof for a theorem Y that hinges on X being true.

> The Incompleteness Theorems show the degree to which mathematicians expect to know why something is true, and if they cannot, why not.

This is a separate matter. Whether in theory it is possible for you to ascertain whether X is true or not, and whether you fully understand (down to first principles) why X is true before you use it as a stepping stone for other results are different issues.