Re: JSH: A simple error
- From: José Carlos Santos <jcsantos@xxxxxxxx>
- Date: Sun, 27 May 2007 15:07:33 +0100
On 27-05-2007 14:45, jstevh@xxxxxxxxx wrote:
Posters in attacking my proofs showing inconsistency with the ring of
algebraic integers routinely move outside the ring. Here is a post
meant to show you how they do it.
In the ring of integers, consider x^2 + 3x + 2 = 0, which of course
factors as
x^2 + 3x + 2 = (x+2)(x+1)
and now solve for it using the quadratic formula, but kind of weird by
NOT resolving the square root then
x = (-3 +/- sqrt(1))/2
and now make the substitution, x=2y, so you get
How can I make that substitution? You wrote that we are working in the
ring of integers. And, given an integer _x_ it is often false that I
can write an integer _x_ as 2y for an integer _y_. In a ring, you cannot
divide by arbitrary elements.
4y^2 + 6y + 2 = 0, so you can divide by 2 to get
2y^2 + 3y + 1 = 0,
and your solution now becomes
y = (-3+/- sqrt(1))/4
which is two solutions where one is not an integer.
So you moved outside the ring of integers.
So what's the trick?
The trick consists in stating that we are working in the ring of
integers and then doing some operation which cannot be performed on that
ring.
Well, with integer solutions you can resolve the square root and throw
away one solution, which is what most people routinely do, so they say
that sqrt(4) = 2.
With integer solutions, the solutions are (-3 + sqrt(1))/2 = -1 and
(-3 - sqrt(1))/2 = -2. That makes two solutions, not one.
Best regards,
Jose Carlos Santos
.
- References:
- JSH: A simple error
- From: jstevh
- JSH: A simple error
- Prev by Date: Re: Free action of a group on R^n
- Next by Date: Re: Free action of a group on R^n
- Previous by thread: JSH: A simple error
- Next by thread: Re: A simple error
- Index(es):
Relevant Pages
|