On 2012-05-24, WM <
muec...@rz.fh-augsburg.de> wrote:
> On 24 Mai, 08:31, Uirgil <uir...@uirgil.ur> wrote:
>> In article
>> <
65dc68ad-7594-4268-be98-7095e3369...@5g2000vbf.googlegroups.com>,
>>
>>
>>
>>
>>
>> WM <
mueck...@rz.fh-augsburg.de> wrote:
>> > On 23 Mai, 00:42, Uirgil <uir...@uirgil.ur> wrote:
>> > > In article
>> > > <
9db0b271-e437-4e8a-83e2-e0708f325...@b26g2000vbt.googlegroups.com>,
>>
>> > > WM <
mueck...@rz.fh-augsburg.de> wrote:
>> > > > On 22 Mai, 22:19, Uirgil <uir...@uirgil.ur> wrote:
>> > > > > In article
>> > > > > <
dc2813f8-c73a-4563-add5-48feb4d47...@5g2000vbf.googlegroups.com>,
>>
>> > > > > WM <
mueck...@rz.fh-augsburg.de> wrote:
>> > > > > > On 22 Mai, 10:53, William Elliot <
ma...@panix.com> wrote:
>> > > > > > > On Tue, 22 May 2012, WM wrote:
>>
>> > > > > snip
>>
>> > > > > > > > Then there are at least two irrational numbers without a rational
>> > > > > > > > between them. That is mathematically impossible.
>>
>> > > > > > > Prove it or cram it.
>>
>> > > > > > The proof has been given. I repeat: As every rational q_n is covered
>> > > > > > by an interval I_n, we can conclude that every pair x and x' of
>> > > > > > irrationals is separated by least one interval I_n. Hence x and x'
>> > > > > > cannot belong to the same complementary interval. This holds for every
>> > > > > > pair of irrationals.
>>
>> > > > > > A very simple conclusion. What do you not understand?
>>
>> > > > > Why you claiming what is so obviously false.
>>
>> > > > > Unless EVERY pair of irrationals is separated by EVERY ONE of your
>> > > > > I_n's, which does not ever happen, for some pairs of irrationals there
>> > > > > will be LOTS of your "complimentary intervals' containing both.
>>
>> > > > > And as every complimentary interval is necessarily of positive length,
>> > > > > being of form [0,q) or (q,1] for some rational q with 0 < q < 1, they
>> > > > > are each proved to contain infinitely many irrationals.
>>
>> > > > Without rationals between them?
>>
>> > > > You misunderstand. The complementary intervals do not contain any
>> > > > rationals q_n. They cannot because all q_n are covered by intervals
>> > > > I_n. Therefore the complementary intervals are merely single points.
>> > > > (Limits of sequences of intervals I_n.)
>>
>> > > Then your original definition was sufficiently ambiguous as not to make
>> > > that clear.-
>>
>> > I said: "every rational q_n is covered by an interval I_n". So, after
>> > having understood now, what is your objection?
>>
>> You did not make clear that your complimentation was with respect to the
>> union of all those rational intervals, rather than one compliment for
>> each interval.-
>
> Now it should be clear that at most 1/9 of the unit interval (or even
> of the complete real line) is covered by intervals I_n. That means at
> least 8/9 (infinity) is not covered by intervals I_n. This remaining
> part does not contain any rational numbers. Therefore either two
> irrationals of that remaining part must be connected or disconnected
> by one of the intervals I_n. The former is not possible in mathematics
> (perhaps it is possible in matheology - but that is not of interest
> for me). The second means that we have either only countably many
> irrationals or we are forced to believe that there are more
> irrationals than rationals and nevertheless every pair of irrationals
> is separated by an I_n.
What the heck makes you think they must be separated by exactly one of
your intervals? Inbetween any two distinct reals there are INFINITELY
MANY rationals and thus infinitely many of your I_n intervals. So what
we're talking about is an injection from |R^2 to the power set of |N.
That such an injection exists is a well-known result and is _of course_
consistent with the existence of uncountable infinity.