25.01.2013 Views

popper-logic-scientific-discovery

popper-logic-scientific-discovery

popper-logic-scientific-discovery

SHOW MORE
SHOW LESS
  • No tags were found...

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

354<br />

new appendices<br />

It should be mentioned, in this connection, that Kolmogorov fails,<br />

in his book, to offer an independence proof for his ‘axiom of continuity’<br />

although he claims independence for it. But it is possible to reframe<br />

our proof of independence so that it becomes applicable to<br />

Kolmogorov’s axiom and his set-theoretic approach. This may be<br />

done by choosing, instead of our S1 a system of intervals S3, exactly<br />

like S1 but based upon a sequence C = c1, c2, ..., defined by cn = (0,<br />

2 − n ] rather than upon the sequence A = a1, a2, ..., with an = (0,<br />

1<br />

2 + 2 − n ]. We can now show the independence of Kolmogorov’s axiom<br />

by defining the probabilities of the elements of the sequence A as<br />

follows:<br />

p(c n) = l(c n) + 1 2 = p(a n)<br />

Here l(cn) is the length of the interval cn. This definition is highly<br />

counter-intuitive, since, for example, it assigns to both the intervals (0,<br />

1<br />

2] and (0, 1] the probability one, and therefore to the interval ( 1 2, 1] the<br />

probability zero; and the fact that it violates Kolmogorov’s axiom<br />

(thereby establishing its independence) is closely connected with its<br />

counter-intuitive character. For it violates the axiom because lim<br />

p(cn) = 1 2, even though p(c) = 0. Because of its counter-intuitive character,<br />

the consistency of this example is far from self-evident; and so the<br />

need arises to prove its consistency in order to establish the validity of<br />

this independence proof of Kolmogorov’s axiom.<br />

But this consistency proof is easy in view of our previous independence<br />

proof—the proof of the independence of our own first definition<br />

with the help of the example S1. For the probabilities p(an) and p(cn) of<br />

the two examples S1 and S3 coincide. And since by correlating the two<br />

sequences, A and C, we may establish a one-one correspondence<br />

between the elements of S1 and S3, the consistency of S1 proves that of S3. It is clear that any example proving the independence of Kolmogorov’s<br />

axiom must be equally counter-intuitive, so that its consistency<br />

will be in need of proof by some method similar to ours. In other<br />

words, the proof of the independence of Kolmogorov’s axiom will<br />

have to utilize an example which is, essentially, based upon a<br />

(Boolean) definition of product such as ours, rather than upon the<br />

set-theoretic definition.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!