25.01.2013 Views

popper-logic-scientific-discovery

popper-logic-scientific-discovery

popper-logic-scientific-discovery

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

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

(90) p(a, c) = p(b, c) → p(ā, c) = p(b¯, c) 64<br />

Consequently, we have<br />

(91) p(ā b¯ c¯, d) = p(ā b¯ c¯, d) 62, 89, 40<br />

(92) p(ā b¯ c¯, d) = p(ā b¯ c¯, d) 90, 91<br />

This is the law of association for the Boolean sum. By substituting in<br />

(40) the complements of a and b, we find<br />

(93) p(ā b¯, c) = p(b¯ ā, c) 40, 90<br />

This is the law of commutation for the Boolean sum. In the same way<br />

we get<br />

(94) p(ā ā, b) = p(a, b) 30, 89, 90<br />

This is the law of idempotence for the Boolean sum. From (87) we<br />

obtain<br />

(95) p(a, b) = p(a, bcc¯), 87, 40, A2<br />

(96) p(a, b)p(b) = p(ab) 95, B2, 75<br />

This may also be written<br />

(97) p(b) ≠ 0 → p(a, b) = p(ab)/p(b) 96<br />

This formula shows that our generalized concept of relative probability<br />

coincides, for p(b) ≠ 0, with the usual concept, and that our calculus is<br />

a generalization of the usual calculus. That the generalization is a genuine<br />

one can be seen from the examples, given in the preceding appendix<br />

*iv, showing the consistency of our system with the following<br />

formula (E):<br />

(E)<br />

(Ea)(Eb)(Ec) p(a, b) = 1 & p(a, bc) = 0<br />

appendix *v 363<br />

—a formula which is invalid in many finite interpretations of our S but<br />

valid in its normal infinite interpretations.

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

Saved successfully!

Ooh no, something went wrong!