11.07.2015 Views

kniga 7 - Probability and Statistics 1 - Sheynin, Oscar

kniga 7 - Probability and Statistics 1 - Sheynin, Oscar

kniga 7 - Probability and Statistics 1 - Sheynin, Oscar

SHOW MORE
SHOW LESS

Create successful ePaper yourself

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

addition theorem depending on the points which are included in it <strong>and</strong> are corresponding tothe elementary propositions.As to the totalities of the second type considered above, they include, because of their verystructure, only such joins 34 , that are reducible to finite or countable joins, so that we do nothave to mention non-measurable totalities; consequently, all the propositions of the totalities,both of the fourth <strong>and</strong> the second type, acquire quite definite probabilities once thearithmetizing function F(z) is chosen.3.2.7. Arithmetization of Totalities of the Third TypeOn the basis of what was said about totalities of the fourth type, we already know that afterarithmetization only a countable totality of possible elementary propositions can be left in theconsidered totality. The difference between arithmetized totalities of the third <strong>and</strong> the fourthtypes only consists in that there exist infinite joins making sense in the fourth, but not in thethird type; we do not therefore mention here probabilities of such joins. All the other joinswill have the same probabilities in both totalities.Summarizing all that was said about the arithmetization of infinite totalities, we see that, towhichever type they belong, this procedure is entirely determined by the function F(x) 35 , onwhose choice the very type of the totality also depends because a point of discontinuity ofF(z) corresponds to each elementary proposition, <strong>and</strong> vice versa. If we admit the generalizedconstructive principle, we obtain, depending on the nature of F(z), totalities of the second <strong>and</strong>the fourth types. If, however, we hesitate to attach sense to some infinite joins (<strong>and</strong>combinations), our totalities should be attributed to the first or the third type.3.2.8. Arithmetizaton of the Totality of IntegersIntegers <strong>and</strong> their finite joins provide an example of a totality of the third type. If we linkto them all the possible infinite joins, we obtain a totality of the fourth type with a countabletotality of elementary propositions. Its arithmetization is usually achieved on the basis of theassumption that all numbers are equally possible. This premise however is obviouslyinadmissible because it would have implied that the probability of each number is zero, i.e.,that no number could have been realized, <strong>and</strong>, in addition, the generalized addition theoremwould have been violated because the sum of the probabilities of a countable totality ofpropositions with probability 0 would have been unity.The difficulty of selecting a law of probability for the numbers depending, in eachconcrete case, on the statement of the problem at h<strong>and</strong>, cannot justify the choice of a law,even if it is simple, contradicting the main principles of the theory of probability. We mayconsider such a limit of probabilities of some propositions, that corresponds to a graduallyincreasing restricted totality of numbers, under the assumption that in such totalities thenumbers are equally possible, but that limit is not the probability of a definite proposition ofour infinite totality.Another inadmissible assumption, connected with the one just mentioned, is made as oftenas that latter, viz., that the probability for the number N, when divided by a prime number ato provide a remainder , does not depend on the remainder left after dividing N by a primenumber b. Indeed, let o = 0, 1 = 1 be the two possible remainders after dividing N by two; o = 0, 1 = 1 <strong>and</strong> 2 = 2, the remainders after dividing it by three; etc. Then, according to theassumption, the probabilities of all the infinite combinations ( <strong>and</strong> <strong>and</strong> …) are equal, butmost of these combinations are impossible, because, when dividing N by a greater number,all the remainders obtained become equal to N, so that, for example, the combination (0, 1, 0,1, …) of the remainders is impossible. And it would follow that also impossible are thosecombinations which correspond to integers.It is also possible to attach another meaning to all the combinations if we connect each ofthem with the series

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

Saved successfully!

Ooh no, something went wrong!