请输入您要查询的字词:

 

单词 ProofOfBanachTarskiParadox
释义

proof of Banach-Tarski paradox


We deal with some technicalities first, mainly concerning theproperties of equi-decomposability. We can then prove the paradoxMathworldPlanetmath in aclear and unencumbered line of argument: we show that, given two unitballsMathworldPlanetmath U and U with arbitrary origin, U and UU areequi-decomposable, regardless whether U and U are disjoint ornot. The original proof can be found in [BT].

Technicalities

Theorem 1.

Equi-decomposability gives rise to an equivalence relationMathworldPlanetmath on thesubsets of Euclidean space.

Proof.

The only non-obvious part is transitivity. So let A, B, C besets such that A and B as well as B and C areequi-decomposable. Then there exist disjoint decompositionsA1,An of A and B1,Bn of B such that Akand Bk are congruentMathworldPlanetmathPlanetmath for 1kn. Furthermore, there existdisjoint decompositions B1,,Bm of B and C1,Cmof C such that Bl and Cl are congruent for 1lm. Define

Bk,l:=BkBl for 1kn,1lm.

Now if Ak and Bk are congruent via some isometryMathworldPlanetmath θ:AkBk, we obtain a disjoint decomposition of Ak by settingAk,l:=θ-1(Bk,l). Likewise, if Bl and Cl arecongruent via some isometry θ:BlCl, we obtain adisjoint decomposition of Cl by settingCk,l:=θ(Bk,l). Clearly, the Ak,l and Ck,ldefine a disjoint decomposition of A and C, respectively, intonm parts. By transitivity of congruenceMathworldPlanetmathPlanetmath, Ak,l and Ck,l arecongruent for 1kn and 1lm. Therefore, A andC are equi-decomposable.∎

Theorem 2.

Given disjoint sets A1,,An, B1,Bn such thatAk and Bk are equi-decomposable for 1kn, theirunions A=k=1nAk and B=k=1nBk areequi-decomposable as well.

Proof.

By definition, there exists for every k, 1kn an integerlk such that there are disjoint decompositions

Ak=i=1lkAk,i,Bk=i=1lkBk,i

such that Ak,i and Bk,i are congruent for 1ilk. Rewriting A and B in the form

A=k=1ni=1lkAk,i,B=k=1ni=1lkBk,i

gives the result.∎

Theorem 3.

Let A, B, C be sets such that A and B are equi-decomposableand CA, then there exists DB such that Cand D are equi-decomposable.

Proof.

Let A1,,An and B1,Bn be disjoint decompositions ofA and B, respectively, such that Ak and Bk are congruent viaan isometry θk:AkBk for all 1kn. Letθ:AB a map such that θ(x)=θk(x) if xAk. Since the Ak are disjoint, θ is well-definedeverywhere. Furthermore, θ is obviously bijectiveMathworldPlanetmathPlanetmath. Now setCk:=CAk and define Dk:=θk(Ck), so that Ck andDk are congruent for 1kn, so the disjoint unionMathworldPlanetmathD:=D1Dn and C are equi-decomposable. Byconstruction, θ(C)=D. Since C is a proper subsetMathworldPlanetmathPlanetmath of A andθ is bijective, D is a proper subset of B.∎

Theorem 4.

Let A, B and C be sets such that A and C areequi-decomposable and ABC. Then B and C areequi-decomposable.

Proof.

Let A1,,An and C1,Cn disjoint decompositions ofA and C, respectively, such that Ak and Ck are congruent viaan isometry θk for 1kn. Like in the proof oftheorem 3, let θ:AC be the well-defined,bijective map such that θ(x)=θk(x) if xAk. Now,for every bB, let 𝒞(b) be the intersectionMathworldPlanetmath of all setsXB satisfying

  • bX,

  • for all xX, the preimageMathworldPlanetmath θ-1(x) lies in X,

  • for all xXA, the image θ(x) lies in X.

Let b1,b2B such that 𝒞(b1) and 𝒞(b2) are notdisjoint. Then there is a b𝒞(b1)𝒞(b2) such thatb=θr(b1)=θs(b2) for suitable integers s andr. Given b𝒞(b1), we haveb=θt(b1)=θt+s-r(b2) for a suitable integer t,that is b𝒞(b2), so that𝒞(b1)𝒞(b2). The reverse inclusion followslikewise, and we see that for arbitrary b1,b2B either𝒞(b1)=𝒞(b2) or 𝒞(b1) and 𝒞(b2) aredisjoint. Now set

D:={bB𝒞(b)A},

then obviously DA. If for bB, 𝒞(b) consistsof the sequenceMathworldPlanetmath of elements,θ-1(b),b,θ(b), which is infiniteMathworldPlanetmathPlanetmath in bothdirections, then bD. If the sequence is infinite in only onedirection, but the final element lies in A, then bD as well.Let E:=θ(D) and F:=BD, then clearly EFC.

Now let cC. If cE, then θ-1(c)D, so𝒞(θ-1(c)) consists of a sequence,θ-2(c),θ-1(c), which is infinite inonly one direction and the final element does not lie in A. Nowθ-1(c)𝒞(θ-1(c)), but since θ-1(c)does lie in A, it is not the final element. Therefore thesubsequent element c lies in 𝒞(θ-1(c)), in particularcB and 𝒞(c)=𝒞(θ-1(c))A, socBD=F. It follows that C=EF, and furthermore Eand F are disjoint.

It now follows similarly as in the preceding proofs that D andE=θ(D) are equi-decomposable. By theorem 2, B=DF and C=EF are equi-decomposable.∎

The proof

We may assume that the unit ball U is centered at the origin, thatis U=𝔹3, while the other unit ball U has an arbitraryorigin.Let S be the surface of U, that is, the unit sphereMathworldPlanetmath. By theHausdorff paradoxMathworldPlanetmath, there exists a disjoint decomposition

S=BCDE

such that B, C, D and CD are congruent, and E iscountableMathworldPlanetmath. For r>0 and A3, let rA be the set ofall vectors of A multiplied by r. Set

B:=0<r<1rB,C:=0<r<1rC,D:=0<r<1rD,E:=0<r<1rE.

These sets give a disjoint decomposition of the unit ball with theorigin deleted, and obviously B, C, D and CDare congruent (but E is of course not countable). Set

A1:=BE{0}

where 0 is the origin. B and CD are triviallyequi-decomposable. Since C and B as well as D and C arecongruent, CD and BC are equi-decomposable. Finally, Band CD as well as C and B are congruent, so BC andBCD are equi-decomposable. In total, B and BCD are equi-decomposable by theorem 1, so A1 and U areequi-decomposable by theorem 2. Similarly, we conclude thatC, D and BCD are equi-decomposable.

Since E is only countable but there are uncountably many rotationsof S, there exists a rotation θ such thatθ(E)BCD, so F:=θ(E) is a propersubset of BCD. Since C and BCD areequi-decomposable, there exists by theorem 3 a proper subsetGC such that G and F (and thus E) areequi-decomposable. Finally, let pCG an arbitrary pointand set

A2:=DG{p},

a disjoint union by construction. Since D and BCD, Gand E as well as {0} and {p} are equi-decomposable, A2and U are equi-decomposable by theorem 2. A1 and A2are disjoint (but A1A2U!).

Now U and U are congruent, so U and A2 areequi-decomposable by theorem 1. If U and U aredisjoint, set H:=A2. Otherwise we may use theorem 3 andchoose HA2 such that H and UU areequi-decomposable. By theorem 2, A1H and U(UU)=UU are equi-decomposable. Now we have

A1HUUU,

so by theorem 4, U and UU are equi-decomposable.

References

  • BT St. Banach, A. Tarski, Sur la décompositiondes ensembles de points en parties respectivement congruentes,Fund. math. 6, 244–277, (1924).
随便看

 

数学辞典收录了18232条数学词条,基本涵盖了常用数学知识及数学英语单词词组的翻译及用法,是数学学习的有利工具。

 

Copyright © 2000-2023 Newdu.com.com All Rights Reserved
更新时间:2025/5/4 6:34:03