Defining \(\mathbb Z\) in \(\mathbb Q\) (Q5962625)

From MaRDI portal
scientific article; zbMATH DE number 6541583
Language Label Description Also known as
English
Defining \(\mathbb Z\) in \(\mathbb Q\)
scientific article; zbMATH DE number 6541583

    Statements

    Defining \(\mathbb Z\) in \(\mathbb Q\) (English)
    0 references
    0 references
    15 February 2016
    0 references
    By Matiyasevich's theorem, a subset \(X\subseteq\mathbb N\) is Diophantine, i.e. expressable in the form \(\{y\in\mathbb N:\exists{\vec x}\in\mathbb N^np(y,\vec x)=0\}\) for some \(n\in\mathbb N\), \(p\in\mathbb Z[\vec x,y]\), if and only if \(X\) is recursively enumerable. Combined with the existence of recursively enumerable sets that are not recursive, this implies the unsolvability of Hilbert's 10th problem, which asked for a general decision procedure for the solvability of Diophantine equations. On the other hand, it is well-known that the theory of \(\mathbb R\) in the language of ordered rings is decidable. One of the main open questions suggested by these results is whether the solvability of polynomial equations with rational coefficients in \(\mathbb Q\) is decidable. One way to attack this problem would be to show that \(\mathbb Z\) has a Diophantine definition over \(\mathbb Q\), thus reducing the question to Matiyasevich's theorem. A more relaxed question is then whether \(\mathbb Z\) is definable over \(\mathbb Q\) at all. This was answered in the positive by a classical result of \textit{J. Robinson} [J. Symb. Log. 14, 98--114 (1949; Zbl 0034.00801)], who showed that \(\mathbb Z\) has a \(\Pi_2\)-definition over \(\mathbb Q\). For a negative solution to Hilbert's 10th problem over \(\mathbb Q\) along the lines sketched above, a \(\Sigma_1\)-definition would be needed. The present paper brings the complexity considerably down by showing that \(\mathbb Z\) is \(\Pi_1\)-definable over \(\mathbb Q\) (Theorem \(1\)), thus showing that \(\mathbb Q\setminus\mathbb Z\) is Diophantine over \(\mathbb Q\) (Corollary 2) and that the \(\Pi_2\)-theory of \(\mathbb Q\) is undecidable (Corollary 3). The proof proceeds in four steps: In step 1, a certain \(\Pi_2\)-definition of \(\mathbb Z\) in \(\mathbb Q\) is given by adapting a construction of \textit{B. Poonen} [Am. J. Math. 131, No. 3, 675--682 (2009; Zbl 1179.11047); Math. Res. Lett. 16, No. 1, 165--170 (2009; Zbl 1183.14031)]. Step 2 consists in proving that the set of rational \(p\)-adic integers is Diophantine in \(\mathbb Q\). Step 3 gives \(\Sigma_1\)-definitions for some Jacobson radicals relevant in step 2. Finally, step 4 uses model-theoretical results to express certain existentially expressed formulas as universal formulas. The author mentions that the Diophantine subsets of \(\mathbb Q\) obtained in step 2 may lead to a different proof strategy of the undecidability of Hilbert's 10th problem over \(\mathbb Q\) than the one suggested above and suggests further ways to show that \(\mathbb Z\) is Diophantine in \(\mathbb Q\) in Section 4. The paper requires a background in algebraic number theory. As its main result is a landmark in the theory of Diophantine equations, it should appeal to logicians, number theorists and recursion theorists.
    0 references
    Hilbert's 10th problem
    0 references
    Diophantine equation
    0 references
    definablity
    0 references
    undecidability
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references