Small values of polynomials: Cartan, Pólya and others (Q1383389)

The author's summary: ``Let \(P(z)\) be a monic polynomial of degree \(n\), and \(\alpha,\varepsilon>0\). A classic lemma of Cartan asserts that the lemniscate \(E(P;\varepsilon):=\{z:\;| P(z)| \leq\varepsilon^n\}\) can be covered by balls \(B_j\), \(1\leq j\leq n\), whose diameters \(d(B_j)\) satisfy \(\sum_{j=1}^p\left(d(B_j)\right)^\alpha\leq e(4\varepsilon)^\alpha\). For \(\alpha=2\), this shows that \(E(P;\varepsilon)\) has an area at most \(\pi e(2\varepsilon)^2\). Pólya showed in this case that the sharp estimate is \(\pi\varepsilon^2\). We discuss some of the ramifications of these estimates, as well as some of their cousins, for example when \(P\) is normalized to have \(L_p\) norm 1 on some circle, and Remez' inequality.'' Let us add that the paper is completed by recalling versions of Cartan's lemma for multivariate polynomials due to \textit{A. Cuyt, K. Driver} and the author [Proc. Am. Math. Soc. 124, No. 7, 2123-2136 (1996; Zbl 0853.30003)] and \textit{B. Paneah} [J. Approximation Theory 78, No. 3, 402-409 (1994; Zbl 0807.32004)]. The author closes this interesting article by the pessimistic remark: ``Notions of capacity are far more complicated in the multivariate case, and several basic questions remain unresolved. One of the main problems is the lack of an explicit formula for the Green's function.'' Actually, the things are not going so bad, since one can define a natural multivariate counterpart of the Green's function as a solution of the homogeneous complex Monge-Ampère equation. For this function, called often plurisubharmonic extremal function, one can find explicit formulas in many special cases. For bacground and details, see \textit{M. Klimek}, Pluripotential theory (1991; Zbl 0742.31001).