On the higher dimensional Mordell conjecture over function fields (Q1191034)

This paper gives an improvement of an older result by the same author, published in Kodai Math. J. 11, No. 1, 1-4 (1988; Zbl 0653.14017). The general question is, whether the \(K\)-rational points on a projective variety of general type defined over the field \(K\) can be Zariski dense. Relevant fields \(K\) for this question are number fields, where this is a conjecture attributed to Vojta, and where the case of a variety of dimension 1 coincides with the celebrated `Mordell conjecture', first proven by Faltings in 1983, for which now several proofs are available. - - Another important case is that of a function field \(K\) over the complex numbers. Here it is well known that one has to put in the extra condition that the variety is not isotrivial, which means that it shouldn't be birational over \(K'\) to a variety of the type \(V_ 0\otimes_ \mathbb{C} K'\), with \(K'\) a finite extension of \(K\) and \(V_ 0\) some projective variety over \(\mathbb{C}\). This function field analogue of the generalized Mordell conjecture is still unsettled, too; \textit{S. Lang}, \textit{E. Bombieri} and \textit{J. Noguchi} have conjectured it. When the variety is a curve over \(K\) it is a celebrated theorem, proven independently by \textit{Grauert} and \textit{Manin} in the 60's [\textit{R. F. Coleman} published a correction of an overlooked gap in Manin's proof fairly recently, cf. Enseign. Math., II. Ser. 36, No. 3/4, 393-427 (1990; Zbl 0729.14018)]. In the present paper, criteria involving the projective bundle \(P=\mathbb{P}\bigl(\Omega^ 1_{V/K}\bigr)\) are given, which imply that the \(K\)-rational points on the variety of general type \(V\) are not Zariski dense. The author explains how this result can be regarded as an extension of results of \textit{J. Noguchi} [Math. Ann. 258, 207-212 (1981; Zbl 0459.14002)] and of Deschamps. -- By basic techniques, the general statement is reduced to the case where \(K\) is the function field of a curve. An other important general result which the author proves and uses successfully, is that sufficiently high powers of the sheaf \({\mathcal O}(1)\) on a projective bundle \(\mathbb{P} \bigl( \Omega^ 1_ V \bigr)\) are generated by global sections.