Meromorphic functions sharing four small functions (Q5900982)

The classical uniqueness theorems for meromorphic functions tell that whenever two meromorphic functions share five, resp. four, extended complex values IM, resp. CM, then \(f=g\), resp. \(f=M\circ g\), where \(M\) is a Möbius transformation. Later on, \textit{G. G. Gundersen} proved that CM in the four-point theorem can be replaced with 2CM + 2IM, see [Trans. Am. Math. Soc. 277, 545--567 (1983; Zbl 0508.30029)]. He also pointed out that 4CM can not be replaced with 4IM, while the combination 1CM + 3IM remains open until now. Later on, \textit{M. Shirosaki} extended the claim of the four-point theorem (with 4CM) to the case of four small target functions (instead of four extended complex numbers), see [Tôhoku Math. J. 45, 491--497 (1993; Zbl 0802.30026)]. In the present paper, the author improves the Shirosaki result admitting truncated sharing instead of the 4CM assumption as follows: Let \(f,g\) be two distinct nonconstant meromorphic functions, and let \(a_{1},a_{2},a_{3},a_{4}\) be four distinct meromorphic functions small with respect to \(f\) in the usual sense of Nevanlinna theory. Suppose that \(f,g\) share \(a_{1},a_{2},a_{3},a_{4}\), counting with multiplicities truncated by \(2\). Then \(a_{1},a_{2},a_{3},a_{4}\) contains exactly two elements, say \(a_{3},a_{4}\) with \(\Theta (a_{j},f)=1\), \(j=3,4\), the cross ratio of \(a_{1},a_{2},a_{3},a_{4}\) equals to \(-1\), and again \(f=M\circ g\). The result obtained is optimal in the sense that truncation by \(1\) is not possible by previous examples provided by Gundersen (loc. cit.).