Coincidence classes in nonorientable manifolds (Q2491521): Difference between revisions

The classical Lefschetz coincidence theorem is concerned with the existence of coincidences of two maps \(f,g:M\to N\) between compact orientable manifolds of the same dimension. Parallel to Nielsen fixed point theory, \textit{H. Schirmer} [J. Reine Angew. Math. 194, 21--39 (1955; Zbl 0066.41701)] established a similar theory for coincidences but the setting is again for maps between orientable manifolds. \textit{R. Dobrenko} and \textit{J. Jezierski} [Rocky Mt. J. Math. 23, No. 1, 67--85 (1993; Zbl 0787.55003)] introduced the notion of semi-index in order to develop a Nielsen coincidence theory in which the manifolds may be non-orientable. Assume that \(f\) and \(g\) are transverse, a typical coincidence class takes the form \[ C=\{a_1,\dots, a_k, b_1,c_1,\dots,b_n,c_n\} \] where the \(a_i\)'s are mutually un-related under homotopy and the \(b_i\) and \(c_i\) are related. The semi-index of \(C\) is \(k\). Thus the Nielsen number \(N(f,g)\) (using semi-index) is defined and coincides with the classical one if \(M\) and \(N\) are both orientable. If \(b_i=c_i\) for some \(i\), we say that the class \(C\) contains a self-reducing coincidence in which case \(C\) is said to be \textit{defective}. In this paper, the author studies the coincidence problem for two maps \(f,g:M\to N\) in terms of lifts \(\widetilde f, \widetilde g\). In particular, the semi-index of a coincidence class of two lifts \(\widetilde f, \widetilde g\) to finite covers is computed. As an application, the author proves (Theorem 5.1) that if \(M\) is non-orientable and \(N\) is an orientable Jiang-type space (i.e., coincidence classes have coincidence index of the same sign in the usual sense) then all essential coincidence classes must be defective.