p-independence implies pseudomodularity (Q752736)

Let K be a field of prime characteristic p. Assume that K is finitely generated over \(K^ p\), the field of pth powers of the elements of K, and that K is distinct from \(K^ p\). The notion of p-independence was defined by \textit{S. MacLane} [Duke Math. J. 4, 455-468 (1938; Zbl 0019.39201)], who observed that the geometric lattice of fields k, \(K^ p\subseteq k\subseteq K\), is not always modular. A semimodular lattice L with rank function r(x) is called pseudomodular [cf. \textit{A. Björner} and \textit{L. Lovász}, Acta Sci. Math. 51, 295-308 (1987; Zbl 0643.05023)] if for any x,y\(\in L\) the set \(P_{x,y}=\{z:\) \(z\leq y\) and \(r(x\vee z)-r(z)=r(x\vee y)-r(y)\}\) has a unique least element. The author proves the following Theorem. Let p be a prime and K finitely generated over \(K^ p\) (K is an inseparable extension of exponent 1 over \(K^ p)\). Then the geometric lattice of fields k, \(K^ p\subseteq k\subseteq K\), is pseudomodular.