Locally almost square Banach lattices (Q6042192)

The present paper concerns some concepts of almost squareness (ASQ) in a Banach space \(X\), defined and studied in [\textit{T.~A. Abrahamsen} et al., J. Math. Anal. Appl. 434, No.~2, 1549--1565 (2016; Zbl 1335.46006)], namely, W(eak)ASQ and L(ocal)ASQ. Here ASQ \(\Rightarrow\) WASQ \(\Rightarrow\) LASQ, and none of the implications are in general reversible. So, the author asks if one can obtain LASQ \(\Rightarrow\) WASQ under additional assumptions. Say that a Banach lattice \(X\) is positive LASQ if, for every \(y\in S_X\) and \(\varepsilon>0\), there is \(x\) in the positive part \(S_X^+\) of \(S_X\) with \(\|y\pm x\|\leq 1+\varepsilon\). This is just LASQ with the extra assumption that \(x\) comes from \(S_X^+\). No AL-space can be positive LASQ, but (Prop.~2.3) if an AM-space is LASQ, it is already positive LASQ. The main result of the paper is the following (Thm.~2.5): If \(X\) is positive LASQ, then for every \(y\in S_X\) there exists a weak-null net \((x_\alpha)\subset X\) with \(\lim_\alpha\|x_\alpha\|=1\) and \(\lim_\alpha\|y\pm x_\alpha\|=1\). When \(B_{X^{\ast\ast}}\) is weak-star sequentially compact, the proof can easily be adjusted to show that in such cases positive LASQ implies WASQ (Thm.~2.7). The author asks (Question~2.6) if LASQ just by itself implies the conclusion of Thm.~2.5 for Banach spaces in general. The paper ends by studying how positive LASQ may imply diameter~2 properties. Recall that LASQ implies the local diameter~2 property (every slice of \(B_X\) has diameter~2). The author proves (Thm.~3.1) that, if \(X\) is positive LASQ, then \(X\) has the diameter~2 property (every non-void relatively weak-open subset of \(B_X\) has diameter~2).