Further results on supernumary polylogarithmic ladders (Q2366035)

This paper is a sequel to the paper of \textit{L. Lewin} and \textit{M. Abouzahra} [Aequationes Math. 39, No. 2/3, 210-253 (1990; Zbl 0729.33015)] and continues the construction of `ladders' for polylogarithm functions. A method due to Zagier is used to find further `cyclotomic equations' and computer algebra then finds the ladders. Various bases are treated, including the roots of the equations \(x^ 4+x^ 3=1\), \(x^{11}+x^ 7=1\), \(x^ 4-x^ 3-x^ 2-x+1=0\) and \(x^{10}+x^ 9- x^ 7-x^ 6-x^ 5-x^ 4-x^ 3+x+1=0\) (the last two being interesting Salem numbers). A new phenomenon appears to arise in the concluding sections. To give one example, Watson found ladders \(L_ 2(\alpha)\), \(L_ 2(\beta)\), \(L_ 2(\gamma)\) for the dilogarithm based on the roots \(\alpha,-\beta,- 1/\gamma\), say, of the cubic \(x^ 3+2x^ 2-x-1=0\) (where the notation is devised to give \(0<\alpha,\beta,\gamma<1)\). None of these extends to a ladder for the trilogarithm, but the author finds numerically that \(L_ 3(\alpha)+L_ 3(\beta)+L_ 3(\gamma)={11\over 14}\zeta(3)\). Despite much recent progress by Zagier and others, the full resolution of these mysteries is still to come.