The context is i, j, k as indices in programs. No doubt FORTRAN was influenced by prior use such as you cite. But in no case does i used as an index come from i designating an imaginary number, which is what I aimed to refute.