Current spindle models explain "anaphase A" (movement of chromosomes to the poles) in terms of a motility system based solely on microtubules (MTs) and that functions in a manner unique to mitosis. We find both these propositions unlikely. An evolutionary perspective suggests that when the spindle evolved, it should have come to share not only components (e.g., microtubules) of the interphase cell but also the primitive motility systems available, including those using actin and myosin. Other systems also came to be involved in the additional types of motility that now accompany mitosis in extant spindles. The resultant functional redundancy built reliability into this critical and complex process. Such multiple mechanisms are also confusing to those who seek to understand how chromosomes move. Narrowing this commentary down to just anaphase A, we argue that the spindle matrix participates with MTs in anaphase A and that this matrix may contain actin and myosin. The diatom spindle illustrates how such a system could function. This matrix may be motile and work in association with the MT cytoskeleton, as it does with the actin cytoskeleton during cell ruffling and amoeboid movement. Instead of pulling the chromosome polewards, the kinetochore fibre's role might be to slow polewards movement to allow correct chromosome attachment to the spindle. Perhaps the earliest eukaryotic cell was a cytoplast organised around a radial MT cytoskeleton. For cell division, it separated into two cytoplasts via a spindle of overlapping MTs. Cytokinesis was actin-based cleavage. As chromosomes evolved into individual entities, their interaction with the dividing cytoplast developed into attachment of the kinetochore to radial (cytoplast) MTs. We believe it most likely that cytoplasmic motility systems participated in these events.