Embryo development (ontogeny) depends on developmental gene regulatory networks (dGRNs), but dGRNs depend on pre-existing spatial anisotropies that are defined by early embryonic axes, and those axes are established long before the embryo’s dGRNs are put in place. For example, the anterior-posterior axis in Drosophila and the animal-vegetal axis in Xenopus and echinoderms are initially derived from the architecture of the ovary through processes mediated by cytoskeletal and membrane patterns rather than dGRNs. This review focuses on plasma membrane patterns, which serve essential ontogenetic functions by providing targets and sources for intracellular signaling and transport, by regulating cell-cell interactions, and by generating endogenous electric fields that provide three-dimensional coordinate systems for embryo development. Membrane patterns are not specified by DNA sequences. Because of processes such as RNA splicing, RNA editing, protein splicing, alternative protein folding, and glycosylation, DNA sequences do not specify the final functional forms of most membrane components. Still less does DNA specify the spatial arrangements of those components. Yet their spatial arrangements carry essential ontogenetic information. The fact that membrane patterns carry ontogenetic information that is not specified by DNA poses a problem for any theory of evolution (such as Neo-Darwinism) that attributes the origin of evolutionary novelties to changes in a genetic program—-whether at the level of DNA sequences or dGRNs. This review concludes by suggesting that relational biology and category theory might be a promising new approach to understanding how the ontogenetic information in membrane patterns could be specified and undergo the orchestrated changes needed for embryo development.