T and B cells share a common somatic gene rearrangement mechanism for assembling the genes that code for their antigen receptors; they also have developmental pathways with many parallels. Shared usage of basic helix-loop-helix E proteins as transcriptional drivers underlies these common features. However, the transcription factor networks in which these E proteins are embedded are different both in membership and in architecture for T and B cell gene regulatory programs. These differences permit lineage commitment decisions to be made in different hierarchical orders. Furthermore, in contrast to B cell gene networks, the T cell gene network architecture for effector differentiation is sufficiently modular so that E protein inputs can be removed. Complete T cell-like effector differentiation can proceed without T cell receptor rearrangement or selection when E proteins are neutralized, yielding natural killer and other innate lymphoid cells.