Systems biology faces a choice between reductionist and holist approaches, but practising systems biologists are often unaware of what the implications of each path will be. Modern neo-holism, as manifested in Robert Rosen’s Relational Biology, concludes that the functions of complex systems are irreducible to the functions of their component parts, and also implies that the current foundations of computational theory are inadequate for systems biology. By contrast, modern neo-reductionism replaces classical conceptions of inter-theory reduction with the looser concept of supervenience, in the process reassuring us that we can make progress in systems biology with computational theory as we know it today. However, the price to pay for this is a shift away from modelling to realisational strategies. Either way, the entire field of systems biology may have to change course if it is to accomplish its goals.