Wolpert's ‘French Flag Problem’ explains how morphogen gradients can provide embryonic cells with positional information. Morphogen interpretation has largely been assumed to be a cell-autonomous effect, with each cell measuring concentration relative to fixed thresholds; however, the role of cell-cell communication has been unexplored. Now, Hyung-Chul Lee, Cato Hastings, Claudio Stern and colleagues combine computational modelling and experimental embryology to determine whether the chick epiblast uses cell-cell communication to interpret morphogens for primitive steak (PS) formation. By grafting VG1-expressing cells or ACTIVIN-soaked beads (both of which induce PS development) into the epiblast, the authors show that a single PS-inducing region causes ectopic PS formation but, counterintuitively, introducing a row of PS-inducing factors across the embryo does not. However, separating the PS-inducing factors by a control region or a region supplying an inhibitor of the PS (BMP4) promotes two ectopic PSs. The authors use computational modelling to distinguish between a concentration threshold and a relative (neighbourhood) concentration model. As predicted by their ‘Neighbourhood watch’ model, grafting an ACTIVIN-soaked bead (at a concentration insufficient to induce PS alone) between two BMP4 beads forms an ectopic PS. Together, these data indicate that cells respond to morphogen concentrations relative to their neighbours.