This study investigates the structure and dynamics of interactions in schooling fish, specifically golden shiners. By analyzing the trajectories of fish in two- and three-fish shoals, the researchers map the mean effective forces as a function of fish positions and velocities. They find that speed regulation is a dominant component of fish interaction, with changes in speed transmitted to both behind and ahead fish. Alignment emerges from attraction and repulsion, with fish copying directional changes made by those ahead. The study challenges the standard assumption in physics-inspired models that individual motion results from averaging responses to each neighbor. Instead, three-body interactions make a substantial contribution to fish dynamics, and pairwise interactions qualitatively capture the correct spatial interaction structure in small groups, persisting in larger groups of 10 and 30 fish. The interactions revealed here may help account for the rapid changes in speed and direction that enable real animal groups to stay cohesive and amplify important social information. The study also shows that directional information flows from the front to the back of the group, while speed information flows bidirectionally. The analysis of three-fish shoals reveals nonpairwise interactions that are not present in existing models of animal groups. These interactions include synergistic effects when a fish has two independent reasons to accelerate or decelerate. The study highlights the importance of three-body interactions in maintaining group cohesion and suggests that models based on pairwise interactions may overlook important interactions. The results suggest that individual interactions are nonisotropic and noncentral, with the speeding force depending on the front-back distance of neighbors and the turning force on their distance to the side. The study also emphasizes the need for new collective motion models that capture these interactions and for further research on how individual movement behaviors relate to emergent group-level properties. The findings have implications for understanding collective behavior in other species and for developing more accurate models of animal group dynamics.This study investigates the structure and dynamics of interactions in schooling fish, specifically golden shiners. By analyzing the trajectories of fish in two- and three-fish shoals, the researchers map the mean effective forces as a function of fish positions and velocities. They find that speed regulation is a dominant component of fish interaction, with changes in speed transmitted to both behind and ahead fish. Alignment emerges from attraction and repulsion, with fish copying directional changes made by those ahead. The study challenges the standard assumption in physics-inspired models that individual motion results from averaging responses to each neighbor. Instead, three-body interactions make a substantial contribution to fish dynamics, and pairwise interactions qualitatively capture the correct spatial interaction structure in small groups, persisting in larger groups of 10 and 30 fish. The interactions revealed here may help account for the rapid changes in speed and direction that enable real animal groups to stay cohesive and amplify important social information. The study also shows that directional information flows from the front to the back of the group, while speed information flows bidirectionally. The analysis of three-fish shoals reveals nonpairwise interactions that are not present in existing models of animal groups. These interactions include synergistic effects when a fish has two independent reasons to accelerate or decelerate. The study highlights the importance of three-body interactions in maintaining group cohesion and suggests that models based on pairwise interactions may overlook important interactions. The results suggest that individual interactions are nonisotropic and noncentral, with the speeding force depending on the front-back distance of neighbors and the turning force on their distance to the side. The study also emphasizes the need for new collective motion models that capture these interactions and for further research on how individual movement behaviors relate to emergent group-level properties. The findings have implications for understanding collective behavior in other species and for developing more accurate models of animal group dynamics.