Symbolic artifacts present a challenge to theories of neurocognitive processing due to their dual nature: they are both physical objects and vehicles of social meanings. While their physical properties can be read of the surface structure, the meaning of symbolic artifacts depends on their embeddedness in cultural practices. In this study, participants built models of LEGO bricks to illustrate their understanding of abstract concepts. Subsequently, they were scanned with fMRI while presented to photographs of their own and others’ models. When participants attended to the meaning of the models, we observed activations associated with social cognition and semantics. In contrast, when attending to the physical properties, we observed activations related to object recognition and manipulation. Furthermore, when contrasting own and others’ models, we found activations in areas associated with autobiographical memory and agency. Our findings support a view of symbolic artifacts as neuro-cognitive trails of human social interactions.