T. rex cognition was T. rex-like-A critical outlook on diverging views of the neurocognitive evolution in dinosaurs.

A recent debate has emerged between Caspar et al. (2024) and Herculano-Houzel (2023) on inferring extinct dinosaur cognition by estimating brain neuron counts. While thought-provoking, the discussion largely overlooks the function of cognition, as well as partly neglects the difficulties involved in estimating neuron numbers, which according to us leads to oversimplified conclusions. We use this exchange as a springboard to further explore how extinct cognition might be studied and the potential pitfalls involved. One of the main emphases is on introducing basic concepts and contemporary views of cognition and its evolution. In relation to this, we highlight the shift in thermobiology during the Mesozoic-from ectothermy to endothermy-and its major impact on cognition and brain evolution. We also examine the challenges of estimating neuron counts in extinct dinosaurs based on current knowledge and take issue with several aspects of the approaches used by both Caspar et al. and Herculano-Houzel. At the same time, we challenge Caspar et al.'s claim that telencephalic neuron numbers, if estimable, would be largely uninformative about extinct dinosaur cognition, while also disagreeing with Herculano-Houzel's somewhat reductive view. We further emphasize the value of comparative cognitive studies in extant animals, alongside neural correlates, to infer the cognitive evolution of non-avian dinosaurs. We briefly outline how cognition is studied in living species and the extent to which such research can inform evolutionary inference. Our focus here is on non-avian theropods, as they are central to the current debate and belong to the lineage that led to modern birds.