Abstract
Computer tomographic scanning is now a standard technique for studying the internal features of fossil structures. This enables comparisons with related modern species and speculation concerning function and even behaviour. We express here a concern that inferences about dinosaur hearing and further implications about, e.g. communication or hunting skills, are sometimes stretched beyond what can reasonably be gleaned from fossil data. We summarize current knowledge about structure-function relationships in the avian auditory inner ear and provide guidance for evidence-based inference of hearing capabilities from bony features. In particular, we point out limitations and caveats regarding inferences that are based on one isolated feature alone, typically cochlear length. As an example illustrating some of these pitfalls, we use a recent analysis (Choiniere et al. 2021 Science 372, 610-613 (doi:10.1126/science.abe7941)) that concluded that Shuvuuia deserti, a theropod dinosaur, showed pronounced sensory specializations, including 'specialized hearing acuity, rivalling that of today's barn owl'. We re-analysed the skeletal features of Shuvuuia's inner ear and argue that the analogy between hearing in Shuvuuia and the extant barn owl was based on an ill-chosen metric in assessing the relative length of the cochlear duct and a questionable assumption concerning inner-ear structure.