Bringing dinosaurs to life: A scientific revolution in palaeobiological methods

Here, we explore the spread of hypothesis testing in modern palaeobiology in the past 60 years. Our focus is dinosaurs, and we show how standard techniques in functional biomechanics, ecology, and evolution are now applied even-handedly to modern and fossil taxa to supplant the earlier, more speculative approaches. This new, analytical approach combines what we term neontological toolkits , sets of established observations and rules from the modern world that we can apply to past situations, together with the extant phylogenetic bracket, the definition of a wider clade within which the fossil taxon of interest resides. Earlier suggestions by philosophers that the ‘historical sciences’ like geology, palaeontology, archaeology, and astronomy are not open to testing are refuted, as is the idea that palaeontology must devote itself solely to describing one-off patterns and not seeking generalisable hypotheses or testing processes. Underdetermined some of it may be, but palaeobiology can still go a long way in reconstructing the evolution, ecology, function, and behaviour of ancient organisms, as we show from examples in the evolution, ecology, and function (feeding, locomotion, senses, growth, thermophysiology) of dinosaurs. • There has been a revolution in testability of palaeobiological inferences • The shift in research mode began with a classic work by R. McNeil Alexander in 1976 about how to calculate dinosaur running speed • Since then, similar modes of research have been adopted in inferring many details of feeding, locomotion, senses, growth, thermophysiology, and even colour of dinosaurs • We term the package of testable assumptions from modern biology, the neontological toolkit