Tom stepped into another scientific controversy when he proposed that the small dinosaurian ancestors of birds evolved wing feathers, not to fly, but to brood their young. Flying, according to his theory, came later. Article pdf

The relationship between birds and dinosaurs has become obvious in light of fossil discoveries. Around 100 million years ago animals existed that were feathered, but lacked the long wing feathers modern birds use for flight. Theorists were divided into two camps over how non-flying dinosaurs evolved the long feathers that enabled them to take to the air. On one side, the "arborialists" hypothesized that tree-dwelling dinosaurs evolved flight by gliding down from trees. The "cursorialists" postulated that flight arose when small, two-legged running (cursorial) dinosaurs somehow evolved feathers capable of lifting themselves off the ground.

Each side of the controversy had its merits and detractors. Key weaknesses in both theories involve the first stages: how could animals without wings begin jumping from trees without being injured or killed? On the other hand, why would ground-dwelling creatures begin flapping their arms when feathers were too short for liftoff?

Tom weighed both arguments and after thoroughly researching knowledge in the field, he and scientific illustrator Mark Orsen developed a new version of the old cursorial theory: the Brooding-To-Flight Hypothesis.

In this concept, small, two-legged running dinosaurs, which are known from fossil evidence to have had short feathers all over their bodies, evolved long feathers on their forelimbs, not to fly at first, but to brood their young. Over time, longer and longer feathers evolved, first to augment the brooding function, and then when the feathers were long enough, to augment the animal's airborn time when it leaped up from the ground to avoid enemies or while hunting.

This theory has attracted interest from paleontologists because it is the first to offer both a rationale as to why the feathers lengthened, and also a scenario in which gradual lengthening makes sense: any small improvement to the ability to shelter offspring would offer a survival advantage, whereas both the ground-up and trees-down flight hypotheses fail to explain the early stages, because short feathers are useless for flight.

Although Tom's theory has yet to gain wide acceptance in the field of paleontology, it was endorsed by the originator of the cursorial theory, John Ostrom, and was included in the book, Feathered Dragons: Studies of the Transition from Dinosaurs to Birds. Short abstract pdf