Dinosaurs Evolve—and Fly 277dinosaurs that could be related to birds had clavicles or collarbones, yet all birds have their
clavicles fused into the “wishbone,” which serves as an important “spring” in the flight
stroke. (Clavicles have since been found in a number of theropods, removing this objection;
they are just delicate and rarely preserved). Of course, this is pure ancestor worship. Heil-
mann’s preferred candidates (such as Euparkeria) among the archosaurs are simply primitive
with respect to all dinosaurs. In the process, he ignored all the derived similarity between
dinosaurs and birds that Huxley had demonstrated, largely because most paleontologists
thought of dinosaurs as big and specialized and couldn’t imagine birds originating from
these huge land animals. Heilmann’s thinking was also dictated by scenarios about how
birds evolved flight by gliding from branch to branch (the “trees down” hypothesis), and
large terrestrial dinosaurs don’t fit that scenario.
The “birds as dinosaurs” hypothesis remained unfashionable until the 1970s, when Yale
paleontologist John Ostrom (a good friend of mine, he died in 2005) looked at the specimens
in Europe again. He found one specimen in the Teyler Museum in the Netherlands that had
been misidentified as a pterosaur in 1855, but when he looked closer, he saw faint feather
impressions and knew he had another specimen of Archaeopteryx. Meanwhile, a specimen
in the Eichstatt Museum found in 1951 had been misidentified as the Solnhofen dinosaur
Compsognathus, until F. X. Mayr found feather impressions on it 20 years later. The fact that a
small dinosaur and Archaeopteryx could be so easily mistaken for each other was a revelation
for Ostrom. He revived Huxley’s hypothesis and added a long list of evidence to support it
further. He was influenced by the fact that in the 1960s he had discovered and described the
highly specialized dinosaur Deinonychus (misidentified as “Velociraptor” in the Jurassic Park
movies), which shows amazing anatomical similarities to the earliest birds.
After Ostrom’s initial papers, the controversy over the “birds are dinosaurs” hypothesis
raged for several years but quickly resolved because the evidence soon became overwhelm-
ing. Hundreds of specialized shared-derived characters support the hypothesis (Gauthier
1986), and there are no competing hypotheses with even a fraction of that support. All but
a tiny minority (less than 1 percent) of paleontologists are convinced by the data, especially
with the discovery of feathered nonavian dinosaurs from China that we shall discuss shortly.
Those few who insist that “birds are not dinosaurs” have no competing hypothesis with any
strong support, and their case consists largely of trying to nitpick individual characters that
occur both in birds or dinosaurs. They do not address the overwhelming evidence of most of
the characters. Indeed, they don’t use or seem to understand cladistics, which is part of the
problem. Another part of the problem is that they are wedded to the “trees down” hypoth-
esis of the origin of flight and cannot imagine terrestrial dinosaurs evolving flight from the
“ground up.” Ken Dial (2003) showed that their emphasis on the “trees down” origin of
flight is misguided. Many birds, such as chukar partridges, use the lift of their wings to help
them run up steep inclines but seldom use them for flight. It is easy to see how dinosaurs
(which already had feathers for insulation, as we will see later) could have adapted these
structures to make climbing steep inclines easier, and from there began to develop short
glides and eventually true flight.
But evolutionary scenarios must not drive the analysis, and scientists still have to
abide by the rules of science and provide positive evidence and well-supported alternative
hypotheses, not “just-so” stories. As detailed by Norell (2005:215–229), their attacks on the
widely accepted hypothesis mostly amount to ridiculous and often self-contradictory snip-
ing, without proposing an alternative hypothesis of their own. In that respect, they resemble