It's not the fall that kills you...it's the sudden stop at the end
In an apparent effort to take low-confidence science to new heights, John Kappelman and colleagues recently published an article to address the question: "How did Lucy die?" ("Perimortem fractures in Lucy suggest mortality from fall out of tall tree." Nature 2016: 537; 503-507).
Lucy is the most famous and well-preserved specimen of Australopithecus afarensis, believed to be an ancient hominin and human evolutionary ancestor. Chapter 6 of The Scientific Approach to Evolution explains how only very-low-confidence science can suggest that Lucy was an evolutionary ancestor to humans. Chapter 4 discusses the cause of death of King Tut, explaining that only very low confidence science can address this question. So, it should come as no surprise that the cause of death of Lucy can only be addressed by very low confidence science.
Kappelman argues that Lucy died by falling out of a tree. He compares the nature of Lucy's bone fractures to modern bone fractures that resulted from falls. Seeing similarities, he concludes that falling is the most likely cause of Lucy's fractures. This leads to an incredible succession of increasingly unsupported speculation, aimed at encouraging a belief that Lucy's is an evolutionary ancestor to humans. His argument goes like this:
1. Lucy has many broken bones, broken in ways that resemble trauma from falling.
2. Therefore, Lucy fell to her death
3. There is evidence of trees in Afar during Lucy's era
4. Therefore, Lucy fell from a tree
5. This is additional evidence that Lucy lived in trees
6. Because Lucy was bipedal, intermediate between humans and tree-dwelling primates, she was clumsy in trees, which explains why she fell.
7. All this reinforces the belief that Lucy is an evolutionary ancestor to humans.
John Hawks, a professor of anthropology at the University of Wisconsin, offers an important critique:
He calls out the need to compare Lucy's fractures to those of other fossils. He suggests that elephants, rhinos, and horse fossils can show similar bone fractures, but are unlikely to have fallen out of trees. He also suggests that the great extent of Lucy's bone fractures would suggest that she fell out of an airplane, not a tree. He speculates that these fractures may have occurred after death.
My interest lies in the low-confidence nature of this science. Kappelman's work clearly has little appreciation for the 6 criteria of high-confidence science:
1. Repeatable: Lucy's death cannot be repeated.
2. Directly measurable and accurate results: Measurements on Lucy's bones (supposedly 3.18 million years old) are a very indirect way to address cause of death.
3. Prospective, interventional study: The study of bones that are believed to be 3.18 million years old certainly counts as retrospective. The nature of the work is observational, not interventional. The door is wide open for bias.
4. Careful to avoid bias: The strong desire to support Lucy as an evolutionary ancestor to humans is a bias that is evident throughout the work.
5. Careful to avoid assumptions: As John Hawks points out, showing similarity of Lucy's fractures to modern fractures that result from falls does not imply that Lucy fell. Kappelman assumes that it does.
6. Sober judgement of results: Although Kappelman does include a sprinkling of hedging terms like "suggest" and "probably", the nature of this work requires a much more generous application of hedging terms. Kappelman's publication does not include a "limitations" section, where he could candidly discuss the limitations of his study. Is this too much to ask? Finally, by extending his sketchy findings to suggest support for an arboreal existence of Lucy and connection to human evolution, Kappelman has clearly amplified his findings well beyond the scope of his observations.
Because the study is based on very-low-confidence science, the influence of bias can easily overpower the influence of the sparse data. As a result, the subjective conclusions will be an endless source of debate.