142 Evolution and the Fossil Record
vertebrae of the backbone. Bit by bit, however, more biologists are coming to terms with the
cladistic revolution in systematics, accepting the results, and learning to use less familiar
but natural groups such as “amniotes” and “tetrapods” rather than “reptiles” and “amphib-
ians” as traditionally used. I am proud to say that my historical geology textbook (Dott and
Prothero 1994) was the first to introduce cladistics to its textbook market and to avoid para-
phyletic groups, and many other textbooks are beginning to catch on to what professional
systematists had accepted more than 30 years ago.
Ancestor Worship
That a known fossil or recent species, or higher taxonomic group, however primitive
it might appear, is an actual ancestor of some other species or group, is an assump-
tion scientifically unjustifiable, for science never can simply assume that which it has
the responsibility to demonstrate. . . . It is the burden of each of us to demonstrate the
reasonableness of any hypothesis we might care to erect about ancestral conditions,
keeping in mind that we have no ancestors alive today, that in all probability such
ancestors have been dead for many tens of millions of years, and that even in the fos-
sil record they are not accessible to us.
—Gary Nelson, “Origin and Diversification of Teleostean Fishes”Fossils may tell us many things, but one thing they can never disclose is whether they
were ancestors of anything else.
—Colin Patterson, EvolutionSome aspects of cladistic theory have proven more difficult for many scientists to accept. For
example, a cladogram is simply a branching diagram of relationships among three or more
taxa. It does not specify whether one taxon is ancestral to another; it only shows the topology
of their relationships as established by shared derived characters. In its simplicity and lack
of additional assumptions, it is beautifully testable and falsifiable, so it meets Popper’s cri-
terion for a valid scientific hypothesis. The nodes are simply branching points supported by
shared derived characters, which presumably represent the most recent hypothetical com-
mon ancestor of the taxa that branch from that node. But strictly speaking, cladograms never
put real taxa at any nodes, but only at the tips of the branches.
Many scientists, however, would like to say more than just “taxon A is more closely
related to taxon B than it is to taxon C.” Instead, they would draw relationships with one
taxon being suggested as ancestral to another. This is the more traditional family tree type
of phylogeny, which not only suggests relationships, but shows a pattern of ancestry and
descent as well. But as Tattersall and Eldredge (1977) point out, a family tree makes far more
assumptions than does a cladogram. Some people are happy to make those assumptions, but
the strict cladists are not so comfortable with them.
The biggest sticking point is the concept of ancestry. We tend to use the term “ancestor”
to describe certain fossils, but we must be careful when making that statement. If we want to
be rigorous and stick to testable hypotheses, it is hard to support the statement that “this par-
ticular fossil is the ancestor of all later fossils of its group,” because we usually can’t test that
hypothesis. Because the fossil record is so incomplete, it is highly unlikely that any particular