Systematics and Evolution 143fossil in our collections is the remains of the actual ancestor of another taxon (Schaeffer et al.
1972; Engelmann and Wiley 1977).
But there’s another reason why cladists avoid the concept of ancestry. To be a true ances-
tor, the fossil must have nothing but shared primitive characters compared to its descen-
dants. If it has any derived feature not found in a descendant, it cannot be an ancestor.
Consequently, for decades, traditional taxonomists looked only at shared primitive charac-
ters so they could construct ancestor-descendant trees, thereby missing all the derived char-
acters that showed they were on the wrong track. One of the great advantages of cladistics
is that it has solved many previously insoluble problems by getting away from paraphyletic
wastebasket groups and “ancestor worship” and focusing on derived characters only. For
these reasons, hard-core cladists like Gary Nelson (quoted earlier) refuse to recognize the
concept of ancestor at all, except in the hypothetical sense of the taxa at the nodes of the
cladogram. Instead of ancestor and descendant, cladists prefer to talk about two taxa at
the tips of the branches as being sister groups. Neither is ancestral to the other, but they are
each other’s closest relatives.
But there are circumstances where the fossil record is so complete that it is possible to
say that “the fossils in this population represent the ancestors of this later population.” My
friend and fellow former graduate student Dave Lazarus (now a curator at the Museum für
Naturkunde in Berlin) and I (Prothero and Lazarus 1980) provided just such an example
from the extraordinary fossil record of planktonic microfossils. In these unusual circum-
stances, we have deep-sea cores covering all of geologic time since the Jurassic for most of
the world’s oceans and every centimeter of sediment in most of those cores is filled with
thousands of microfossils. With an extraordinarily dense and continuous record such as this,
we really can say that we have sampled all the fossil populations that lived in the world’s
oceans and can establish which samples are most likely the ancestors of later populations.
Since our paper, a number of studies have been done to establish how complete the fossil
record needs to be to determine the probability that one population is ancestral to another
(Fortey and Jefferies 1982; Lazarus and Prothero 1984; Paul 1992; Huelsenbeck 1994; Fisher
1994; Smith 1994; Clyde and Fisher 1997; Hitchin and Benton 1997; Huelsenbeck and
Rannata 1997). Nowadays, paleontologists are a lot more relaxed about the concept of ances-
try than they were during the early, bitterly polarized debates over cladistics in the 1970s.
Most paleontologists use the word ancestor (as I will throughout this book) very loosely to
describe a fossil that has all the right anatomy and is older in time to potentially be ancestral
to some later form. But we all recognize subconsciously that, in the strictest sense, telling
whether a particular fossil is actually the ancestor of another is not a testable hypothesis.
Instead, we look to fossils to show us the transitional anatomical features of ancestors that illus-
trate the path that evolution took.
Naturally, this debate, which revolves around subtle philosophical distinctions about
ancestry, has been a bonanza for the quote-mining creationists. They pull dozens of quota-
tions (like the one from Gary Nelson) completely out of the context I have just described
and claim that these statements show that there are no such things as ancestors in the fossil
record. (Prothero and Lazarus [1980] disproved that decisively.) The debate was all about
whether we could tell whether a particular fossil could be recognized as an ancestor and how to do
phylogeny, but even the most hard-core cladists do not doubt that ancestors existed! None of
the debate is about whether life has evolved. After all, what would be the point of doing a
cladogram, which is a phylogeny, if you didn’t accept the fact of evolution?