I will start with what the book is not about. It is not a primer on how evolution works. Apart from a brief discussion of how Darwin came to discover natural selection, there is nothing about the various ways evolution can occur. There is a bit on species definitions, where he explains the limitations of Mayr’s Biological Species Concept (the “interbreeding” one), but nothing on the various modes of speciation. It does not discuss the difference between fact and theory—in other words, it is aimed at an audience who already knows something about evolution and knows that common descent is a fact.
The purpose of the book is to tackle one, and really only one, type of pervasive misconception, one that many people hold: the misconception of the “The Missing Link”.
In order to do this Gee wanders through a lot of historical territory, examining the ways the fossil record has traditionally been interpreted, and points out that paleontology for most of its history suffered from the very human tendency to tell ourselves stories about the past. Most of paleontology and the fossil record has been couched in narrative terms: this happened, and then it led to this, and then this, and as a result we have (birds, whales, humans, etc.). For example, the history of tetrapods was, for a century or more, told in terms of fishy creatures evolving limbs in order to survive on land. The evolution of horses is one of progression towards the modern horse; increasing height and tooth crowns, decreasing number of toes. Humans stepped along in a progressive chain of increasing uprightness and brain power. A story was told of becoming; new fossil finds were slotted in to move the tale along.
And as Gee describes, as satisfying as this narrative approach is, it is a misleading, erroneous, and unscientific way of understanding the fossil record. We cannot say that this fossil was the ancestor to that fossil; we cannot even try to legitimize it by saying that this fossil represents the ancestral species represented by that fossil. As he says in the infamous quote, we cannot infer cause and effect from fossils. We do not have enough of them, we do not know if the individual that became the fossil ever reproduced or not, we cannot legitimately create a line of fossils, call them ancestors and descendants, and also call it science. It is not science because we cannot test such a claim: it cannot be falsified.
But here’s the thing: while it is not legitimate to infer lines of descent from fossils, and construct scenarios about why such and such a lineage grew legs/grew big brains, we can in fact construct testable hypotheses about the relationships of one fossil species to another. We can reconstruct the evolutionary history—the phylogeny--of any group of organisms, as long as there are more than two, and we can do this because we know that all organisms are related by common descent, even if we cannot identify any specific ancestors. We use cladistics.
Cladistics is basically an algorithm for recovering nested patterns. It can work in any system where a feature in one object gets passed down to “descendant” objects. It works in biology because of the facts of reproduction and heredity. It also is used in recovering the history of manuscripts: in the days of manual copying, mistakes introduced in one copy were often passed down to “descendant” copies.
What this tells us is that the algorithm does not rely upon any a priori assumptions about evolution, beyond the unarguable facts of basic heredity. We don’t need to know how evolution works, we don’t need to know anything about any dates on any of the objects. As Gee writes:
“Cladistics is concerned with the pattern produced by the evolutionary process; it is not concerned with the process that created the pattern, or the swiftness or slowness with which that process acted” (p. 151)
And of course we do not need to know anything about the relationships among the objects of interest: that is the purpose of cladistics—to recover those unknown patterns. The patterns are recovered by analyzing features of the objects in question, and tracking how they are passed down. A group is formed because of the features the objects within the group all share that are also present in the ancestor of the group, but are not present in other groups.
We know that traits are passed down from ancestors to their descendants, that is basic biology. We know that ancestors usually have more than one descendant, which means that each descendant has the potential to become the ancestor of a lineage that is distinct from any other of its siblings. This is how families work, and this is how common descent works. To deny common descent is to deny the reality of families. Gee’s point, though, is that when we get into Deep Time, we do not have all the members of the family we would require to construct an ancestor-descendant lineage, we only have occasional bits of information in the form of individual fossils. We know that there are ancestors, that had descendants; we do not know what those specific ancestors were. But cladistics allows us to reconstruct the relationships among the bits of information we DO have, in a way that does not rely upon invention or wishful thinking. We use available data to construct hypotheses of relationships.
Why have I emphasized “hypothesis of relationships”? Because this is critical to the entire theme of Gee’s book: Irrespective of what any individual may want to be true about the relationships, the data will determine the outcome, not a priori beliefs, wishful thinking, the research directions of a career, intuition, “common sense”, or authoritative statements. One’s hypotheses can be tested, by anyone with access to the data, or by the discovery of new fossils, or by refinements in the algorithms themselves. So if we have a bunch of tetrapod or dinosaur fossils, and The Expert believes that fossil B is a “missing link” between fossils A and D, and publishes his work claiming that there is a sequence of ancestry and descent represented by fossils A--> B --> D, this will likely get a lot of media attention, with breathless headlines about how the discovery of B has “Changed Everything We Know About ________ Evolution!!!11!!one!” (insert group of choice). And the misconception of The Great Chain of Being, or the Ladder of Progression will be reinforced, because it makes intuitive sense to us, it satisfies a deep-seated requirement to link things together and create a story of life.
And it is Wrong. Gee calls this sort of thing “Evolution by copywriter”, and “voodoo paleontology”. This was pervasive in the past, but in today’s paleontology such pronouncements are usually avoided by the scientists themselves (with some unfortunate exceptions). Rather than make breathless announcements about new Missing Links, they will reanalyze the data with the new information, testing previous hypotheses of relationships, and along the way testing any claims that A --> B --> D form a lineage. And while they can never prove such a scenario to be true, it may indeed bedisproved, or falsified.
It is theoretically possible that species A may be the ancestor to B, but that is something we can never know for certain. However, cladistics provides a way of testing such claims. Basically, if one wishes to propose that fossil A is the ancestor to fossil B, then fossil A and B must represent sister species on a cladogram—they must be each other’s closest relatives (share a common ancestor)—if they do not, then fossil A cannot be the ancestor to fossil B, no matter how much one might wish it. If A is NOT the sister to B on a cladogram, it means there are intervening ancestors between A and B.
Here is an example of this, from human evolution, a field of paleontology unfortunately prone to Voodoo Paleontology and the antediluvian notions of progression and “becoming”: we are most interested in our own story, after all. But entrenched attitudes can be hard to dig out. This example is not in Gee’s book, but from my own files.
In 1999 a new species of hominin was described, Australopithecus garhi. The discoverers provide descriptions of a relatively complete cranium and dentition, and several postcranial elements (arm and leg bones). They then make this claim: “This species is descended from Australopithecus afarensis and is a candidate ancestor for early Homo” (Asfaw et al., 1999:629). We can see how they have created an untestable story of becoming, as Gee puts it. And in this case it is the scientists themselves doing it.
But all is not lost; cladistics is becoming much more common amongst paleoanthropologists. In 2004 David Strait and Fred Grine published a massive paper that showed that the disagreements between molecular phylogenies and morphological phylogenies based on craniodental characters of extant humans and great apes largely disappeared when fossil hominins were added to the data. They also tested several proposed phylogenies of fossil hominins, including the A. garhi claim. Their consensus cladogram, utilizing a data matrix of almost 200 characters, showed that A. garhi could not be ancestral to Homo: it did not represent the sister species to the Homo clade. Therefore the claim that A. garhi was a “missing link” between A. afarensis and Homo was falsified. Asfaw et al. protested this result of course, but the point remains: their proposal was “story-telling” with no objective way to test it, while Strait and Grine made no a priori assumptions, and just let the data speak for itself. Their result, being itself a hypothesis of relationships, can be tested by anyone willing to try it.
To sum up this rather lengthy exposition:
The history of life cannot be characterized by a story of progress. There is no chain, no links, no steps on a ladder, no “becoming” something (a bird, a modern human). There are just taxa, and their relationships to each other--either close, or more distant. There are traits that are shared, by larger or smaller groups, that show us the pattern of descent with modification, and we can recover that pattern by using cladistics. We don’t have to make up stories about why it happened that particular way.
And this is what Gee was talking about in his quote. He was NOT saying that ancestors did not exist, or that common ancestry was false, or even that the relationships of one fossil to another cannot be determined. He was simply saying that inventing a chain of ancestors and descendants cannot be supported by testable data, that any such scenario is speculation. It is perfectly legitimate to say that A. garhi is a distant cousin of ours, with a few common ancestors and some closer relatives in between. It is not legitimate to say that A. garhi IS one of those ancestors: we can never know (and the evidence is against it anyway).
We can even apply this to the other quotes that Our Faithful Poster was putting up: Eldredge was not saying that horse evolution was false; he was lamenting the fact that the display showed a linear progression of ancestors and descendants. In point of fact, horse evolution is remarkably well represented in the fossil record, and is remarkably bushy and speciose. There were many different groups that lived at the same time, each with its own set of characteristics that set it off from related groups, so in the context of horses it is especially misleading to suggest that there is a single line of ancestry and descent.
