Linear versus branching evolution

Jul 26 2012 Published by under Uncategorized

In the last post I pointed out that probably the most important problem with the word "evolution" is that it has two distinct meanings in biology. Transformational evolution refers to a linear set of changes in a single object, whereas variational evolution refers to a branching sequence created by unequal survival among a group of objects. Variational evolution is a process of continual branching and extinction, apparently without end, whereas transformational evolution is a simple linear process, usually leading to a predictable end. A phylogeny is principally designed to communicate variational evolution, since that is the important one in evolutionary biology.

The main point that I wish to make about this is that people seem to  easily grasp transformational evolution but not variational evolution. As David Baum & Stacey Smith have noted in their recent book (2012):
"We do not know why it should be so, but we have learned from working with thousands of students that, without contrary training, people tend to have a one-dimensional and progressive view of evolution. We tend to tell evolution as a story with a beginning, a middle, and an end. Against that backdrop, phylogenetic trees are challenging; they are not linear but branching and fractal, with one beginning and many equally valid ends. Tree thinking is, in short, counter-intuitive."

A phylogeny displays a set of lineages proceeding through time. We can trace an ancestor down to any one of many different descendants, and it is a single line from that ancestor to each descendant. This looks superficially like transformational evolution. However, the phylogeny is a whole series of these lines, inter-linked by their connections through different ancestors and descendants, and so a phylogeny actually represents variational evolution.

This is illustrated in the next figure, which is modified from one in an earlier post, showing the family tree of the current Swedish Royal Family. In this case I have emphasized a single line leading through those people who became monarch. So, even though the family tree is a branching one, with many children in each generation, we can still "see" a single lineage if we choose to. In this case that is exactly what history does, of course, since history books tend to be about monarchs rather than their their non-monarchial relatives.

The family tree of the current Swedish Royal Family, with the lineage of the monarchs emphasized.

However, even in this case the lineage of monarchy is not really linear. There is one branch (the dashed line) that leads to a dead end (an "extinction"), where the line of monarchial descent terminated because none of the offspring were alive to inherit the crown. So, even in this simple case, seeing a single transformational lineage is inappropriate.

The issue is that people fundamentally have more trouble with inter-linked and over-lapping structures, such as a branching phylogeny, compared to a single linear sequence. A linear sequence can be grasped almost instantly, whereas inter-linked structures take time. A family tree is like a telegraph pole if you trace only one lineage, but it's not really a family tree if you do that.

Probably the single most famous image in evolutionary biology is the one showing a series of primate species, each walking from left to right. This first appeared in a 1965 book by the anthropologist Francis Howell, but it has now become an integral part of modern life in many modified forms, as illustrated by this Google search. (The full-size picture, with labels, can be viewed here.)

Zallinger’s linear depiction of human evolution for the book by Howell.

In this context, human evolution seems to be presented as a transformation series. What we are apparently being told is that each species evolved into the next species in the sequence, leading ultimately to humans. That is, human evolution has been transformational, rather than variational. Equally importantly, since transformational evolution often leads to a single predictable end, we are apparently being told that humans are the "goal" of this evolutionary sequence. This has a certain psychological charm. As Sean Nee (2005) has noted: "Our persistence in placing ourselves at the top of the Great Chain of Being suggests we have some deep psychological need to see ourselves as the culmination of creation."

Biologists reject this anthropocentric view, of course. However, the image is such a powerful one, and the message is so clear and so easily grasped, that this one single image seems to dominate most people's view of biological evolution.

That many members of the general public see this image as representing "evolutionary theory" is clearly attested by scores of web pages (e.g. this one, and this one). Indeed, Laurence Smart has gone so far as the claim that: "It is found in many science and evolution textbooks, and is exhibited at museum displays about human evolution", which has unfortunately been all too true.

The diagram is apparently based on a much earlier one, which was published as the frontispiece to Thomas Henry Huxley’s book about primate anatomy (1863), which itself was a collection of Huxley's oral and written work from 1860-1862. The important thing to note is that this appeared several years after Darwin and Wallace published their works emphasizing that biological evolution is variational rather than transformational, and yet it still communicates the idea that evolution is transformational. I sometimes wonder just how much Huxley really understood Darwin's idea, even though history credits him with being one of Darwin's staunchest supporters.

Huxley's infamous frontispiece.

Huxley's image is apparently the one that took hold of both the public and the scientific imagination, rather than Darwin's variational idea. Even educational institutions such as museums used the same idea for their public displays, as claimed by Smart and as illustrated here.

A display of a series of skeletons showing the evolution of humans, at the Peabody Museum circa 1935.

Once again, these earlier illustrations apparently show a linear series, and so that is naturally how they are interpreted by the viewer. However, the same image could also be presented without the apparent linear series. The next picture shows a version as modified for Wikipedia. Note that there are three main changes: (i) The gibbon has been re-sized to match the other species (it is shown at twice its size in the Huxley original); (ii) the species have been re-ordered, notably without humans as the apparent culmination of the series; and (iii) two species have been re-orientated, so that there is no obvious sense of transformation.

Huxley's diagram as modified for Wikipedia.

It is worth noting that Huxley was a persistent menace at confusing transformational and variational evolution. In the 1870s he did it again, this time with horse evolution. The paleontologist Othniel Charles Marsh published a description of newly discovered horse fossils from North America, describing them as a sequence, with the fossil species "Eohippus" being transformed into a totally different modern descendent "Equus" through a series of clear intermediates. Huxley was very active at popularizing this inappropriate transformational series. Sadly, some years later the American Museum of Natural History assembled a famous exhibit of these fossil horses; and this story of the horse family was soon included in most biology textbooks.

What we are dealing with here is mis-communication. Both Huxley and Howell probably did understand the concept of variational evolution, but they each presented images that completely subverted that idea by showing transformational evolution, instead. This mis-interpreted image has come to dominate people's understanding of biological evolution, because a linear series is much more easily grasped than is a branching sequence. Ancestors lead to descendants both in transformational evolution and in variational evolution, but the relationships between the ancestors and descendants are quite different. In transformational evolution we can potentially see all stages of the transformation at the same time, because different objects will be at different stages of their transformation — we can, for example, see infants, children, adolescents and adults all around us, at any time. But in variational evolution, the ancestors are gone and all we can see are the descendants — gibbons are not human ancestors but instead are descendants of a (now extinct) ancestor that they share with humans.

Much of the problem with mis-communication is due to: (a) sloppy language (ambiguous use of words), and (b) sloppy diagrams (poor choice of iconography). In the absence of body language, such as when talking face to face, all we have for communication are words and pictures, and so we need to get both of them right if we are to communicate effectively. Some modern-day examples of this problem will be discussed in the next few posts.


David A. Baum, Stacey D. Smith (2012) Tree Thinking: An Introduction to Phylogenetic Biology. Roberts & Company Publishers, Greenwood Village, CO.

Francis C. Howell (1965) Early Man. Time-Life International, New York.

Thomas H. Huxley (1863) Evidence as to Man’s Place in Nature. Williams & Norgate, London.

Sean Nee (2005) The great chain of being. Nature 435: 429.


One response so far

  • Jim Thomerson says:

    Some good points here. I've seen the term patristic evolution for change within a single lineage, vs cladistic evolution which produces biodiversity. I'm a late, and somewhat reluctant convert to Hennigian cladistics. In Hennig's view, all ancestors are hypothetical. Given the long history of trying to construct ancestor to descendant lineages, it took me a while to accept that. I read of paleoanthropologists still claiming that species A is the ancestor of species B. So I have the impression that patristic thinking is still common among students of human evolutionary history.