Similarity Happens! The Problem of Homoplasy
I'll bet you think that evolution has to do with explaining how we are all related by common descent. And I'll bet you think that one of the chief pieces of evidence for common descent is homology, defined as similarity of form due to shared ancestry. This is pretty basic -- Darwin's theory is an argument from similarity.
However, biologists have known for some time that similarity is not always and everywhere the product of common descent. Organisms can display similarities of sequence, form, or life history that cannot be accounted for by their family tree. Homoplasy is the technical term assigned to such tree-jumping similarities, and "convergent evolution" is the process by which they evolved. Conway Morris has written extensively on it.
In the past, evolutionary biologists have dealt with homoplasy by ignoring it. Any trait identified as due to homoplasy was eliminated from their tree-drawing efforts. But now that we have access to DNA sequence data, we are finding more and more cases of homoplasy -- similarity in sequence or structure that can't possible be due to common descent -- similarity that jumps across trees. Phylogeneticists are urging caution, because the conflicting signals from different sequences can confuse tree-drawing algorithms. The problem is deep, and crosses all taxonomic levels -- it is not confined to just bacteria where horizontal gene transfer is common, or to the shallow branches of recently diverged species, where incomplete lineage sorting might be invoked. And guess what? They still deal with it by eliminating the supposed homoplasy from the data set.
At some point though, one has to ask what the homoplasy is telling us. Why is it there? What can account for both kinds of similarity -- homoplasy and homology? Some biologists retreat to a reflexive adaptationist position -- sequences converge because they make organisms better adapted to their environment, and sequences diverge because they make them better suited to their environment. In other words, similarity happens!
Our work suggests, however, that numerous slight successive modifications can't explain either kind of change. The neo-Darwinian mechanism (random genetic change, natural selection, and drift) can shift activity from one pre-existing form or function to another, but it is not capable of true innovation -- either the wonderful diversity of forms we see around us, or the reapplication of form needed to build eyes or wings, not once but many times, and in many ways.
What kind of process can do both? I suggest that intelligent design -- the deliberate goal-oriented repurposing of existing structures to produce useful, new functions -- is capable of generating both kinds of innovation. A designing agent can diversify an existing structure into many variations on a theme, or the agent can take a structure and rework it to give it a new purpose. Humans do these things all the time.
In future posts, we will talk about additional hallmarks of design that we expect to find in biology. Stay tuned.
Cross-posted at Biologic Perspectives.