One of the most important insights Charles Darwin had was that evolution does not follow a pre-ordained path. There is no evolutionary endpoint that organisms are striving toward. The "endless forms most beautiful" we observe in nature are both shaped by adaption to local conditions and constrained by the contingencies of their history, and it is impossible to predict what life might look like 1 million, 10 million, or 100 million years from now.
But paleontologist Simon Conway Morris begs to differ. He argues that some evolutionary outcomes are inevitable. If this were true then it could be argued that our species had been planned all along, thus allowing for a closer correspondence between cherished religious beliefs and what we know about nature.
Conway Morris makes his case primarily through identifying evolutionary convergences, a natural phenomenon in which two unrelated organisms independently evolve the same trait. If a particular trait has evolved multiple times, Conway Morris argues, then it can be treated as an inevitable outcome of evolution which therefore suggests (in his view) that life is being pulled in a particular direction by some supernatural force. In his latest exposition of this idea, published in the journal Naturwissenschaften, Conway Morris appeals to the evolution of birds to help support his thesis.
According to Conway Morris, "birds" evolved at least three times. In addition to the earliest recognized bird Archaeopteryx, the four-winged dromaeosaur Microraptor and its relative Rahonavis might be considered dinosaurs that independently acquired some degree of flight. For Conway Morris the convergent evolution of flying dinosaurs several times means that evolution is following a predictable pathway; if birds did not evolve from one lineage of feathered dinosaurs then they certainly would have evolved from another.
But there are some severe problems with this interpretation. First, it is still not entirely clear how Archaeopteryx, Microraptor, and Rahonavis flew, if they could fly at all. They were all small, bird-like dinosaurs that possessed what we call "flight feathers," but this does not mean that they all flew or flew the same way. They may have been gliders rather than fliers, especially Microraptor, and while each is relevant to understanding the origin of birds we cannot honestly interpret each as definite evolutionary step towards birds of today.
Indeed, feathers and other "bird" characteristics were widely shared among dinosaurs that were not ancestral to birds. Birds evolved only once, and the alternative lineages Conway Morris sees as anchored to Microraptor and Rahonavis never came to be. He does not address why this might be so, but it has everything to do with the complementary roles of contingency and constraint in evolution.
In many cases, the dinosaurs that possessed "avian" traits such as feathers could not be easily modified into flying creatures. Many were too big, had the wrong kind of feathers, or just did not have a lifestyle in which gliding or flying would have been advantageous to survival and reproduction. Little quirks of evolutionary history and ecology constrained how feathered dinosaurs could be modified from one generation to the next, and it was in only one lineage, by chance, that the circumstances caused the first birds to evolve. Likewise, the fact that birds survived the mass extinction at the end of the Cretaceous could not have been predicted beforehand. Had they perished, we probably would not be spending so much time talking about them now as they would be, in our biased view, just another evolutionary dead end.
Despite all this, however, Conway Morris concludes: "Avian theropods are, therefore, pre-ordained." I cannot help but think of this as a case of confirmation bias. In battling a straw man, "ultra-Darwinian" view, Conway Morris treats every case of convergence as evidence of inevitability in evolution, thereby ignoring the roles of contingency and constraint in shaping the tree of life. If a group of dinosaurs of similar body shape inherits feathers from a common ancestor, for example, then of course it would be expected that some of them might be adapted in similar ways given their shared characteristics. The resulting convergences would not be the result of inevitability, but based upon the constraints of surviving and reproducing given a common starting state.
Evolutionary convergence is a real pattern, and an interesting one at that, but it has to be understood as being couched within evolutionary history. It is not productive to simply cherry-pick occurrences of traits evolving more than once and then state that it was all meant to be.
At a grand scale, though, I think that the wider diversity of dinosaurs undercuts Conway Morris' argument for evolutionary inevitability. Dinosaurs have been around for over 230 million years, and during their time on earth they have diversified into a wide array of unique forms that generally have not been duplicated by mammals. There have been some instances of convergence, as between the armored ankylosaurs and the hard-shelled mammals called glyptodonts, but you would think that if evolution was proceeding in a fore-ordained direction most dinosaurs would have made the "next step" to whatever Conway Morris believes should have come after them. Yet no such signal is readily visible. Hence the evolution of dinosaurs (and all other life) is better viewed through Darwin's perspective, and I think the notion that dinosaurs were not predestined makes them all the more fascinating.