Thursday, March 20, 2014

Anzu the feathered demon

Artist's impression of Anzu wyliei.  Credit: Mark Kingler, Carnegie Museum
of Natural History.  Image source.
The Carnegie Museum of Natural History in Pittsburgh, Pennsylvania is a museum I have visited many times as a child.  For the last decade, the museum's Dinosaur Hall has included a skeleton of a strange, crested, ostrich-like dinosaur that was 11 feet from nose to tail.  Its odd and terrifying appearance led museum employees to nickname it the "chicken from hell."  Yesterday, it was finally given the scientific name Anzu wyliei, named after Anzu, a Mesopotamian feathered demon, and Wylie Tuttle, the grandson of one of the financial supporters of the Carnegie Museum.

The bones of Anzu were found in the Hell Creek Formation in North and South Dakota, which probably contributed to its nickname.  The Hell Creek Formation (also called the Lance Formation in Wyoming and other places) is a very complicated but very dinosaur-rich package of rocks, thought to contain the last of the dinosaurs before their great extinction at the end of the Cretaceous period.  Dinosaurs such as Tyrannosaurus, Triceratops, and Edmontosaurus are frequently found there.

Anzu is a type of oviraptorosaur, a group of dinosaurs related to the therizinosaurs (mentioned previously) and the dromaeosaurids, which include Deinonychus and Velociraptor.  While many paleontologists consider dromaeosaurids to be direct "siblings" to birds on the evolutionary tree, placing oviraptorosaurs as their cousins, others have suggested that oviraptorosaurs are actually the first true birds.  It is no surprise, then, that Anzu is depicted as having feathers.  Is this depiction justifiable, though?

The question of feathers on dinosaurs is a tricky one.  To some extent, the concept of feathered dinosaurs is motivated by the evolutionary claim that modern birds descended from dinosaurs.  Nevertheless, some dinosaurs, including oviraptorosaurs, have been found with preserved feathers, as noted by the authors of the latest study.  Although Anzu itself was not found with preserved feathers, it is reasonable to think that it may have had a few.  However, we must be careful about what is meant by "feather."  Many of the structures found on dinosaurs might be better described as branching hairs, but they are called "feathers" because of their supposed evolutionary relationship to bird feathers.  If dinosaurs were thought to have evolved into mammals rather than birds, these same structures would undoubtedly be called "hairs" in the scientific literature.  The dinosaur "feathers" are not at all structured for flight, and paleontologists acknowledge this.  Their exact use remains somewhat of a mystery, though in my opinion they seem purely decorative on many dinosaurs.

If dinosaurs had feathers, does that mean they evolved into birds?  Of course not!  For one thing, the modifications required for flight are extensive and go far beyond having feathers.  Paleontologists are still trying to figure out how that could have happened.  Besides that, I think it is quite evident that the classic (Linnean) method of dividing up organisms into distinct classes may not quite reflect the true nature of life in either evolutionary or creationist views.  While the Bible does state that God created discrete kinds, these kinds appear to be distributed across a spectrum rather than set in separate categories.  This leaves us with some strange overlaps, such as egg-laying mammals, insect-eating plants, and feathered dinosaurs.  If creationists attempt to stick with the traditional categories, they may find the transitional forms to be more and more convincing of evolution rather than of created variety.

While I do find the modern classification system convenient, it is being progressively hijacked by evolutionary thinking.  Phylogenetic analysis measures the differences between organisms, then it forces those differences into an evolutionary tree.  What if instead those measured differences were placed in a multidimensional plot?  Organisms could be organized strictly based on their similarities and differences, rather than on their claimed evolutionary relationships.  This could even be done from a genetic basis.  Though such an undertaking would be quite extensive and intensive, I suspect that it would give a clearer picture of what the created kinds were and how modern organisms are truly related.


  1. Interesting article. I'm glad you addressed the feathered-dinosaur in the room, I've been wondering about that. When you mentioned the possibility of multidimensional plotting of morphological traits, it reminded me of phenetics, or numerical taxonomy. From what I understand (and remember from the class I took), phenetics attempted to remove user bias and subjectivity from animal classification by applying numerical values to different traits, giving every trait an equal weight. The problem with this was that even through this attempt they introduced their own forms of bias. Although I will say that it seems like it would be worth examining for the methodology. Phenetics was largely replaced by cladistics (or phylogenetic systematics) in the 80s.

    1. Ah, yes, that appears to be quite similar to what I was proposing! Thanks for pointing that out! I had never looked into phenetics before. I do think it is a good idea in theory, but as you say, it is difficult to implement practically due to bias and complexity.