Decades of Divergence in Dino Depiction

Many of my paleontology projects are based on the question of what we think dinosaurs looked like, and why.  We only know about dinosaurs from their fossils, which mostly only show their bones, and are often missing many pieces and are badly deteriorated.  However, when we see pictures of dinosaurs in the media, we see them as fully-realized animals with skin and colors and so on.  How have we gotten to the point where we have come up with such images?

The short answer is that some illustrators work directly with scientists to help flesh out their images, and others do plenty of primary research.  But others just base their drawings on what they’ve seen, and those images quickly become more and more divorced from the science, just like a game of Telephone.  And of course, even the original visual conceptualizations are best guesses, based on those partial fossils.

Not only that, but the science changes over time, with new ideas evolving as new fossils are found!

I have often surveyed everyday people about what they think dinosaurs looked like.  In this case, I went to the printed record, and did a morphological analysis of every image of just three dinosaurs–Tyrannosaurus, Triceratops, and Stegosaurus–that can be found among all my paleontology books.  This includes books aimed at all ages, and books that are both entertainment-oriented and scientifically-oriented.

Future versions will do the same for other popular dinosaurs.

This resulted in a huge data pool.  Here is the final analysis, which is in poster form.  It’s been reduced a bit in resolution to fit here but should be fine for these purposes.  Check it out!


Some of the earlier information, while I was still in the process of analyzing it, can be seen below.


Here is the complete data pool, which allows you to see all the characteristics I listed for each image.  I looked both at their visual characteristics (for instance, their color, or the number of plates a Stegosaurus has) as well as their contextual details (for instance, were they shown fighting with another dinosaur, or were they presented in an extinction scenario).

I did an aesthetic-cladistic analysis of the dinosaur images, as I had done earlier for the Clay-dograms, with these sample results.  Each dinosaur (Tyrannosaurus, Triceratops, and Stegosaurus) has two charts, one breaking down the morphology of their images, and the other breaking down their contextual information.

The visual characteristics:

This slideshow requires JavaScript.

The contextual characteristics:

This slideshow requires JavaScript.


I did a lengthy interactive project with many people, Skin & Bones II, where people covered plastic dinosaur skeletons in clay in an attempt to make the dinosaurs look like what the people thought the dinosaurs looked like in real life. Then I analyzed the skeletons’ morphology according to over 30 visual characteristics.

For Clay-dograms, I took all the data about the skeletons’ morphology and fed it into Mesquite, a program that paleontologists usually use to create cladograms (sets of evolutionary relationships) based on real fossils’ morphologies. I broke the data down by the species of the original plastic dinosaur, then used the program to find the visual relationships that existed among the clay models.

There is one cladogram per species. The names are shaded from dark to light, depending on how many steps a particular model is from its original “visual ancestor.”  It should be noted that these are not necessarily the only cladistic solutions to the data sets, but are one viable possibility.

This slideshow requires JavaScript.

I performed Skin & Bones II, the project the Clay-dograms are based on, on multiple occasions in Ithaca and Syracuse, NY, including a preschool, multiple elementary school classes from kindergarten to 4th grade, a high school class, and random adults at various sites including the New York State Fair.

Here are a few examples of the 300+ models that people made:

This slideshow requires JavaScript.