« Multilevel priors with no hierarchical data? | Main | Gap in Life Expectancy Widens for the Nation »
March 22, 2008
The great chain of being
In his talk on mental models of the structure of the world, Josh Tenenbaum talked about how people think of animals as being classified in a tree structure, and how this structure might differ from those implied by different scientific models. This kind of thing:
Anyway, as an aside, Tenenbaum pointed out that, although the tree structure seems so natural to us, it doesn't have to be this way. He noted that, traditionally, creatures have been organized into a linear "great chain of being" rather than as a tree structure. Then I realized . . . that's how I think of the animal kingdom. It's how we learned things in 9th grade biology. At the bottom are single-celled animals (amoebas and so forth), then gradually through the invertebrates, then the vertebrates, starting with the fish (with the sharks at the bottom because of their primitive structure), then amphibians, then reptiles (amphs are lower than reps because of being more fish-like and primitive, I think), then birds (higher because they're warm-blooded), then animals, with primates at the top and, well, you know what's the #1 primate . . .
Anyway, only when sitting in Tenenbaum's talk did I realize that I'd swallowed this whole great-chain-of-being formulation without even thinking about it. The assumption is that every invertebrate is lower than every vertebrate, that the most complicated bird is lower than any mammal, that all plants are lower than all animals, etc. It's still hard for me to shake this mode of thinking.
I guess it's a good thing I'm not a biologist. (I did publish in the Journal of Theoretical Biology once, but we all know that knowledge of biology is not necessary to publish in that journal.)
Posted by Andrew at March 22, 2008 3:30 PM
Trackback Pings
TrackBack URL for this entry:
http://www.stat.columbia.edu/~cook/movabletype/mt-tb.cgi/1495
Comments
People often misinterpret evolutionary trees.
T. Ryan Gregory, the author of the Genomicron blog, has written an excellent review, “Understanding Evolutionary Trees”. This paper, which appears in the new journal Evolution: Education and Outreach, is available free online through 2008.
Dr. Gregory writes:
The arrangement of the tips is unimportant..., so long as the branching patterns are maintained. In this respect, phylogenies are like a baby’s mobile: every internal node can be rotated without any implications for the pattern in which the branches are connected.
Dr. Gregory provides many good examples in his paper of how to avoid misinterpreting evolutionary trees.
Posted by: Conrad Halling at March 23, 2008 8:40 PM.
It's not necessarily a "tree" structure, as exactly the same mental model can be visualised as a "boxes within boxes" structure. Darwin's famous tree sketch is one way, but I feel sure that Linnaeus, who invented the biological naming system we still use today, used drawers as his mental picture.
The graphical display technique called "treemap" also shows this ambiguity of possible views.
Posted by: derek at March 24, 2008 11:34 AM.