Richard Sternberg Evolutionary Biologist

Current Research

The goal of my research is to better understand the causal relationships between the genotype and the phenotype, or DNA specifications on the one hand and the morphological “groundpattern” on the other.

It has long been presumed that the four-dimensional ontogeny of an organism is encoded in the one-dimensional nucleotide strings, which are commonly referred to as “genes.” But now any quasi-direct genome segment  homology mapping seems unlikely for a number of empirical reasons. To begin with, we currently lack an adequate definition for the term gene. The layers of codes that reside along a single DNA sequence and that enable hundreds of different RNA and protein products to be encoded, the interleaved organization of loci, and the fact that most genomic expression involves “non-coding” regions — to mention but a few — preclude the classical meaning of the word “gene.” So one cannot speak with precision of a gene homology relation, in the sense of “master genes” or “DNA blueprints” or “genetic programs” that determine cell differentiation, without first clarifying what the “gene side” of that relation stands for. Then too there are the many modifying and editing steps that transform an RNA sequence into a product that only partly resembles its DNA substrate, not to mention the translational and post-translational processing or proteins. Aside from dampening the DNA “signal,” each of these events has the potential to allow the operation of cellular sources of “information,” or additional arrows in the genome segment homology mapping. Moreover, it is now clear that anatomical and developmental characters shared by two taxa may nonetheless have separate genomic underpinnings. One interpretation of this observed drifting of the phenotype independently of the genotype is that DNA sequences only specify the building materials of bodyplans, not the groundpattern. That is to say, the gene homology relation may only extend to the RNA and protein levels. There are of course many other empirical problems with any strong version of the gene-to-homology mapping.

To address this complex topic, I am trying to identify formal rules that underpin morphological states, meaning generative functions that entrain DNA specifications. The second part of my research is the identification of ontogenetic events where “informational buffering” may occur — where outputs at one level are “read” by higher developmental processes as “data.” And third, as a corollary of these two endeavors, I am testing various self-organizational models to see if these can account for the origination of novel organ complexes.