pz

(This is in response to some comments on the "Darwin vs the unfaithful" thread.)

What is the general impression everyone has of Richard Goldschmidt? It's mostly uninformed. People may know the term "hopeful monster", and associate it with the idea of extreme saltationism, in which a chance mutation leads to radical changes in form that 'hopefully' succeed. Anyone who has read Ernst Mayr knows that he portrays Goldschmidt as a clueless clod, who didn't know anything about "real" (i.e., population) genetics, and failed to recognize the central importance of population thinking in evolutionary biology.

There is a grain of truth to these pictures of the man. He did think major transformations were important to macroevolution; however, "hopeful monster" is an oversimplification of his ideas, which had significantly more depth and more evidence than is usually recognized. He was also not a population geneticist...but I would argue that his ideas were a fuzzy precursor to modern molecular genetics, and that Mayr is as much a clod for failing to appreciate the importance of that field of study.

The heart of his ideas was the systemic mutation. While most of the interest in evolutionary biology was graded change caused by small, incremental variations in genes, Goldschmidt was much more into the possibility that small changes in single genes or the conformation of the genome could lead to major changes in the pattern of expression of many genes, and cause huge changes in morphology in a single step. He was laughed at. In particular, his use of the misleading term "hopeful monster" became the easy target of a lot of mockery. This is unfortunate, because not only did he have solid evidence for the idea (and some of that evidence is so obvious that it is painful to see how it was ignored!), but the idea is now taken for granted in the concept of regulatory genes in development (although the more radical version of his idea, that systemic mutations were entirely non-genic, and could be a consequence of the spatial conformation of the chromosomes, is not accepted).

The evidence he used is apparent in the major themes of his research. One of the most obvious examples is simply the central problem in developmental genetics: all of the cells of a multicellular organism have exactly the same genetics, yet they develop into multiple, easily discriminated forms. This tells you right off that a genetic description of a cell is insufficient. Those genes have multiple, stable endpoints in their pattern of expression. Apparently, simple external signals are sufficient to launch a cell on trajectories that end in cell types as different as a keratinocyte, a hepatocyte, an oogonium, or a neuron. If this is true of a cell, why couldn't it also be true of the whole organism? Perhaps there are multiple stable patterns of potential morphological organization encoded in the genome.

And, of course, that's true. Another aspect of his research was sexual differentiation. We know this already, that the human genome contains at least two possible patterns of expression, male and female. We can freely interconvert morphologies during development with nothing but a few hormones. Goldschmidt studied the variations in sexual morphologies in the moth, Lymantria, which has more extreme sexual dimorphism than humans. Same genes, different forms: there's more to this morphology stuff than just what's on the chromosomes.

Another example was also apparent in Lymantria: metamorphosis. Same genes, even more different forms -- caterpillar vs. moth. Again, simple hormonal signals are sufficient to trigger a complex reorganization of the pattern of expression of the entire genome.

Goldschmidt was a pioneer in another aspect of genetics and teratology, the study of phenocopies. A phenocopy is an environmentally induced form that mimics a form that can also be caused by a genetic change. The message is the same as that of sexual differentiation and metamorphosis, that simple environmental cues can have huge effects on how the genome expresses itself.

Goldschmidt did have some serious scientific flaws. The last half of The Material Basis of Evolution, for instance, descends in places to the level of a cranky rant. In reaction to the criticism of his ideas, he went too far in denying the relevance of genes. He was also impolitic. Still, there's good stuff in his work, well worth exploring.

Gould predicted that "Goldschmidt will be largely vindicated in the world of evolutionary biology" in his essay, The Return of Hopeful Monsters. It hasn't quite happened yet, except perhaps in a few disciplines, but I think it's in process.

I also have to disagree with this comment about Gould's article:
Quite the contrary, The Return of Hopeful Monsters is a lucid article. I don't see how the simple message could be missed. Gould is saying that Goldschmidt did do some very good work, and focuses on the importance of rate genes, genes that regulate the rate of growth of morphological units, and that are at the center of allometric transformations. He illustrates the point with the examples of the cheek pouches of pocket gophers, paedomorphosis in salamanders, and color variants in moths. He points out that these are examples of discontinuous transformation (a more palatable term than "hopeful monster"). He explains how all this fits into the context of Darwinian theory.

I had no problem seeing exactly what he meant. Perhaps the problem isn't with Gould.

lpetrich

I do think that there has been a big problem with the "Hopeful Monster" hypothesis: it looks like the sort of hypothesis that can potentially explain anything, and therefore really nothing.

Goldschmidt and Gould had mentioned how small changes in development-control genes could have large effects, but they had not been in much of a position to go much further.

But the last few decades have seen an explosion in gene-sequencing technology, one whose consequences Gould himself had noticed in the last few years of his life. One could now sequence the genes responsible for possible "hopeful monster" effects!

And some work has recently been done in that vein.

For example, the Ubx gene of the brine shrimp Artemia allows that shrimp to grow abdominal legs, while the insect version of it does not. And recently, the leg-growth-control difference has actually been found.

Likewise, snakes do not grow front legs because their Hox-gene expression pattern has no region that indicates "grow front limbs here", as illustrated here. This subject is explained in more detail here.

To go in more details front limbs normally grow where Hoxb5 and Hoxc6 are active, but not Hoxc8. But in snakes, the Hoxc8 zone extends forward, beyond the Hoxc6 zone. So snakes lack a front-limb position, as it were.

Edited to add: Here is another nice article on evo-devo biology, as it is sometimes called, including another snake-Hox diagram.