Living in purgatory, I am finding myself getting drawn very strongly to the intellectual history of biology. Having been pulled into studying population genetics because of my interest in understanding the molecular revolution in paleoanthrology, I got drawn first to phylogenetic systematics, then to macroevolution (Jablonski is fascinating), and finally to homology. It is as if my journey were recapitulating the intellectual history of the biological sciences itself. I want to write about the stunning revival of homology over the last few decades. Having been banished in oblivion by the modern synthesis — ‘case closed’ — and derided for the better part of the twentieth century, homology is back baby! But more on that soon enough. For now I thought I should share a particularly insightful excerpt from Günter P. Wagner (2014), Homology, Genes, and Evolutionary Innovation. Although couched as ‘a case for conceptual liberalism,’ it can be read as a call to put forward strong theories, for only they admit systematic progress — an attitude that is severely lacking in the humanities.
A Case for Conceptual Liberalism (p. 78-79, emphasis mine.)
Concepts are mental tools, and tools are made to perform a given set of tasks. A hammer is made to exert strong forces on an object, either to drive a nail into wood or to smash a rock. A concept is also a tool. Its purpose is to bring order to a set of observations and to guide further study and practice (e.g., experimentation and technology). Even with the best tool, we should resist the temptation to overuse it (e.g, if we have a hammer, we should not consider all of reality as nails). This is particularly true for biological concepts, which always oversimplify reality. … It would also be foolish to abandon the species concept only because there are cases for which two populations are neither fully part of one species nor clearly separate species.
To the contrary, recognizing these situations in which the species concept does not fully apply leads us to a deeper understanding of the process of speciation. Studying populations that are in the process of speciation helps us understand how speciation works. Furthermore, these situations would not be recognized as worth studying if they did not stick out against the foil of the biological species concept. Hence, critical evidence regarding the process of speciation becomes recognizable only because it does not fit the species definition. Hence, in a paradoxical way, the situations that deviate from the species concept validate this concept because they help identify cases that deserve further study and from which we, in fact, learn a lot about the biology of populations.
Concepts, theories, and models are most valuable because reality does not always conform to them. This is a basic principle of scientific investigation. Scientific evidence for novel facts arises from discrepancies between well supported theories and observation. Small deviations from the predictions of the Newtonian model of planetary motion provide evidence for the theory of special relativity. Without the Newtonian model, these observations would have no meaning and would not even be “deviations” from anything and could not lead us to deeper insights. Similarly, albeit less glamorous, is the use of the neutral theory of evolution that proposes that most molecular sequence change is due to drift to fixation of selectively neutral mutations. This model makes strong predictions about the rate of nucleotide substitutions over time (constancy of the rate regardless of population size; relationship between within-population variation and divergence between populations). Of course, everyone knows that real sequence evolution does not always conform to these predictions. But the reason we still teach this model is that it provides a yardstick to detect biologically significant facts. For example, the most widely used methods to detect natural selection at the molecular level work by identifying statistically significant deviations from the neutral model. In all the sciences, strong evidence drives discovery with models and discovers new things by finding discrepancies from these models. This is not the same as falsification, in which one discards a model once deviations from the model are discovered. In contrast, models, concepts, and theories that reflect some aspects of reality, but fail in others, are essential tools for scientific discovery.