A Manifesto for Evolutionary Systematics
Richard H. Zander
December 23, 2010
A Manifesto for Evolutionary Systematics
This is a call for serious reappraisal of modern systematics. Do you remember when there was a “paradigm change” some decades ago, and we taxonomists were supposed to reject theories as just-so stories, redefine evolution as relationships on a cladogram, reframe evolution as changes in traits not taxa, abandon scientific induction in favor of hypothetico-deductivism, and revile as mere “intuition” the results of 250 years of taxonomic expertise and discursive reasoning? It is time to consider a total rejection of structuralism hiding in plain sight in systematics.
Phylogenetic analysis extracts from homologous character data sets patterns of evolutionary relationships, just as all basic evolutionary evidence is based on similarity. The software-generated patterns are, however, neither hypotheses nor theories because they do not directly detail possible causal relationships of natural processes that can be investigated and falsified, or supported by additional facts. These patterns are evidence of processes involved in the natural generation of species and higher taxa for a particular group. Interesting and helpful evidence, but only evidence.
Phylogenetics is not science. Phylogenetics rejects forming theories using evidence and induction. Instead it treats patterns of evidence of macroevolution as the phenomenon of evolution itself, and uses deduction alone to form what are essentially apodictic theorems, avoiding theories of process in nature. Following Thomas Aquinas, when one has the correct first principles, all deductions must be correct.
This is structuralism, a “content-free” methodology common in linguistics, anthropology, psychology, and other areas, intended to replace empiricism. Structuralism attempts to introduce an element of “hard science” into non-mathematical and non-physics-based fields. This is done by identifying basic patterns that may be taken as axiomatic or irreducible first principles. In phylogenetics, the structure of the dichotomous tree is “saved” by rejecting all theory-based ancestor-descendant relationships in favor of a forced sister group parenthetical nesting of exemplars. Internal consistency of the method is aided by administrative enforcement of the principle of holophyly (strict phylogenetic monophyly). The statistical certainty or near-certainty of patterns of nested exemplars in molecular systematics is “saved” by ending analysis at the level of the nested exemplars, rejecting induction of process-based theories of ancestor-descendant relationships from those evidential patterns and from other data.
Phylogenetic “testing of hypotheses” is simply comparing patterns of evidence. Even if morphologically and molecularly derived patterns are “consilient,” they are themselves not hypotheses but are only patterns of evidence. If such patterns differ, there is actually information in that difference because the patterns are not theories, that is, there is no comparing of alternative processes. The differences, in fact, may be used to generate process-based theories. Phylogenetic analytic methods of clustering are innocent and can be powerful, but the interpretation of the results (as pattern equals evolution and classification must follow such patterns) is not just flawed but is non-science.
Structuralist thinking necessarily eliminates any reflection of macroevolution in classification. The “tree” of life has no scientific realism or theoretic substance because nodes are not diagnosably named, and the dendrogram is just a visual aid for often complex evidential patterns of nested exemplars, just as is the case in ultrametric cluster analysis. Scientific empiricism rejects solutions involving unnamed or unobservable entities or hidden causes. The introduction of other, less certain data or theories (e.g., from morphometrics, fossils, cytology, biogeography, chemistry, development) as additional evidence for scientific induction of evolutionary process involving descent with modification of taxa would collapse the pattern-based statistical certainty of molecular cladograms. Thus, in cladistics, all data outside the data set that are relevant to macroevolutionary theory are “mapped” on the dendrogram or in some other way relegated to the fundamental structure of the cladogram. This is not science.
Important taxa in conservation and biodiversity research are being synonymized as a kind of epistemological extinction. Their identity is often intellectually scrambled by transfer into disparate and heterogenous groups, or to those that are molecularly diagnosed, or they are buried among molecular cryptic species, genera and families. It is understood that phylogenetics is presently fueled by comparatively high levels of funding, distinction of association with DNA studies, highly sophisticated mathematics supporting apparent statistical certainty of patterns in many cases, and ease of generating results as was the case with the “automatic taxonomy” of phenetics. But without names or sensible nomenclatural handles for macroevolutionarily significant taxa, stewards of biodiversity are stymied.
An immediate return is urged to the practice of evolutionary systematics and process-based science. This involves using all evidence, including both sister-group and ancestor-descendant analysis, the importance of both similarity and difference, and use of both induction and deduction to form theories of evolution of natural groups upon which a robust and responsible classification can be based. Pluralistic taxonomy is not easy and its results are not certain but it is scientific. It is a challenge that can be aided by modern methods of non-ultrametric cluster analysis but never in the context of structuralism.
This Manifesto is mainly targeted towards students, because established phylogeneticists used to “tree-thinking” have difficulty thinking outside that structuralist box, and also towards unaligned alpha taxonomists who should be aware of the non-empirical structuralist basis behind recent massive changes in classification. To be as short and clear as possible, documentation was eliminated from the above. Readers may find additional discussion, examples of taxonomy papers using modern evolutionary systematics, and references to the works of authors with similar or at least relevant observations on the Modern Evolutionary Systematics Web page: http://www.mobot.org/plantscience/resbot/EvSy/Intro.htm