Summary
To be written. What mutation, selection, drift, recombination, and shared patterns can establish about biological change. Nine articles in this section are written; the rest are planned. This page will eventually state where they came out, which arguments survived, and which were retired.
This section was narrowed on 2026-09-05. It had been the parent of most of the biology catalogue, including material that is not a mechanism by any reading. Five groups are now siblings of this page rather than children of it: Observed Evolution, New Genes, Major Evolutionary Transitions, Limits of Evolutionary Change, and Biological Design. What remains here is inference from shared patterns, the geography of life, and the population genetics everything else rests on.
The articles
Linked titles are written. Bold titles are planned.
- Phylogenetic Inference and Congruence — Biological trees agree above chance, but designed systems can too. Common descent gains force from genomic insertions, chromosome order, and fossils, not congruence alone.
- Shared Genetic Mistakes and Common Ancestry — Shared broken genes and mobile-element insertions are the strongest genetic evidence for common ancestry. Homoplasy in these markers is real and documented, but measured at 0.05% to 3%, far too rare to explain the pattern. What the failures do break is the claim that the markers are perfect and need no statistics.
- Biogeography and Common Descent — Island species, Australian marsupials, geographic ancestry patterns, and proposed post-Flood dispersal.
- Evolutionary Biogeography: What the Map of Life Shows — Darwin predicted in 1859 that the depth of the sea between an island and the mainland would predict its mammals. Measured here on modern bathymetry, it does, across eighteen crossings. The vicariance half of the modern theory has fared much worse, and the field says so itself.
- Creationist Biogeography: The Post-Flood Dispersal Model — Rafting, land bridges and rapid speciation are all real, and the model built from them still fails. Measured here: the shallowest walkable route from Asia to Australia needs a 1,264 m sea-level fall, and the Asian fossil trail marsupials should have left is empty.
- Population Genetics — Effective size is not headcount; most mutations are lost by drift; selection weakens below 1/(2Ne); and load is not a death count. Clock and bottleneck estimates depend on demographic models.
- Genetic Entropy — Load does not imply extinction, but weak harmful mutations can fix by drift and Muller's ratchet is real. Long-term fitness loss remains unmeasured.
- Haldane's Dilemma and the Waiting-Time Problem — Haldane's reproductive limit is real, but human adaptation did not use the classic sweeps it constrains. Waiting-time claims assume prespecified sequence targets, a premise not established.
- Population Bottlenecks — A brief bottleneck keeps most heterozygosity but destroys allelic richness. Two founders cap a locus at four alleles. Real founder populations recover in numbers and stay damaged in genes. The size of the biblical bottleneck is defensible; its date is not.
- Development and Constraint — Gene regulatory networks, Hox genes, developmental constraint, epigenetics, cooperation, and the extended evolutionary synthesis. Convergence is covered under Biological Design.
- Evolutionary Computation: What Genetic Algorithms Show About Evolution — Genetic algorithms work, and every famous one works because a designer drew the search space, wrote the score, and smoothed the landscape until selection could climb it. Avida, ev, Weasel and Tierra run at mutation rates, selection strengths and fitness effects with no biological parallel; the one honest simulation, Tierra, plateaued. The critics of these programs overreach too, and the whole dispute reduces to one unmeasured number: how smooth the real fitness landscape is.
Verdict
To be written.