Evolution Study Guide

A Neutral, Evidence-Centered Study Guide

Evolution and the debate surrounding it, examined claim by claim.

Evidence Matrix: High-Level Review

Every chapter from Foundations through Abiogenesis applies the same reasoning pattern in full: define the claim, present the strongest evidence for it, present the strongest challenge to it, separate what each side does and does not establish, name the open questions, and state what would falsify or seriously weaken it. This page compresses that pattern into one fast-review table covering the fourteen topics required by this guide's evidence matrix, so a reader who already worked through the chapters can review the whole guide at a glance, or a reader in a hurry can see the overall shape of the evidence before deciding which chapter to read in full.

How to Use This Matrix

This table is a compressed index, not a substitute for the chapters it summarizes. Each row distills a chapter (or a chapter section) that treats its topic in full, with citations, worked examples, and the complete evidence/challenge/limit pattern; the topic name in the first column links directly to that fuller treatment. Where a cell here seems too brief to be convincing on its own, that is intentional — the full case, including the sources behind it, lives on the linked page.

Each topic's Matrix column lists five fields, following the distinctions this guide draws throughout (see Limits of Inference): Evidence For and Strongest Challenge summarize the strongest material on each side; What Is Directly Observed and What Is Inferred separate direct measurement from historical or theoretical reconstruction for that topic specifically; and Open Question names what remains genuinely unresolved. Each topic name is also followed by this guide's own evidential status label for that claim, drawn from the corresponding chapter's own assessment rather than reconstructed here from scratch. (What would falsify or seriously weaken each claim is covered in the corresponding chapter itself, not repeated in this summary table.)

The first seven rows are the seven evolutionary claims from The Seven Evolutionary Claims, in their original order of increasing scope; confidence generally decreases moving down that block, from directly measured mechanisms toward deep historical inference. The remaining seven rows are cross-cutting topics — biological information, protein innovation, complex molecular systems, mutation-load arguments, waiting-time arguments, Intelligent Design, and abiogenesis — each drawn from its own dedicated chapter and not ordered by scope in the same way.

As throughout this guide, “not yet falsified” and “positively supported” are different claims in every row of this table, for every topic, on every side of every debate it covers.

Evidence Matrix

Overall evidence matrix across the fourteen required topics
Topic Matrix
Heritable genetic variation
Status: Directly observed.
Evidence ForPedigree sequencing directly observes new mutations and their inheritance Four-generation pedigree mutation study; a large trio study confirms rate and parental-age effects Icelandic de novo mutation study.
Strongest ChallengePhenotype is not simply genotype: identical genotypes show substantial phenotypic variation in controlled conditions Phenotypic variation in Drosophila.
What Is Directly ObservedNew DNA sequence variants arising in offspring and being transmitted to further generations.
What Is InferredVery little inference is needed for this narrow claim; it is among the most directly measured claims in the guide.
Open QuestionHow the many distinct types of new variants (point, structural, repeat-length) differ in their downstream functional effects.
Microevolution
Status: Directly observed.
Evidence ForLenski's long-term E. coli experiment tracked allele-frequency change across tens of thousands of generations Genome evolution in the long-term E. coli experiment, including the citrate innovation Genomic basis of citrate utilization in E. coli; a field release shifted stickleback allele frequencies within one generation Stickleback selection experiment.
Strongest ChallengeDiminishing-returns epistasis Diminishing-returns epistasis and background-dependent resistance ceilings Resistance constraints and genetic background show evolutionary pathways are constrained, not unlimited.
What Is Directly ObservedAllele-frequency change under selection, measured directly in laboratory and field populations.
What Is InferredWhether thousands of laboratory generations scale up to geological-timescale macroevolutionary change — addressed separately below.
Open QuestionHow far specific documented pathway constraints generalize beyond the organisms and traits already studied.
Natural selection and adaptation
Status: Directly observed.
Evidence ForGenotype-to-fitness links measured directly: stickleback loci shifting toward stream-associated alleles Stickleback selection experiment, and Darwin's finch beak size tracking HMGA2 variation under drought- and competition-driven selection Darwin's finches, HMGA2 and beak size.
Strongest ChallengeGenetic drift produces variation with no fitness difference Experimental neutral drift and phenotypic variation; pleiotropy and chance also shape which alleles spread Selection, pleiotropy and chance in bacterial evolution; the direction of selection itself can reverse over time Reversing selection in Darwin's finches.
What Is Directly ObservedGenotype-linked fitness differences producing a measured, corresponding change in allele frequency.
What Is InferredThe relative contribution of selection versus drift, pleiotropy, and chance in any one specific historical case.
Open QuestionHow to reliably apportion an observed frequency change between selection and drift when both plausibly contributed.
Speciation
Status: Directly observed in documented cases.
Evidence ForRepeated, independent allopolyploid speciation in Tragopogon within roughly a century Tragopogon polyploid speciation; the “Big Bird” finch lineage forming and becoming reproductively isolated on Daphne Major Big Bird finch lineage.
Strongest ChallengeSpecies boundaries are often porous: Heliconius lineages persist despite ~99% genome-wide homogenization by gene flow Heliconius divergence with gene flow; tropical eel species show extensive historical hybridization while remaining distinguishable over millions of years Historical hybridization among tropical eels.
What Is Directly ObservedSpecific, documented cases of new reproductively or ecologically distinct lineages forming.
What Is InferredWhether every historical species pair formed by a broadly comparable process to the observed cases.
Open QuestionHow continuous versus binary reproductive isolation really is across different taxa and timescales.
Macroevolution
Status: Strongly supported by converging historical evidence.
Evidence ForWhale evolution: convergent ear-anatomy, Indohyus morphology and isotopes, and independent retroposon-insertion evidence all place whales inside the artiodactyl radiation Cetacean hearing transition Indohyus and early whale relatives Retroposon evidence placing whales within artiodactyls; Tiktaalik's fish-tetrapod mosaic anatomy was a successfully predicted discovery Tiktaalik transitional anatomy.
Strongest ChallengeThe fossil record is incomplete, and even celebrated transitional sequences retain acknowledged gaps Early tetrapod/transitional fossil analysis; the Cambrian radiation shows genuinely elevated diversification rates Cambrian evolutionary-rate analysis.
What Is Directly ObservedNothing about deep-time transitions is directly observed; this claim rests entirely on historical reconstruction from fossils, anatomy, chronology, and genetics.
What Is InferredSpecific ancestor-descendant pathways, exact rates, and which fossils are direct ancestors versus close relatives.
Open QuestionHow much of the apparent gap between microevolutionary observation and macroevolutionary claims is closed by converging evidence versus legitimately remaining inference.
Common descent
Status: Very strongly supported.
Evidence ForHuman chromosome 2 carries the specific fusion signatures — internal telomere-like sequences and a degenerated second centromere — predicted by an ancestral fusion event Human chromosome 2 fusion site Degenerate ancestral centromere on human chromosome 2; primates share endogenous retroviral insertions at corresponding genomic positions following the expected branching pattern Primate endogenous retroviral insertion patterns.
Strongest ChallengeInsertion-site homoplasy is measurable, though rare (~0.52% in one primate study, with no independent hits at the specific sites surveyed) LINE insertion-site homoplasy in primates; roughly 30% of the gorilla genome locally conflicts with the dominant human-chimpanzee species tree due to incomplete lineage sorting Gorilla genome and incomplete lineage sorting.
What Is Directly ObservedThe genomic markers themselves — specific sequences, insertions, and structural rearrangements — are directly sequenced and documented.
What Is InferredThat these historically contingent markers derive from one shared ancestral event rather than independent, coincidental origin.
Open QuestionExact branching details in specific contested cases, and how much local gene-tree conflict is attributable to each of several known complicating processes.
Universal common descent
Status: Strongly supported, harder to reconstruct in detail.
Evidence ForDeep molecular conservation across Bacteria, Archaea, and Eukarya, including a comparative analysis of 23 conserved proteins across 45 organisms Conserved proteins across domains; an explicit statistical model comparison favored universal common ancestry overwhelmingly over separate-ancestry alternatives Theobald statistical test of universal common ancestry.
Strongest ChallengeThe statistical test itself was methodologically challenged Critique of Theobald's universal-common-ancestry test Theobald response to methodological critique Analysis of formal tests of universal common ancestry; horizontal gene transfer means a strict universal tree may describe only a small fraction of microbial genome histories Tree of one percent critique Lateral gene transfer in prokaryotic evolution.
What Is Directly ObservedShared molecular machinery — the genetic code, ribosomes, DNA as hereditary material — directly observed across all presently known cellular life.
What Is InferredThat this shared machinery traces to one common ancestral population (LUCA) rather than to independent origins that happened to converge on similar chemistry.
Open QuestionThe nature of LUCA (population versus single organism), the earliest lineage structure, and how much of the deepest genome was vertically versus horizontally inherited.
Biological information
Status: Capacity directly demonstrated; frequency in nature less certain.
Evidence ForNew sequence and inheritance are directly observed Four-generation pedigree mutation study; mutation and selection have improved weak promiscuous activities into strong ones Aharoni et al. promiscuous protein functions and changed hormone specificity via ancestral reconstruction Historical contingency in glucocorticoid receptor evolution; a random-sequence protein was shown to have a selectable biological effect Functional random-sequence protein in E. coli.
Strongest Challenge“Information” names at least four distinct, non-interchangeable senses (new sequence, more genetic material, new function, system complexity); demonstrating a mechanism can produce novelty is a different, narrower claim than demonstrating how often it does (capacity versus frequency).
What Is Directly ObservedNew DNA sequences, duplicated genetic material, and improved or altered protein function in specific, documented laboratory and comparative cases.
What Is InferredHow often these mechanisms, alone or combined, explain the origin of specific historical biological innovations in wild populations.
Open QuestionHow many currently accepted de novo gene candidates would survive a rigorous synteny check against densely sampled close relatives.
Protein innovation
Status: Genuinely two-sided; both accessibility and constraint are well documented.
Evidence ForEnzyme promiscuity lets weak side activities be strengthened by mutation and selection Aharoni et al. promiscuous protein functions; a random 80-residue library of ~6×1012 sequences yielded four unrelated ATP-binding protein families Keefe & Szostak random-sequence ATP-binding proteins.
Strongest ChallengeThe GFP fitness landscape shows most mutations are harmful and epistasis blocks many paths GFP local fitness landscape; Axe's estimate put retention of one demanding enzyme-domain function at roughly one in 1064 hydropathic-compatible sequences Douglas D. Axe, Estimating the Prevalence of Protein Sequences Adopting Functional Enzyme Folds — Journal of Molecular Biology.
What Is Directly ObservedFunctional change in laboratory-evolved proteins, and selectable ATP-binding sequences recovered directly from a random library.
What Is InferredHow rare or accessible functional sequences are in general, beyond the specific studied cases — no universal value has been established.
Open QuestionWhether most biologically important functions sit on well-connected neutral networks (like the studied cases) or resemble the far more isolated case Axe's estimate describes.
Complex molecular systems
Status: Local constraints demonstrated; no universal limit demonstrated on either side.
Evidence ForAncestral reconstruction and direct experimental testing showed fungal V-ATPase's modern three-subunit dependency arose from a single ancestral protein via duplication and complementary loss V-ATPase complexity via duplication and complementary loss.
Strongest ChallengeMany mutational pathways really are blocked by epistasis, with mutational order strongly affecting which routes remain open Epistatic accessibility of mutational pathways — the empirical grounding of the irreducible-complexity concern.
What Is Directly ObservedThe V-ATPase experimental reconstruction and the resulting increase in subunit interdependence, tested directly in living yeast.
What Is InferredWhether comparable co-option or duplication routes explain the origin of other systems (the flagellum, ATP synthase, and others discussed in Intelligent Design) that have not been reconstructed this directly.
Open QuestionNo general method currently exists for calculating in advance how many duplication or co-option steps a given complex system would require from a plausible simpler ancestor.
Mutation-load arguments
Status: Real, established mainstream force; the broader “genetic entropy” claim is disputed.
Evidence ForMutation load, mutation-selection balance, and Muller's ratchet are established mainstream population-genetic phenomena; mutational meltdown has been directly observed in engineered small-population, elevated-mutation-rate laboratory yeast Zeyl, Mutational meltdown in laboratory yeast populations — Evolution.
Strongest ChallengeYoung-earth “genetic entropy” extends this to a claim that essentially all populations inevitably deteriorate Sanford et al., Mendel's Accountant: A New Population Genetics Simulation Tool for Studying Mutation and Natural Selection — International Conference on Creationism proceedings / ICR; whether that holds depends on effective population size, recombination, and beneficial-mutation supply, which vary widely Brian Charlesworth, Effective population size and patterns of molecular evolution and variation — Nature Reviews Genetics.
What Is Directly ObservedFitness decline and extinction under engineered small effective population size and artificially elevated mutation rate in laboratory yeast.
What Is InferredWhether mutation load causes inevitable long-term decline in large, naturally recombining, real-world populations at natural mutation rates — a considerably broader claim than the engineered result.
Open QuestionThe true distribution of fitness effects for near-neutral mutations in real genomes, and how effective population size and recombination actually combine in long-lived natural lineages.
Waiting-time arguments
Status: Genuinely parameter-dependent; no single universal answer.
Evidence ForBehe and Snoke's peer-reviewed model found that under restrictive assumptions (no partial intermediate activity, one predetermined target), fixing two or more specified residues could require populations of 109 or more within 108 generations Behe & Snoke, Simulating evolution by gene duplication of protein features that require multiple amino acid residues — Protein Science.
Strongest ChallengeMichael Lynch showed the conclusions depend heavily on those restrictive assumptions, and waiting times fall substantially once partial activity or alternative solutions are allowed Michael Lynch, Simple evolutionary pathways to complex proteins — Protein Science; Durrett and Schmidt showed two-mutation waiting times vary enormously with effective population size Durrett & Schmidt, Waiting for two mutations: with applications to regulatory sequence evolution and the limits of Darwinian evolution — Genetics.
What Is Directly ObservedNothing directly; these are mathematical population-genetic models, not direct field or laboratory observation of a historical waiting time.
What Is InferredWhether any specific historical innovation actually fit the restrictive Behe-Snoke parameter regime or the more permissive assumptions Lynch modeled.
Open QuestionFor any given historical innovation, what the actual functional-target size was and whether selectable intermediates were genuinely available.
Intelligent Design
Status: Not currently positively established; specific narrow claims are falsifiable in principle.
Evidence ForID's own positive arguments rest on specified complexity Specified complexity and testability and irreducible complexity Behe & Meyer, Irreducible complexity, bacterial flagellum and the Type III Secretory System — Discovery Institute; comparative case studies show real, unresolved gaps in several systems' earliest history (the ribosome, ATP synthase, and cilium among them).
Strongest ChallengeHomology, gene duplication, and co-option evidence substantially weakens strong “simultaneous origin” claims for several systems Pallen & Matzke, From The Origin of Species to the origin of bacterial flagella — Nature Reviews Microbiology V-ATPase complexity via duplication and complementary loss; mainstream institutions argue ID does not yet meet the bar of a testable causal account National Academies evolution resources.
What Is Directly ObservedThat humans, as intelligent agents, can and do engineer genomes today — establishing that intelligent causation is not inherently impossible to investigate, without establishing that it occurred in any specific ancient case.
What Is InferredWhether any specific biological system's origin actually required intelligent intervention, as opposed to being presently unresolved under current evolutionary models.
Open QuestionWhether any presently proposed design signature can be independently validated on known engineered systems and then applied prospectively, rather than defined only after an evolutionary account proves difficult (the argument-from-ignorance concern Scientific criticism of argument-from-ignorance reasoning in ID).
Abiogenesis
Status: Substantially unresolved; not comparable in evidential depth to the evolutionary chapters.
Evidence ForSome individual chemical steps have genuine experimental support: spontaneous formation of certain organic molecules under early-Earth-like laboratory conditions, self-assembly of membrane-like vesicles, and catalytic activity in RNA.
Strongest ChallengeNo complete, experimentally demonstrated pathway exists from nonliving chemistry to an autonomous first living cell; reliable self-replication, the genetic code's origin, homochirality, and the coupling of heredity to metabolism remain substantially unresolved.
What Is Directly ObservedThe individual demonstrated chemical steps themselves — organic-molecule formation, vesicle self-assembly, and RNA catalysis — each tested directly in the laboratory.
What Is InferredThat any of the named model families (RNA-first, metabolism-first, protocell, hydrothermal) describes what actually happened; these remain named, actively studied hypotheses rather than an established consensus account.
Open QuestionThe origin of reliable self-replication and of the genetic code; homochirality; and how heredity and metabolism became coupled into one integrated system.

Reading this table as a whole: the first four rows rest almost entirely on direct observation; the next three shift progressively toward historical and statistical inference while remaining strongly supported; and the final seven cross-cutting rows show a genuinely mixed picture — real capacity and real constraint documented side by side, with Intelligent Design and abiogenesis the two topics where this guide's own chapters are most explicit that a complete, positively established account does not yet exist on either side of the debate.

Sources for This Chapter