The Origin of Speciess04

Chapter II. Variation Under Nature

Charles Darwin · 1872. Read the original ↗

Thesis

Natural populations vary, the boundary between varieties and species is arbitrary and graded, and patterns of variation in large and dominant genera are what one would expect if varieties are incipient species.

The observed claims about variation and the blurred species/variety boundary are well documented and independently supported, and the large-genera prediction was a genuine test. The step from varieties to species by natural selection is an extrapolation whose demonstration is deferred to later chapters. On the language of agency, the chapter mostly uses 'selection' and 'manufactory of species' as shorthand for accumulating observed variation, and nothing in this chapter's own argument depends on foresight or choice.

BStrongly supported

Claims, graded

  1. C1ObservedB

    Wild species show heritable individual differences, including in important organs, and these differences are the raw material on which selection can act.

    “These individual differences are of the highest importance for us, for they are often inherited, as must be familiar to every one”

    Evidence offered Cited cases of variable nerve branching in Coccus and muscles in insect larvae (Lubbock), polymorphic genera, and Darwin's own years of collecting; the heritability is asserted as familiar from domestic breeding.

    1. 1Observation///Variation in wild populations, including in internal organs, is directly documented with named sources.
    2. 2Hypothesis//The claim is specific enough to be wrong, though degree of heritability in wild forms is not quantified.
    3. 3Prediction/It is not framed as a risky prediction; it is a background observation.
    4. 4Experiment/Heritability is inferred from familiarity and domestic breeding, not from controlled tests on the wild cases cited.
    5. 5Replication//Multiple observers (Lubbock, Wallace, De Candolle) report variation independently.
    6. 6Falsifiability//Darwin addresses the objection that important organs never vary and exposes it as circular, but does not rule out non-heritable causes for specific cases.

    What would move it up Breeding or pedigree data showing that the specific wild variants cited (e.g. nerve branching in Coccus) are transmitted to offspring.

  2. C2ObservedB

    The distinction between species, sub-species, varieties and individual differences is graded and arbitrary, with no sharp line marking where one ends and the next begins.

    “These differences blend into each other by an insensible series; and a series impresses the mind with the idea of an actual passage.”

    Evidence offered Disagreements among botanists over British plants (e.g. 251 vs 112 species), Wallace's Malayan butterflies, De Candolle's work on oaks, and Walsh's Phytophagic forms.

    1. 1Observation///The taxonomic disagreements and intergrading forms are documented with numbers and named authorities.
    2. 2Hypothesis//The claim is clear, though 'arbitrary' is a statement about classification rather than a testable biological mechanism.
    3. 3Prediction/De Candolle's remark that better-known groups yield more doubtful forms fits, but is not set up as a prediction.
    4. 4Experiment/This is a survey of classification practice, not a controlled test.
    5. 5Replication///Many independent taxonomists across plants and animals find the same difficulty.
    6. 6Falsifiability//A clean gap between species and varieties would count against it, and the rival view (fixed species) is discussed.

    What would move it up Quantitative genetic comparison of reproductive isolation and divergence across graded forms to show whether discontinuities are real or only a matter of convention.

  3. C3ObservedB

    Species of larger genera have more varieties, and widespread, common species yield more varieties, as predicted if varieties are incipient species.

    “it has invariably proved to be the case that a larger proportion of the species on the side of the larger genera presented varieties, than on the side of the smaller genera.”

    Evidence offered Darwin's tabulation of plants from twelve countries and beetles from two districts, divided into large and small genera, with Hooker's endorsement; the tables themselves are withheld for a future work.

    1. 1Observation//The tabulated counts are reported but not shown, and rest on contested decisions about what counts as a variety.
    2. 2Hypothesis///The prediction is stated in advance and could clearly fail.
    3. 3Prediction///Darwin says he anticipated the pattern from his theory and then tested it against data from many countries.
    4. 4Experiment//A systematic comparison across two groups, though not controlled, and subject to the variety/species ambiguity.
    5. 5Replication/Several countries are included, but all the analysis is Darwin's own with Hooker's review.
    6. 6Falsifiability//He says the opposite pattern would be a problem and contrasts the special-creation view, though that rival could be adjusted to fit and he does not consider alternatives such as shared habitat or taxonomic bias.

    What would move it up Publication of the full tables, plus replication with independently classified floras and faunas that control for the taxonomists' tendency to name more varieties in well-studied or large groups.

  4. C4ExtrapolatedD

    Well-marked varieties are incipient species, and the passage from slight variety to species is the product of the cumulative action of natural selection on adaptive characters.

    “the passage from one stage of difference to another may be safely attributed to the cumulative action of natural selection, hereafter to be explained”

    Evidence offered The graded series of forms between individual differences and species, the similarity of species in large genera to varieties, and the analogy with human accumulation of differences in domestic productions; Darwin explicitly defers the demonstration to later chapters.

    1. 1Observation/Graded forms are observed, but the transformation of varieties into species is not.
    2. 2Hypothesis//The claim is clear but 'incipient species' is defined so loosely that failure is hard to specify.
    3. 3Prediction//The large-genera patterns were predicted, but other views could accommodate them.
    4. 4Experiment—No test of varieties becoming species by selection appears in this chapter.
    5. 5Replication—Nothing offered here that is replicated.
    6. 6Falsifiability/Darwin sets his view against special creation but considers few other alternatives, and says that varieties may also just persist or go extinct.

    What would move it up Observed cases of populations diverging under measured selection to the point of reproductive isolation, with the genetic changes identified and traced.

  5. C5InferredC

    The patterns of variation in large genera are readily explained if species arose as varieties, and are 'utterly inexplicable' if species are independent creations.

    “we can clearly understand these analogies, if species once existed as varieties, and thus originated; whereas, these analogies are utterly inexplicable if species are independent creations.”

    Evidence offered The large-genus statistics, clustering of species like varieties, and similar restricted ranges of varieties and closely allied species; the contrast is drawn only against special creation.

    1. 1Observation//The patterns are observed, though they depend on contested classification.
    2. 2Hypothesis//Descent from varieties is specific, but the creationist rival is stated so that it makes no prediction at all.
    3. 3Prediction//The large-genus pattern was anticipated from the theory.
    4. 4Experiment—No test distinguishes the rivals directly.
    5. 5Replication/The same patterns are reported across several countries and groups, but by one analyst.
    6. 6Falsifiability/Only one rival is considered, and it is stated in a form that predicts nothing.

    What would move it up Showing that a specified rival (for example, independent creation with a stated pattern of design) makes different, testable predictions about genus size and variety counts, and that the data favour the descent view.

Reasoning check

  • Possible
    Claim C5
    “these analogies are utterly inexplicable if species are independent creations.”

    Darwin offers real evidence (the large-genus statistics) and a real test of his own prediction, but sets it against only one rival, special creation. Other explanations, such as shared ecological conditions or taxonomic bias in well-studied groups, are not examined; checking them against the data would settle the point.

  • Possible
    Claim C4
    “may be safely attributed to the cumulative action of natural selection, hereafter to be explained”

    In this chapter the source of variation is observed (heritable individual differences), and selection is credited with accumulating, not creating, it, which largely answers the charge. But the claim that accumulation carries a variety to species level rests on a promise of later chapters; measured selection producing reproductive isolation in a living lineage would settle it.

  • Possible
    Claim C1
    “in the same manner as man accumulates in any given direction individual differences in his domesticated productions.”

    The breeder comparison is used here only for the shared feature that heritable differences can accumulate, which both share. Breeders also supply foresight, a goal and isolation of the line, which nature lacks, and the chapter does not show that nature's accumulation is directed or sustained; later chapters on selection would have to supply this.

What it does well

  • Clearly separates observed variation from inference about what it means, and openly admits the difficulties, such as the lack of a clear line between species and varieties.
  • Makes a prediction in advance (more varieties in large genera) and tests it against data from many regions, which fits the rubric's call for risky prediction.
  • Exposes the circularity of ranking organs as 'important' because they do not vary, and notes where his own pattern could have failed ('it would have been fatal to my theory').
  • Cites many independent observers (Wallace, Lubbock, De Candolle, Watson, Hooker) and gives specific numbers for the British plants.
  • Keeps the agent language of selection mostly in the future tense ('hereafter to be explained'), so the chapter's own observations do not lean on it.

Limits of this evaluation This is one of 17 sections, so the main arguments for natural selection, the struggle for existence and divergence are not evaluated here. Darwin withholds his tables and the catalogue of facts for a future work, so the data cannot be checked in this text. The agency check (reification, tautology, false analogy, filter-as-source) finds little to rate in this chapter because selection is mostly introduced by promissory reference; phrases such as 'manufactory of species' read as figures of speech that the argument does not need. The tautology concern, 'fittest' defined by survival, does not arise since 'dominant' is defined here by commonness and range, which is measured independently. Later findings (genetics, speciation studies) bear on the claims but are not credited here.