The Origin of Speciess17

Chapter XV. Recapitulation And Conclusion

Charles Darwin · 1872. Read the original ↗

Thesis

Darwin summarises his case that species descend with modification from common ancestors, chiefly through natural selection of small heritable variations. He argues that the objections stem from our ignorance, and that the theory explains far more facts than special creation does.

The chapter is built on real observations: variation, over-production and the success of artificial selection. Its central step, that this scales to the origin of complex organs and of all major groups, is inferred or extrapolated, and its answers to the objections rest largely on our ignorance of the record and of the limits of variation. The agency language of selection is mostly shorthand for a mechanism that Darwin spells out, but the case that selection builds, rather than keeps, new structures is not made from observation here.

CInference

Claims, graded

  1. C1ObservedB

    Heritable individual variation, a high rate of increase that causes a struggle for existence, and differential survival of slightly advantaged individuals together produce a real, ever-acting selective process under nature, as breeders' selection does under domestication.

    “In the survival of favoured individuals and races, during the constantly recurrent Struggle for Existence, we see a powerful and ever-acting form of Selection.”

    Evidence offered Observed variation under domestication and nature, a geometrical ratio of increase supported by calculation and by the spread of naturalised species, and the effect of breeders' selection of slight differences.

    1. 1Observation///Variation, heritability, over-production and differential survival are all directly observed, and artificial selection is directly observed.
    2. 2Hypothesis//The claim is clear, but 'slightest advantage' and 'a grain in the balance' are not tied to any measured survival difference.
    3. 3Prediction/No risky advance prediction is offered here, although the claim implies that fitness differences should be measurable.
    4. 4Experiment//Breeding is controlled selection, but the chapter reports no controlled test of selection in nature.
    5. 5Replication//Domestic breeding is widely replicated, while natural selection is supported mainly by general observation.
    6. 6Falsifiability//A population with no heritable fitness variation would count against it, but the text does not set out such a test.

    What would move it up Measure survival and reproduction in wild populations against independently measured traits, and show that the trait differences are inherited and that the population changes in the predicted direction.

  2. C2ExtrapolatedD

    Natural selection acting on small variations is the main cause of species' modification, adaptation and the origin of complex organs and instincts, with no limit to what slow accumulation can achieve.

    “What limit can be put to this power, acting during long ages and rigidly scrutinising the whole constitution, structure, and habits of each creature, favouring the good and rejecting the bad? I can see no limit to this power”

    Evidence offered The success of breeders over short periods, the existence of gradations in nature, and the claim that no one has proved that variation is limited.

    1. 1Observation/Small-scale change under selection is observed, but the building of complex organs is not.
    2. 2Hypothesis//The claim of stepwise accumulation is definite enough to fail, for example if a complex organ had no functional intermediates.
    3. 3Prediction/The chapter says the facts 'might even have been anticipated', but the examples are explained after the fact.
    4. 4Experiment/Domestic breeding is controlled, but it shows only modest change, and Darwin concedes that the steps in great changes are not seen.
    5. 5Replication/Breeding results are replicated; the origin of complex adaptations is not independently reproduced.
    6. 6Falsifiability/The burden is placed on critics to show a limit, and the objections are answered by appeal to ignorance of the record.

    What would move it up Show, in a living or fossil lineage, the stepwise origin of a complex structure with each step's function and heritable basis. Or show experimentally that selection on continuing variation passes the limits that the breeders' record suggests.

  3. C3InferredC

    Homology, embryology, rudimentary organs, classification into nested groups, biogeography and the geological succession are all better explained by descent with modification than by independent creation.

    “The similar framework of bones in the hand of a man, wing of a bat, fin of the porpoise, and leg of the horse ... at once explain themselves on the theory of descent with slow and slight successive modifications.”

    Evidence offered A list of patterns in nature (nested groups, island endemics related to the nearest mainland, homologous limbs, similar early embryos, rudimentary organs) each said to be 'inexplicable' on special creation.

    1. 1Observation///The patterns of homology, distribution, embryology and rudiments are real and well documented.
    2. 2Hypothesis//Common descent is clear, but the chapter treats the rival as one unspecified 'theory of creation'.
    3. 3Prediction//Some patterns are presented as expected, such as island biota resembling the nearest mainland, but they are largely explained after the fact.
    4. 4Experiment—Descent from a common ancestor in the deep past cannot be tested by a controlled experiment.
    5. 5Replication//The patterns are repeated across many groups and regions, but the chapter cites no independent test of the inference.
    6. 6Falsifiability//A comparison with creation is made, but the rival is stated in a limited form, and the chapter does not say which pattern would refute descent.

    What would move it up Test advance predictions of descent, such as a fossil or molecular pattern that would contradict a single nested tree. Compare the result against specific rival hypotheses, including independent design with reuse of a common plan.

  4. C4InferredC

    The gaps in the fossil record and the absence of pre-Cambrian fossils are explained by extreme imperfection of the geological record, and the fossils we have show slow, graduated change across consecutive formations.

    “I can answer these questions and objections only on the supposition that the geological record is far more imperfect than most geologists believe.”

    Evidence offered Arguments about sedimentation, the small area explored, intermittent deposition and the rare preservation of fossils, plus the observation that consecutive formations hold more closely related fossils.

    1. 1Observation//Incomplete preservation and the closer relation of adjacent formations are observed.
    2. 2Hypothesis/The degree of imperfection needed is not stated, so the claim is hard to fail.
    3. 3Prediction/The chapter predicts that links will not be found often, and no test of that is given.
    4. 4Experiment—No test of what a sample of the record of a given completeness would show is offered.
    5. 5Replication/Geological examples are cited, but the claim rests on one argument.
    6. 6Falsifiability/Darwin admits 'it is extremely difficult' to say, and the absence of links is itself used as supporting evidence.

    What would move it up Find intermediates in strata with good preservation, such as pre-Cambrian ancestors or finely graded lineages, and calibrate how complete the record should be for a given group.

  5. C5ExtrapolatedD

    All animals descend from four or five progenitors, and plausibly all life from one primordial form, by the same process.

    “I believe that animals are descended from at most only four or five progenitors, and plants from an equal or lesser number.”

    Evidence offered Shared chemistry, cellular structure, reproduction and germinal vesicle, intermediate forms between the kingdoms, and the same patterns of homology and embryology; Darwin says the inference to one prototype is 'chiefly grounded on analogy'.

    1. 1Observation/Shared features of living things are observed, but ancestry is not.
    2. 2Hypothesis//The claim of one or a few progenitors is definite enough to fail.
    3. 3Prediction—No prediction is stated; the claim is inferred from patterns already known.
    4. 4Experiment—No test is possible at this scale of time within the chapter.
    5. 5Replication/Shared features recur across groups, but no independent tests of ancestry are offered.
    6. 6Falsifiability/Darwin calls the step 'immaterial' and says analogy 'may be a deceitful guide', so no result is set out that would count against it.

    What would move it up Test independently whether separate lineages share one ancestral set of genes and molecular machinery, and compare that against rival explanations such as separate origins with common design constraints.

  6. C6AssumedE

    Because selection works only for each being's good, it will tend toward perfection, and the observed imperfections and cruelties in nature are what selection should produce.

    “And as natural selection works solely by and for the good of each being, all corporeal and mental endowments will tend to progress towards perfection.”

    Evidence offered The claim follows from the definition of selection as preserving useful variations, and examples such as the bee's sting and parasitic wasps are said to be unsurprising.

    1. 1Observation—No measure of progress toward perfection is observed.
    2. 2Hypothesis/'Perfection' is vague, and the claim is also made to cover imperfection, which weakens it.
    3. 3Prediction—It accommodates both perfect and imperfect structures and so predicts neither.
    4. 4Experiment—No test is described.
    5. 5Replication—No result is offered to replicate.
    6. 6Falsifiability/Darwin notes that 'the wonder ... is that more cases of the want of absolute perfection have not been detected', which is accommodation rather than a test.

    What would move it up Define 'improvement' by a measure independent of survival, then test whether lineages under selection increase on it, and which kinds of imperfection selection should permit.

Reasoning check

  • Possible
    Claim C2
    “What limit can be put to this power, acting during long ages and rigidly scrutinising the whole constitution, structure, and habits of each creature, favouring the good and rejecting the bad?”

    Selection is described as 'scrutinising', 'favouring' and 'rejecting', as if it were an agent. Darwin does spell out the mechanism (differential survival of heritable variants) elsewhere in the chapter, so the argument survives with the agent words removed, but the 'power' that has 'no limit' is then a statement about variation and survival rates. Measured variation and survival data would settle whether the limit question is empirical.

  • Possible
    Claim C2
    “If man can by patience select variations useful to him, why, under changing and complex conditions of life, should not variations useful to nature's living products often arise, and be preserved or selected?”

    Breeders have foresight, a goal, control of mating, and isolation of a line, and nature has none of these. Darwin partly closes the gap by stressing unconscious selection and the 'great agency' it has had, and by showing the process through ratio of increase and struggle. The comparison would be settled by measuring whether unguided differential survival sustains cumulative change in wild lineages.

  • Clear case
    Claim C2
    “It has often been asserted, but the assertion is incapable of proof, that the amount of variation under nature is a strictly limited quantity.”

    The conclusion that selection has no limit rests on the fact that no one has shown a limit. The text offers no direct test of unlimited accumulation. Darwin repeats the point in 'I can see no limit to this power', and it is rated the same way wherever it appears.

  • Possible
    Claim C2
    “As natural selection acts solely by accumulating slight, successive, favourable variations, it can produce no great or sudden modifications”

    Darwin says variation is 'caused, or at least excited' by conditions, and that its causes are largely unknown, so the source of new variants is acknowledged rather than shown. The chapter credits selection with 'production' of forms, though its text also treats selection as preserving. A demonstration that the variation needed for each step actually arises, in the right amount, would settle it.

  • Possible
    Claim C6
    “If, then, animals and plants do vary, let it be ever so slightly or slowly, why should not variations or individual differences, which are in any way beneficial, be preserved and accumulated through natural selection, or the survival of the fittest?”

    'Beneficial' and 'fittest' are tied to survival, and the text offers no independent trait measure here. Elsewhere Darwin does name traits such as vigour, weapons and better adaptation to physical conditions, which gives partial independent content. Measuring those traits ahead of time and testing whether they predict survival would settle it.

  • Possible
    Claim C3
    “It can hardly be supposed that a false theory would explain, in so satisfactory a manner as does the theory of natural selection, the several large classes of facts above specified.”

    Darwin compares his explanation with creation across many facts, and cites the wave theory of light as a precedent for this method. But the rival is stated in a loose form and many facts are explained after they were known. Advance predictions that a rival would not make, or a fair statement of the rival's predictions, would settle it.

  • Possible
    Claim C3
    “Nevertheless, they do not pretend that they can define, or even conjecture, which are the created forms of life, and which are those produced by secondary laws.”

    The chapter frames the choice as descent with modification or independent creation of each species, with sudden unexplained change as a brief third option. Darwin does consider that third option and rejects it for stated reasons, so the dilemma is only partly closed. Considering rival models of ancestry and of common design would settle it.

  • Clear case
    Claim C4
    “But it deserves especial notice that the more important objections relate to questions on which we are confessedly ignorant; nor do we know how ignorant we are.”

    The chapter answers missing intermediates and Cambrian ancestors by saying the record is unknown, and it treats the unknown as grounds for confidence. No evidence of the degree of imperfection required is given. Darwin also accepts that Thomson's time limit is a grave objection and answers only that we do not know the rate of change.

What it does well

  • States the objections to his theory honestly and at length, including the missing intermediates, the Cambrian problem and Thomson's age of the earth.
  • Separates secure observations (variation, over-production, artificial selection) from the more speculative claims, and says the one-prototype inference is 'chiefly grounded on analogy'.
  • Corrects a misreading by quoting his earlier statement that selection is 'the main but not the exclusive means', and concedes a larger role for other causes of variation.
  • Shows how one explanation covers many independent classes of fact, such as island biota, homology and nested classification, and offers this as a method with precedent.
  • Gives a quantitative basis for the struggle for existence in the geometrical ratio of increase.

Limits of this evaluation This is one chapter, a recapitulation, and much of the evidence (the ninth-chapter sterility data, the pigeon breeding, the geological arguments) is deferred to earlier chapters. The grades reflect what is argued in this section and the knowledge of 1872. Later work on genetics, the age of the earth, and molecular and fossil evidence bears on several claims, and mostly strengthens the case for common descent and for the observed operation of selection. On the focus: the agent words are usually backed by a spelled-out mechanism, the breeder analogy is partly qualified by unconscious selection, and the 'fittest' wording has some independent content, so these are rated 'possible'. The 'no limit' step is rated 'clear'.