The Origin of Speciess07

Chapter V. Laws Of Variation

Charles Darwin · 1872. Read the original ↗

Thesis

Variation arises from causes that are largely unknown but are related to the organism's nature, the conditions of life, use and disuse, and correlated growth. Common descent combined with natural and sexual selection explains patterns in how variable different kinds of characters are, and the accumulation of beneficial variations is said to account for the important modifications of structure.

The chapter rests on real observations, especially equid stripes, island beetles and breed variation, and it handles the language of agency better than most popular accounts: the agent wording is shorthand for survival differences, and the source of variation is admitted to be unknown. Most explanations are inferences without direct tests, much of the supporting data is withheld, and the closing claim that accumulated beneficial differences produced all important modifications goes beyond the cases shown, which are mostly reductions of existing parts.

CInference

Claims, graded

  1. C1ExtrapolatedD

    The steady accumulation of beneficial variations by natural selection has produced all the more important modifications of structure adapted to each species' habits.

    “it is the steady accumulation of beneficial differences which has given rise to all the more important modifications of structure in relation to the habits of each species.”

    Evidence offered The chapter points to the large differences among domestic pigeon breeds produced by fanciers' selection, and to natural cases of lost or reduced structures (wings, eyes, tarsi, carapaces). It offers no case of a new complex structure built up in this way.

    1. 1Observation/Selective accumulation is observed in breeds and loss of parts is seen in nature, but the building of important new structures is not observed.
    2. 2Hypothesis/The claim is so general ('all the more important modifications') that few specific results could count against it.
    3. 3Prediction/The chapter makes no advance prediction that the accumulation of beneficial variants would produce a particular complex structure.
    4. 4Experiment/Breeding is a controlled form of selection, but its direction and persistence are set by the breeder.
    5. 5Replication/Breeding results are widely repeated, but not for the construction of novel organs in unmanaged populations.
    6. 6Falsifiability/Alternatives are mainly set aside by Darwin's statement that he can see no other explanation, and the chapter admits profound ignorance of how variation arises.

    What would move it up Observe or experimentally reconstruct the stepwise origin of a new multi-part structure, with measured fitness differences at each step, or show that a continuous functional path of heritable variants exists and is selected.

  2. C2InferredC

    Patterns in how variable characters are (specific more than generic, unusually developed parts, secondary sexual characters, rudimentary parts) follow from common descent together with natural and sexual selection not yet having fixed recent changes.

    “All being mainly due to the species of the same group being the descendants of a common progenitor, from whom they have inherited much in common”

    Evidence offered Darwin asserts the pattern and gives a few cases (Pyrgoma barnacles, birds, beetle tarsi, wing neuration, Pontella). He states that the 'long array of facts' cannot be introduced here, and that the rule does not obviously hold in plants.

    1. 1Observation/The variability patterns are asserted or illustrated by a few cases, with the main data withheld and the plant case unresolved.
    2. 2Hypothesis//The explanation does imply expected patterns, such as the bat wing being no more variable than other parts.
    3. 3Prediction//Some consequences are drawn from the theory, such as constancy for long-inherited parts, but the pattern is largely known before it is explained.
    4. 4Experiment—No controlled test of the explanation is offered.
    5. 5Replication/Other naturalists are cited as having made similar remarks, but there is no independent test of the explanation.
    6. 6Falsifiability/Both high variability (selection not yet fixed, or absent) and low variability (selection long applied) are accommodated, and the rival is dismissed rather than tested.

    What would move it up Measure variability across traits in many taxa with an explicit measure of recent divergence, including plants, and test whether recently modified traits are more variable as predicted.

  3. C3InferredC

    The flightlessness of many Madeira beetles is mainly due to natural selection against flight, since flying individuals were blown out to sea, combined probably with disuse.

    “These several considerations make me believe that the wingless condition of so many Madeira beetles is mainly due to the action of natural selection, combined probably with disuse.”

    Evidence offered Counts of wingless species and genera on Madeira, beetles blown to sea, the greater share of wingless forms on the exposed Desertas, the absence of wing-dependent groups, and enlarged wings in flower-feeding insects.

    1. 1Observation//Wingless proportions, wind exposure and feeding habits were observed, but differential survival was not.
    2. 2Hypothesis//The mechanism is specific: flying individuals are lost to wind, flightless ones survive, and wing-dependent insects keep wings.
    3. 3Prediction//Larger wings in flower-feeding insects and more wingless forms on the exposed island fit the hypothesis, though Wollaston's observations came first.
    4. 4Experiment—No survival or heritability test is reported.
    5. 5Replication/Island flightlessness is reported elsewhere, but the mechanism is not independently tested here.
    6. 6Falsifiability//Darwin separates selection from disuse and notes cases where the expectation could fail, but does not rule out other causes.

    What would move it up Measure heritable wing variation and survival of winged versus wingless forms in windy and sheltered Madeira populations, and compare to the predicted loss rates. Later genetics also counts against the inheritance of use and disuse effects that Darwin invokes alongside selection.

  4. C4InferredC

    Natural selection tends to reduce any structure that has become superfluous, because saving the nutriment it would use gives an advantage.

    “natural selection is continually trying to economise in every part of the organisation.”

    Evidence offered The loss of the carapace in cirripedes living parasitically inside other cirripedes, argued to be an advantage because less nutriment is wasted.

    1. 1Observation//The reduced carapace in protected parasitic cirripedes was observed by Darwin.
    2. 2Hypothesis/The nutritional benefit is stated as an advantage but its size is not specified.
    3. 3Prediction/Reduction of useless parts is expected, but no risky prediction is made.
    4. 4Experiment—No test of fitness benefit from the saved nutriment is reported.
    5. 5Replication/Darwin says other instances could be given but supplies none.
    6. 6Falsifiability/Loss of the part is equally explained by disuse or by relaxed selection, and these are not separated.

    What would move it up Measure the energetic cost of the structure and show that reduced forms leave more offspring than less-reduced forms.

  5. C5InferredC

    Variation arises from causes that are largely unknown and is related to the nature of the organism and the conditions of life. The direct action of conditions is real but modest.

    “Our ignorance of the laws of variation is profound. Not in one case out of a hundred can we pretend to assign any reason why this or that part has varied.”

    Evidence offered Variation is more frequent under domestication and in wide-ranging species. Darwin cites scattered reports of colour and leaf effects from climate, and notes contrary cases and the difficulty of separating climate from selection.

    1. 1Observation//The patterns in frequency of variation are observed, though many effects rest on second-hand reports.
    2. 2Hypothesis/The claim is deliberately loose, naming 'a tendency to vary' from unknown causes.
    3. 3Prediction—The chapter makes no prediction of which variations will arise.
    4. 4Experiment/Comparisons across climates are suggestive, but the kidney-bean test Darwin calls for was never done.
    5. 5Replication/Several observers report similar effects, but they are mixed with counterexamples.
    6. 6Falsifiability//Darwin states counterexamples and the confound between selection and direct action, which keeps the claim open to test.

    What would move it up Raise matched stocks under controlled conditions over many generations and measure whether the rate and direction of heritable variation changes, with selection excluded.

  6. C6ObservedB

    Stripes appear on the legs and shoulders of horses, asses and hybrids of the horse genus, and blue-barred plumage reappears in crossed pigeons, as a recurring tendency toward characters of an earlier form.

    “We see several distinct species of the horse genus becoming, by simple variation, striped on the legs like a zebra, or striped on the shoulders like an ass.”

    Evidence offered Darwin's own observations and reports from named correspondents (Poole, Gray, Gosse, Martin) of stripes in many breeds, foals and hybrids, plus an inquiry about face-stripes in Kattywar horses prompted by the hybrid case.

    1. 1Observation///The stripes were directly seen by Darwin and by several named observers, across breeds and countries.
    2. 2Hypothesis//A latent tendency to reappear is stated clearly enough that failure to find the pattern would count against it.
    3. 3Prediction//The face-stripe query to Colonel Poole was a prediction that came out as expected.
    4. 4Experiment//Crosses between breeds and species act as a form of comparison, though they are not designed tests of a single factor.
    5. 5Replication//Several independent observers report the same pattern, mostly second-hand.
    6. 6Falsifiability//The rival of ancient crossing with a dun stock is weighed and rejected with a stated reason.

    What would move it up Quantify stripe frequency by coat colour and age in large, documented samples, and test inheritance in controlled crosses.

  7. C7InferredC

    The horse, ass, hemionus, quagga and zebra descend from a common parent that was striped like a zebra.

    “I see an animal striped like a zebra, but perhaps otherwise very differently constructed, the common parent of our domestic horse”

    Evidence offered The reappearance of stripes, especially in hybrids and foals, interpreted as reversion to an ancestral character, and the parallel with pigeon breeds crossed to produce the blue ancestral type.

    1. 1Observation/The ancestor itself is not observed, only stripe patterns in living animals.
    2. 2Hypothesis//A specific ancestral state is proposed that later evidence could contradict.
    3. 3Prediction/Stripes in hybrids were expected on the hypothesis, but the same observation fits other readings of latent tendency.
    4. 4Experiment/Pigeon crosses give a real model for reversion, but no test is made on equids for the ancestral state.
    5. 5Replication/The pigeon parallel is repeated, but the equid ancestor has no independent check in the chapter.
    6. 6Falsifiability//The crossing rival is considered, but the possibility of striping as a repeated independent variation is left open.

    What would move it up Independent phylogenetic, genetic and fossil evidence for the ancestral coat state of equids, showing that the stripe genes are retained from an ancestor rather than newly arisen.

Reasoning check

  • Clear case
    Claim C2
    “On the view that each species has been independently created, with all its parts as we now see them, I can see no explanation.”

    Darwin's positive case is real: descent with incompletely fixed change predicts the pattern, and comparing rivals is legitimate. But the rival is dismissed because he cannot see an explanation, not by stating what it predicts and checking that against the data. The same move recurs in 'I do not see that any explanation can be given' for specific versus generic characters.

  • Possible
    Claim C1
    “we have reason to believe that it is the steady accumulation of beneficial differences which has given rise to all the more important modifications of structure in relation to the habits of each species.”

    The evidence offered is real: large differences among pigeon breeds and natural cases of loss of wings, eyes and carapaces. But the natural cases are reductions of existing parts, the chapter admits profound ignorance of how variation arises, and the leap is to the construction of all important structures. An observed case of stepwise construction of a new complex structure would settle it, and the fuller argument is deferred to other chapters.

  • Possible
    Claim C2
    “There may truly be said to be a constant struggle going on between, on the one hand, the tendency to reversion to a less perfect state, as well as an innate tendency to new variations, and, on the other hand, the power of steady selection to keep the breed true.”

    Darwin states the difference openly (it is 'the will of man' that accumulates variation, which 'answers to' survival in nature) and gives the Madeira beetles as a consistent pressure acting without foresight. But the parallel for variability in recently modified organs assumes natural pressures are as steady and directional over long periods as a fancier's fixed goal, and this is not measured. Data on the direction and persistence of selection in wild populations would settle it.

  • Possible
    Claim C2
    “I cannot make out that it applies to plants, and this would have seriously shaken my belief in its truth, had not the great variability in plants made it particularly difficult to compare their relative degrees of variability.”

    The rule of high variability in unusually developed parts fails to appear in plants, and the reason given is that comparison is difficult because plants are so variable. This is plausible and candidly stated, but it is offered only to protect the rule and has no test of its own. Variability measured in plants against specified comparators would settle it.

What it does well

  • Darwin openly admits the limits of his knowledge, writing that 'our ignorance of the laws of variation is profound', and does not claim that selection explains the origin of variation.
  • He keeps variation and selection separate in the agent language: 'variability is in some manner excited, but it is the will of man which accumulates the variations', and he says the conditions 'determine whether this or that variety shall survive'.
  • The Madeira beetle argument states a mechanism in terms of measurable traits (flying least, being blown out to sea) and draws testable consequences such as the exposed Desertas and enlarged wings in flower-feeding insects.
  • The equid stripe case combines many independent observers, a prediction check (the face-stripe inquiry), and an explicit consideration and rejection of the rival crossing hypothesis.
  • He reports counterexamples and confounds, such as the furriers' fur example, shells that do not always conform, and blind Bathyscia, and proposes a clear experiment on kidney-bean hardiness while noting it has not been done.

Limits of this evaluation This is one of 17 sections of a historical work, and Darwin defers support to other chapters and to a promised future work. Per the rules, those gaps are rated as they stand here. He is not marked down for lacking genetics, but later work shows no inheritance of the effects of use, disuse or mutilation, which counts against that part of the chapter. On the focus: reification was not diagnosed because phrases like 'trying to economise' and 'the power of steady selection' are backed by spelled-out survival mechanisms and the argument works without the agent words. 'Favoured' is tied to independent traits (flight propensity, nutriment wasted), though the advantage of saved nutriment is assumed rather than measured, so tautology was not diagnosed. Filter-as-source was not diagnosed because variation is named as the source and said to be unknown. The remark that independent creation 'makes the works of God a mere mockery and deception' is an openly stated theological aside, and the descent argument works without it. Reliance on what 'naturalists' will admit about secondary sexual characters is only a mild appeal to consensus, since a few examples are given.