I recently read Arthur Koestler's 1967 book The Ghost in the Machine. In it, Koestler criticises the neo-Darwinian theory of evolution—beneficial random mutations preserved by natural seleciton—as insufficient to explain the formation of complex forms like eyes and eggs. The issues Koestler has with the theory are ones that I've been trying to wrap my head around since before I read the book, but I'm aware that:

a) the book is half a century old;

b) Koestler was not a biologist or scientist; and

c) neo-Darwinian theory/the modern synthesis seems to have stood the test of time

so I'm wondering how accurate Koestler's account of the theory is, and if he is wrong what the retorts to his claims are. Here are two examples he gives of complex forms:

[The giant panda] has on its forelimbs an added sixth finger, which comes in very ‘handy’ for manipulating the bamboo-shoots which are its principal food [but] that added finger would be a useless appendage without the proper muscles and nerves [and t]he chances that among all possible mutations those which produced the additional bones, muscles and nerves should have occurred independently are of course infinitesimally small.


The decisive novelty of the reptiles was that, unlike amphibians, they laid their eggs on dry land...[b]ut the unborn reptile inside the egg still needed an aquatic environment...[i]t also needed a lot of food...[s]o the reptilian egg had to be provided with a large mass of yolk for food, and also with albumen—the white of egg—to provide the water. Neither the yolk by itself, nor the egg-white itself, would have had any selective value...[e]ach change, taken in isolation, would be harmful, and work against survival.

Instead of random mutations and external selection, he suggests that ‘internal selection’ works at all levels, from chemical upwards, to correct ‘misprints’ long before the developed organism is exposed to any sort of external selection. The implication that there must therefore be some plan towards which embryonic development works is supported by two examples:

the growing eye-bud of the embryo is an autonomous holon, which, if part of its tissue is taken away, will nevertheless develop into a normal eye


[the fruit fly has a recessive gene that when paired with another in a fertilised egg will produce an eyeless fly.] If now a pure stock of eyeless flies is made to inbreed, then the whole stock will have only the ‘eyeless’ mutant gene, because no normal gene can enter the stock to bring light into their darkness...within a few generations, flies appear in the inbred ‘eyeless’ stock with eyes that are perfectly normal.

His other main point is that evolution takes a zig-zag path, evolving down until reaching an evolutionary dead-ends before retracting to ‘an earlier or more primitive, but also more plastic and less committed stage—followed by a sudden advance in a new direction’. For example:

[A]mphibian...ancestry...goes back to the most primitive type of lung-breathing fish; whereas the apparently more successful later lines of highly specialised gill-breathing fishes all came to a dead end.


...the human adult resembles more the embryo of an ape than an adult ape

Is Koestler's science just faulty, or are these valid criticisms that've been resolved since?

  • $\begingroup$ I am entirely unfamiliar with Koestler, but based on the excerpts you posted it sounds like he just making assertions to justify his own hypothesis. Does he make reference to any actual experiments? $\endgroup$
    – canadianer
    Dec 4, 2018 at 1:38
  • 1
    $\begingroup$ Regarding the panda example: polydactyly occurs in many mammals often as the result of a single mutation. This example is commonly used in teaching developmental biology as it shows the effects that single mutations can have on tissue patterning during embryogenesis. You can read the Wikipedia article for a primer on this. If such a mutation happens to have a selective benefit for pandas, it is entirely plausible that it could take hold in a population. $\endgroup$
    – canadianer
    Dec 4, 2018 at 1:44
  • $\begingroup$ Regarding the reptile example: I don't have any specific information on this, but evolutionary theory does not say that deleterious mutations cannot become established in a population (assuming for the sake of argument the mutations leading to reptilian eggs were in fact deleterious). It depends on the strength of selection against them, among other factors such as genetic hitchhiking or heterozygote advantage. $\endgroup$
    – canadianer
    Dec 4, 2018 at 1:57
  • $\begingroup$ @canadianer For the zig-zag theory of evolutionary progress he cites Garstang, for the embryonic tissue reformation Weiss. $\endgroup$
    – 08915bfe02
    Dec 4, 2018 at 2:04
  • $\begingroup$ Regarding the fruit fly example: It is difficult to comment without knowing the details of the actual experiment. A well studied gene in fruit flies is called eyeless (actually the Pax6 transcription factor with homologs in all sighted animals studied) because mutations to it can prevent eye development. In such mutants, it is hypothetically possible for back mutations to restore the original phenotype (I'm not aware if this has actually been observed, but I'm not aware of a lot of literature on fruit flies). $\endgroup$
    – canadianer
    Dec 4, 2018 at 2:07

3 Answers 3


Maybe both. Certainly his understanding of limb developement doesn't match modern understanding. The quote you provide seems to indicate that he thought the appearance of an additional digit would require a multitude of coordinated mutations. Our current understanding seems to indicate that it's all about changes in the regulation of genes. For example in fruit flies mutations in the antennapedia gene can cause antenna to grow where legs should be or visa versa. Another example is polydactyly(extra toes) in cats which be be caused by mutations to a single regulatory region.

within a few generations, flies appear in the inbred ‘eyeless’ stock with eyes that are perfectly normal.

I'm not a fruit fly expert, but this seems highly dubious to me. There is a well studied eyeless mutation, but it's dominant, not recessive. Furthermore getting two copies of this mutant eyeless gene is lethal at the embryonic stage. This means that a colony of adult eyeless flys are necessarily all heterozygous. That is, they have one copy of the mutated eyeless gene, and one copy of the normal eyeless gene. By Mendel's laws, a quarter of the offspring from the completely eyeless generation are going to have two copies of the normal eyeless gene, and so will have normal eyes. I wouldn't be surprised if the flies with normal eyes outcompete the flies without eyes, so eventually the eyeless flies will disappear from the colony. I can't tell you whether Koestler has mis-characterized this gene, or is talking about some other gene all together. Citation needed.


Koestler's entire argument can be summed up is an argument of irreducible complexity, which always stems from a lack of understanding of the biological systems. All his arguments stem from a lack of knowledge about the things he is describing. Lets handle them in order.

  1. Panda thumb. lack of understanding. The panda thumbs does not move, it is just a spur of bone. but becasue it is a wrist bone it still has muscles on it so the argument is wrong in two ways.

  2. Reptile eggs. Amphibian eggs have yolks and albumen. so this again is just a lack of knowledge.

  3. Internal selection is again based on a lack of knowledge, the "plan" embryonic development works from are called HOX genes, which regulate development. And there are organisms without them and indeed behave just like you would expect. Eyeless fruit flies do not lack the genes to make eyes they have a mutation in the HOX genes that trigger eye development, (lake a "make eye here" on a blueprint) and just as single mutation can prevent the gene from functioning, another can make it functional again, we can can even trigger eye development in the wrong place, say on the legs.

  4. Is just confusing, It sounds like he thinks there needs to be some overarching plan to evolution instead of the chaotic whatever works mess that actually happens. In which case he does not even understand what evolution is, so his arguments against it make even less sense. I certainly can't imagine why he would claim teleost fish are dead ends when they are the most numerous vertebrate on the planet.


Time — i.e. the advances in our knowledge of molecular genetics — has clearly disposed of Koestler’s arguments, as it has of many of those of people who cannot see the wood for the apparently diseased tree.

Other answers provide chapter and verse, but what I wish to contribute in this answer is something about the intellectual bias that influenced Koestler’s views. Most of us will be aware that many anti-evolutionary arguments are made from religious conviction, rather than from disinterested scientific rationalism, but the poster and the younger members of this list may not be aware of a different kind of intellectual bias.

Although Koestler is most famous for his book “Darkness at Noon” in which he attacked the evils of Stalinism, after which he rejected the Communist party, this did not change the (in many respects admirable¶) socialist principles that had lead to his joining the party in the first place.

So, what does this imply? It does not imply a rejection by Koestler of Darwin’s theory of evolution, but a rejection of the idea of a genetic mechanism in which man is subject to the dice of random mutation. He still clings to a doctrinal belief in Lamarckism in which man evolves by the incorporation of characteristics acquired by his own efforts†. It is ironic that despite repudiating Stalin, he hung on to Stalin’s dogmatism that led to Lysenko and the consequent murder of Soviet geneticists as well as the death by starvation of untold thousands of Russians.

And the moral for today? There are several. But it is politic to restrict myself to suggesting one should question the motives of those who cry “epigenetics” when they haven’t the first idea of what they are talking about.

¶ In my opinion. (Not particularly relevant, but to show I am not criticizing him from a right-wing standpoint.)

† The actual context at the time he wrote it was different — a rejection of social Darwinism. I personally regard the latter as having nothing to do with genetics or biology, but in order to repudiate the one, he seems to have returned to his repudiation of neo-Darwinism.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .