Myth 3: Darwin was actually a Lamarckian
This one is subtle. It implies that Darwin, because he lacked a Mendelian account of heredity, was not actually a "true" (or Neo-)Darwinian. The error depends on the extent of what is named as a school of thought in science and why.
As far as I know, the term "Lamarckian" for those who think heredity is acquired during the parental generation and passed on to progeny, was coined by August Weismann, a pre-genetics developmental biologist with an interest in heredity. He contrasted this with his own view, that germ cells - the cells in which inheritance is passed on (sperm and egg in animals) - are isolated from what happens to the post-conception body, or in his terminology, they are sequestrated, now called the "Weismann Barrier". This was 60 years or so before DNA was given structure as a mechanism, and 10 years before genetics proper got going, so it really has little to do with DNA.
Darwin held the common view that inheritance was strengthened or weakened by the amount a trait was used by the maturing animal; this was not unique to Lamarck, and goes back to ancient accounts in which, via sympathetic magic, the developing fetus could be modified by the experiences of the pregnant mother. An example of this is in Genesis chapter 30. It is not right to call this "Lamarckian" or "Lamarckism", as it was neither invented by Lamarck, nor is it strictly the sin that Weismann named thus, of thinking that the traits that were acquired during the lifetime of the parent would be passed on to the progeny. Darwin did not think that at all. He held that the traits arose in ways that were amenable to investigation (only not as yet), which caused variation. But because he had a problem of the loss of variation by averaging, he needed to explain how variation could be inherited so that natural selection could act upon it.
R. A. Fisher has a nice explanation of this in the first chapter of his 1930 Genetical Theory of Natural Selection . Assuming as (nearly) all did in Darwin's day that heredity blends - that is, the parent's individual differences average out in the progeny - after around 1000 generations all variance is lost. So variance needs to be maintained either by some other factor, such as use and disuse - Darwin's term for the mechanism we wrongly term "Lamarckian" - or natural selection needs to be many factors higher than we think it really is. It is often said that if Darwin had read his Mendel, he'd have seen the solution to the blending problem, but I doubt it. Yes, Mendel's hypothesis of particulate inheritance in fact does solve the blending problem, but it took scores of geneticists nearly twenty years to show this in the case of natural selection, the first proof being given in William Castle's work in the second decade of the century and the new science of genetics, and later developed in detail by Fisher, Sewall Wright, and J. B. S. Haldane, among others. Whether Darwin would have been able to free himself from what he thought was a vast amount of empirical evidence, mostly from breeders and farmers, of use and disuse, and hence of blending, is debatable.*
But here's what Darwin did not think: that the variation itself was acquired from parental experience, which is what we typically think of as being Lamarckian. This is what Mayr termed soft inheritance. Darwin held that variation arose in ways that he did not understand but were likely to be explicable in natural terms, but which were not correlated to the present or future needs of the organisms, just as we do today. Any variant that was used was more strongly inherited, and he developed his ill-fated Pangenesis theory to account for that "fact", but this didn't make the variants occur or more likely. So in any meaningful sense, Darwin was no Lamarckian.
The very use of the term "Lamarckian", like the use of the contrasting term "Weismannian" (Griesemer 2005), is an act of battle. It raises a standard under which scientists and those fellow travellers like myself can rally against the enemy. To say that Darwin was a "Lamarckian" (especially when he was not in this regard) is a way of strengthening those who have, by their own account, refined Darwin and stripped him of error. Poor Darwin, who lacked the resources and data to know better! We have fixed his mistakes. On the other side, those who continually assert their heterodoxy by calling themselves "Lamarckian" (e.g., Jablonka and Lamb 1995, Steele, Lindley and Blanden 1998) seek to mark out their originality and again, claim to be fixing the mistakes of the past. Steele may be a Lamarckian in Weismann's (though I rather doubt in Lamarck's) terms, but Jablonka and Lamb are not. They are anti-Weismann, not anti-Darwin. The existence of mechanisms for taking somatic, non-nuclear, inheritance into subsequent generations in no way undercuts Darwin. Arguably it doesn't even undercut Weismann, who held that inheritance at the cellular level was sequestered, not that it was DNA based (DNA at that time had not even been discovered, let alone given a role in inheritance). And from an evolutionary perspective, we need to show that this inheritance, termed "epigenetic", persists long enough to have an evolutionary effect, which to my knowledge hasn't yet been done.
Rather ironically, the history of the term "Lamarckian" shows us that there were a group of biologists who meliorated some of Darwin's focal ideas - in particular the role of natural selection - who termed themselves "Neo-Lamarckians", against whom Weismann strove. The tumult of ideas that were about in the late nineteenth century, a period in which natural selection was largely ignored or treated as secondary (Bowler 1983), led directly to the establishment of Mendelian genetics, and so to be "Lamarckian" meant one was immediately excluded from the new science, although as Koestler both documents and demonstrates (1971), neo-Lamarckism was still around late into the twentieth century. Natural selection itself was deprecated by the biological community from around 1870 or so to 1920, when mathematical models began to show its efficacy in a Mendelian world. Empirical support took even longer to acquire, although it is now ubiquitous (Bell 1996).
We need to be careful not to make any historical figure what they are not, and as I have repeatedly said, this is a common flaw in the use of history by scientists. Darwin was no Lamarckian. He was himself, and his errors are his own.
* Darwin's use of anecdotal evidence is a singular moral lesson on why anecdote is not the singular of data. He saw what he expected on the basis of others seeing what they expected. It is to overcome these biases of seeing what you expect that Pearson and Fisher, among others, developed the use of statistical analysis of data sets. In a nice historical irony, both were continuing what they thought of as the Darwinian program.
Bell, Graham. Selection: The Mechanism of Evolution. New York: Chapman and Hall, 1996.
Bowler, Peter J. The Eclipse of Darwinism: Anti-Darwinian Evolution Theories in the Decades around 1900. Baltimore and London: John Hopkins University Press, 1983.
Castle, William E. "The Effect of Selection Upon Mendelian Characters Manifested in One Sex Only." Journal of Experimental Zoology 8, no. 2 (1910): 185-92.
———. Heredity: In Relation to Evolution and Animal Breeding. New York, London: D. Appleton and Company 1911.
———. Studies of Inheritance in Guineapigs and Rats. Vol. Pub. No. 241. Washington: Carnegie Institution of Washington, 1916.
———. "Are Genes Linear or Non-Linear in Arrangement?" Proceedings of the National Academy of Sciences USA 5 (1919): 500–06.
———. "Is the Arrangement of the Genes in the Chromosome Linear? ." Proceedings of the National Academy of Sciences USA 5 (1919): 25–32.
———. "Piebald Rats and Selection: A Correction." Amer. Naturalist 53 (1919): 370.
Fisher, Ronald Aylmer. The Genetical Theory of Natural Selection. Oxford UK: Clarendon Press, (rev. ed. Dover, New York, 1958), 1930.
Griesemer, James R. "The Informational Gene and the Substantial Body: On the Generalization of Evolutionary Theory by Abstraction." In Idealization Xii: Correcting the Model. Idealization and Abstraction in the Sciences (Pozna, edited by Martin R. Jones and Nancy Cartwright, 59-115. Amsterdam: Rodopi Publishers, 2005.
Jablonka, Eva, and Marion J. Lamb. Epigenetic Inheritance and Evolution: The Lamarckian Dimension. Oxford; New York: Oxford University Press, 1995.
Steele, Edward John, Robyn A. Lindley, and Robert Vincent Blanden. Lamarck's Signature: How Retrogenes Are Changing Darwin's Natural Selection Paradigm, Frontiers of Science. St Leonards, N.S.W.: Allen & Unwin, 1998.
As far as I know, the term "Lamarckian" for those who think heredity is acquired during the parental generation and passed on to progeny, was coined by August Weismann, a pre-genetics developmental biologist with an interest in heredity. He contrasted this with his own view, that germ cells - the cells in which inheritance is passed on (sperm and egg in animals) - are isolated from what happens to the post-conception body, or in his terminology, they are sequestrated, now called the "Weismann Barrier".
You can tell that Weismann worked on animals rather than plants, can't you?
It is often said that if Darwin had read his Mendel, he'd have seen the solution to the blending problem, but I doubt it. Yes, Mendel's hypothesis of particulate inheritance in fact does solve the blending problem, but it took scores of geneticists nearly twenty years to show this in the case of natural selection...
If I recall my history of biology correctly, wasn't Mendelian inheritance seen by at least some researchers as a significant problem for Darwinian evolution prior to the Modern Synthesis? Because early research into heredity tended to focus on clearly-marked either/or characters (such as Mendel's peas being either tall or short, without medium-height plants), rather than more continuously varying characters (single-gene rather than multi-gene characters, to use modern terminology), Mendelian inheritance was initially thought to suggest a more saltational mode of evolution.
I've been enjoying these posts, by the way.
I don't get it: it seems from what you write that for Darwin use and disuse would not cause variation, but affect the pattern of inheritance of said variation.
Darwin held the common view that inheritance was strengthened or weakened by the amount a trait was used by the maturing animal.
That would entail that Darwin had a fairly articulated view of the difference between variation and inheritance. He obviously had some idea about that (just think of the common view of 'adaptation' of individuals to their environments), but I think you are attributing him too much insight. I don't see how to reconcile your views (if I understand them) with things like the following, from ch. 5 of the Origins.
From the facts alluded to in the first chapter, I think there can be no doubt that use in our domestic animals has strengthened and enlarged certain parts, and disuse diminished them; and that such modifications are inherited.
My impression is that Darwin thought that the study of the causes of variation was already within his reach. These causes would have included use and disuse, and changes in the conditions of life, while inheritance could be (momentarily) taken as given.
This is a most excellent post, and thank you for it! I am going to totally link this up in my latest post about a study that claims evidence for "Lamarckian" inheritance... somewhat wrongly.
Anyhoo, there's much that I (always!) can learn about this stuffs.
John, are you familiar with a rather entertaining gloss on Darwin's life called 'Darwin: the Man and his Warfare', by Henshaw Ward? I have an original (1928) edition. Ward argues rather forcefully that Darwin never embraced any full-blown version of Lamarckism. He also argues that Lyell never became an 'evolutionist', or at least not one that would satisfy Darwin. Interesting stuff...SH
Don't know it, nor can I find an online copy...
I need help on something:
How many distinct lineages did Lamarck think there were? Did he believe that every living species had evolved more or less in parallel from spontaneously generated simple organisms, making each species its own lineage? (In other words, the similarities between, say, a cat and a dog were due to them evolving in parallel, but each had a separate origin as a very simple organism rather than recent separation of their lineages.) Or did he merely think that there were only several origins due to spontaneous generations and that similarity between species were generally due to common descent?
He does discuss wild and domestic variants, so there he does accept recent division of lineages, but otherwise I'm confused on precisely what his views are on common descent vs separate parallel lineages, given his emphasis on continuous spontaneous generation. Or was he deliberately vague to give himself some wiggle room?
Lamarck thought that every new monad established a new evolving lineage. A species was just an arbitrary segment of that continuously evolving lineage. There was a largely constrained pathway, with some alternative branches, in the Philosophie Zoologique, but it isn't a phylogeny so much as a scala naturae in temporal terms. The constraint was physical - not so much convergent evolution as this was all that a monad could achieve.
The Zoological Philosophy (the English translation) is available from The Internet Archive, and the "tree" is on page 179. In the original, the lines are dots, indicating a progression rather than a historical link.
That is a weird notion. Perhaps the most important post-Lamarckian advance in evolutionary theory is not natural selection, but common descent from a single primordial ancestor (or at least a few of them rather than continuous generation). Or were there advocates of common descent before Lamarck?
Lamarck himself did not hold to common descent. The first person to publish that view of whom I know (there are probably others) is Heinrich Bronn, in 1858. He beat Darwin and Wallace by a matter of months, and we know Darwin had common descent before he had natural selection, around 1837 or so. I think Darwin is original in that regards.
But his grandfather had written "...would it be too bold to imagine, that all warm-blooded animals have arisen from one living filament, which the THE GREAT FIRST CAUSE endued with animality...?" in his . He was less over the top elsewhere: "All animals therefore, I contend, have a similar cause of their organization, originating from a single living filament, endued indeed with different kinds of irritabilities and sensibilities, or of animal appetencies; which exist in every gland, and in every moving organ of the body, and are as essential to living organization as chemical affinities are to certain combinations of inanimate matter." He goes so far as to unite "vegetable" and animal life to a shared "filament": "Shall we then say that the vegetable living filament was originally different from that of each tribe of animals above described? And that the productive living filament of each of those tribes was different originally from the other? Or, as the earth and ocean were probably peopled with vegetable productions long before the existence of animals; and many families of these animals long before other families of them, shall we conjecture that one and the same kind of living filaments is and has been the cause of all organic life?"
So the notion that some living things share a common cause is not novel. But Lamarck did not hold to it.
Thanks for the information. This must be Lamarck's greatest error - and is it a doozy. My conclusion: Darwin rules.
I made an error above - Franz Unger had a theory of common descent in 1852. My bad.
I object to your connection of Lamarckian evolution to sympathetic magic. Situations like that in Genesis aren't that similar. One could conceive of inheritance mechanisms that are inherently Lamarckian (if for example, each body part in some way contributed information for the creation of the corresponding body part in the offspring). Lamarck's idea is thus much more reasonable than what Jacob does in Genesis 30 where uses speckled sticks to make the offspring speckled . It is a disservice to Lamarck and the scientists who worked with his ideas to make that comparison.
As I tried to make clear, "Lamarckian" is a bit of a misnomer. The popular and traditional notions about inheritance derive from folk biology, and the Genesis case is one of those folk biological notions. You are right about Lamarck - he is a much better thinker than the folk biology around him. I was not suggesting he was relying on sympathetic magic. In fact he was a rather hard materialist when it came to biology.
Changing the topic slightly, it's worth noting that in all the celebrations this year that implicitly concatenate evolution per se with Darwin's Origin, the fact that this year is also the 200th anniversary of Lamarck's theory of evolution is studiously never mentioned.
Whatever Lamarck actually meant or didn't mean in his ideas about inheritance (and do we incessantly throw pangenesis in Darwin's face every time we mention his scientific works?), his contributions to biology were enormous (including giving us the word "biology").
If in 2009 we intend to celebrate Darwin as person and scientist, well and good. However when we expand the festivities to include the theory of evolution in general, then Lamarck deserves better than to be a non-person at the party.