The French biologist Jean-Baptiste de Lamarck (1744–1829) outlined a theory of evolutionary change in 1809, 50 years before Darwin's On the Origin of Species. Lamarck's basic idea was that organisms change in adapting to their environment and then pass on to their offspring the new characteristics they have acquired. Since then, Lamarck has been ridiculed for presumably implying, for example, that giraffes developed their long necks by stretching, generation after generation, toward the leaves of trees. Most modern biologists are adamant that nothing of the sort occurs, ever. But the molecular immunologist Edward J. Steele is attempting to revive Lamarckism: he and his colleagues claim to have found evidence for a Lamarckian hereditary mechanism in the immune system.

The immune system is an evolutionary puzzle in its own right: How is it that our bodies can quickly respond to so many different kinds of attacks? Is all this information in the genes? If so, then how does our immune system defend against new diseases? Part of the answer comes from the fact that some immune system cells contain genes that mutate with unusual frequency. The most common type of mutation is a sort of genetic "typo" that occurs when a cell's DNA is transcribed into RNA, the molecule that helps to assemble proteins. These mutations allow the immune system to test out different defenses until it finds one that does the job.

Steele hypothesizes that the altered RNA then reverts back into DNA. Indeed, such "reverse transcription" of RNA back into DNA has been observed frequently in other contexts. But the troublesome question for Lamarckians is this: Could this new DNA then be carried to the reproductive genes (in the sperm and egg cells), replace the original DNA there, and so be passed on to an organism's offspring? Steele and company believe this is possible, and they have devised an elegant, but speculative, story to describe how it might happen using known biological mechanisms. They believe a virus could carry the altered DNA to the reproductive cells and replace the DNA in those cells.

But even if the process Steele and his colleagues describe is possible, does it ever actually occur? Evolutionary mechanisms are never observed directly, so we must make do with circumstantial evidence. Steele and his colleagues claim to have found such evidence, namely a "signature" of past events that is "written all over" the genes that carry instructions for immune system responses. They claim that a distinct pattern of mutations concentrated in particular areas of these genes "strongly suggests" that, in the past, information has been transferred into DNA in the reproductive organs. Other biologists are not so easily swayed. They suggest there may be other, less radical explanations for the pattern of mutations that Steele cites.