Cloning an animal would produce an exact physical and behavioral duplicate of the original.

On February 22, 1997, Ian Wilmut's team at the Roslin Institute in Edinburgh announced that they had produced a live lamb from the nucleus of an adult mammary gland cell. The lamb's name was Dolly, chosen in honor of Dolly Parton in recognition of the cell's anatomical origin. The announcement dominated front pages for weeks. Scientists and journalists reached immediately for a single explanatory metaphor: the photocopy. Cloning, readers and viewers were told, would produce biological duplicates, copies of people, identical in every way that mattered.

The metaphor was wrong.

Wilmut's paper, published in Nature on February 27, 1997, described a technical achievement that had seemed impossible: somatic cell nuclear transfer, in which the nucleus from a differentiated adult cell is introduced into an enucleated egg cell, reprogrammed to an embryonic state, and coaxed through development. Dolly was genetically identical to the Finn Dorset ewe who had donated the mammary cell. But Dolly was not a copy of that ewe. She could not be.

The reason lies in the distinction between a genome and an organism. DNA carries instructions, but it does not carry the reading of those instructions. Epigenetics, the patterns of chemical modification that determine which genes are active and which are silenced in a given cell, is shaped by environment, timing, and developmental history in ways that cannot be transferred by copying a nucleus. Two organisms with identical DNA may activate their genes in substantially different sequences and patterns, producing different physical and behavioral outcomes.

In Dolly's case, this was visible almost immediately. Her telomeres, the protective sequences at the ends of chromosomes, were shorter than expected for a lamb her age, resembling those of the six-year-old sheep who had donated the cell, suggesting that the reprogramming had been imperfect. She developed arthritis at an unusually young age, along with lung disease, and died in 2003 at six years old, well short of the typical sheep lifespan. Whether these conditions were related to her origin remained debated, but they made clear that cloning was not a technology for producing perfect replicas.

Subsequent cloning experiments reinforced the point. CC the cat, the first cloned domestic cat, produced by Texas A&M University in 2001, was a genetic copy of a calico cat named Rainbow but bore different coat markings, a consequence of the random inactivation of one X chromosome in each cell during development, a process that produces calico patterning and that cannot be predicted or replicated from nuclear DNA alone. Behaviorally, CC differed from Rainbow as well, being more playful and less timid.

The public understanding of cloning that crystallized in 1997 drew on science fiction, where clones had long been imagined as interchangeable replicas. That fiction had a coherent internal logic: if DNA is the blueprint for a person, a copy of the blueprint should produce a copy of the person. The logic overlooked the building process. A blueprint for a house does not tell you how long the timber cures, what the weather was during construction, or who hammered which nail. The process of development, from fertilized egg through gestation and into a lifetime of environmental exposure, is not recorded in the DNA and cannot be reconstructed from it.

By the early 2000s, the science of epigenetics had developed enough to explain what cloning experiments kept demonstrating: genetic identity is not biological identity. Two organisms sharing a genome are more like identical twins than like copies, and even twins, who develop in the same uterus at the same time, diverge substantially over a lifetime. A clone born into a different body, a different era, a different environment, would diverge further still.

The photocopy never existed. It was a metaphor built for headlines, not for biology.