The Friendliest Gene

In 1961, the cardiologist J. C. P. Williams described a syndrome he had not seen described before: children with a particular cardiac defect who also displayed an unusual constellation of traits — elfin features, a love of music, profound intellectual disability in some domains, and, most strikingly, an almost total absence of the social wariness that typically develops in infancy. These children approached strangers with warmth and trust so complete it looked, to some observers, like joy. Williams-Beuren syndrome, as it came to be called, results from the deletion of about twenty-five genes on chromosome 7. Among them, researchers would eventually discover, are genes that also appear in the dog.

In 2017, a team led by evolutionary biologist Bridgett vonHoldt at Princeton University published a paper in Science Advances that drew a line between two ideas that had seemed, until then, entirely unconnected: the genetic architecture of human hypersociability and the evolutionary mystery of how the dog became the only animal in four billion years of life on Earth to volunteer, and then sustain, a genuine social partnership with our species.

The Genomic Turn in Domestication Science

For most of the twentieth century, the story of dog domestication was told in terms of behaviour: wolves that were less fearful of humans were tolerated around camp fires, got food scraps, reproduced more, and over generations were selected toward tameness. The story was plausible and widely repeated. What it lacked was a mechanism — a molecular account of how a wild animal becomes, within a few hundred generations, the creature that searches your face when you are sad.

vonHoldt's team compared the genomes of 18 dogs and 10 grey wolves, looking for structural variants — insertions, deletions, rearrangements — that differed systematically between the species. What they found in a genomic region orthologous (functionally similar) to the human Williams-Beuren syndrome critical region (WBSCR) on chromosome 6 of the dog stopped them.

The team found that dogs carried deletions and insertions in a region containing several genes — including GTF2I and GTF2IRD1 — that in humans are deleted in Williams-Beuren syndrome and are associated with the condition's characteristic hypersociability. The more disrupted a dog's versions of these genes, the more time it spent in proximity to and seeking contact with an unfamiliar human during a behavioural test. Wolves, by contrast, showed little variation in the region and spent far less time with humans regardless.

Williams-Beuren and What It Illuminates

Williams-Beuren syndrome affects approximately 1 in 7,500 people worldwide, according to the National Institutes of Health. People with the syndrome typically have excellent verbal ability and musical facility alongside significant difficulty with spatial tasks. But the most socially salient feature is what clinicians call "hypersociability" — an absence of the normal social wariness that leads most people to be cautious around strangers. People with WBS often describe everyone they meet as a potential friend and show heightened oxytocin responses to social cues.

The GTF2I and GTF2IRD1 genes implicated in this trait encode transcription factors — proteins that regulate the expression of many other genes. Their role in social behaviour is not yet fully understood at the molecular level, and vonHoldt's team was careful to describe the relationship as an "association" rather than a direct cause. The study design was observational; the sample was small by genomic standards. What it offered was not a definitive mechanism but a striking hypothesis: that the domestication of the dog may have involved, at least in part, a chance variant in the molecular architecture of sociability — one that made certain wolves not just tolerant of humans but drawn to them.

A Different Kind of Domestication

What makes this finding significant beyond its novelty is what it implies about the nature of domestication itself. Most animals we have domesticated — cattle, sheep, horses, chickens — were domesticated for utility: meat, fibre, traction, eggs. The relationship is transactional and does not require genuine mutual affiliation. Dogs are different. The archaeological and genetic evidence now fairly firmly places the split of domestic dogs from wolves somewhere between 15,000 and 40,000 years ago (the range reflects ongoing debate about single versus multiple domestication events), and throughout that entire span, the primary thing dogs have provided humans is not traction or protein but companionship. They stayed close. They watched our faces. They learned to read us.

The Limits of the Parallel

Scientists who study Williams-Beuren syndrome have been, in the main, cautious rather than hostile about the vonHoldt findings. The parallel is intriguing, they note, but imperfect. WBS in humans involves the deletion of all twenty-five or so genes in the critical region; the variants vonHoldt identified in dogs are structural rearrangements that may alter expression rather than eliminate function. The behavioural phenotype in dogs — sociable, attentive, gaze-seeking — overlaps with WBS hypersociability but is not identical to it, and dogs with the most disrupted versions of the genes were not, by any measure, cognitively impaired in the way WBS individuals often are. The dogs were, in short, friendly rather than differently-abled.

Subsequent research has begun to map the molecular picture in more detail. A 2021 paper in PLOS Genetics by Ségurel and colleagues, examining GTF2I expression across tissues in dogs versus wolves, found differential expression patterns consistent with a role in social behaviour — adding a layer of mechanistic plausibility to vonHoldt's structural observations. The picture is not yet complete, and the scientists involved say so plainly. But the direction of travel is consistent.

A Stranger Who Feels Like Family

There is something both humbling and quietly moving in this research. The dog that sits beside you, tracking your face, angling for proximity — that behaviour may not be purely learned. It may be, in some part, a genetic predisposition so deep it parallels the machinery of a human syndrome. Dogs may not just have been trained to like us. Some of them may have been built to.

The question of whether that constitutes a kind of love is one that science is not yet equipped to answer, and may never be. But it changes the register of the relationship. You are not just a food source that has been classically conditioned into a social partner. You may be, from the dog's genomic perspective, something that feels, for reasons it cannot articulate and that predate language by tens of thousands of years, like home.