On the front panel of a limestone throne base excavated at Nimrud in 1962, two kings are shown gripping hands to seal a pact. The scene commemorates the Assyrian king Shalmaneser III backing the Babylonian king Marduk-zakir-shumi I against a rebellious brother, and it was carved around 846 BCE. It is one of the oldest clear pictures we have of a gesture almost everyone reading this performed within the last week. The usual explanation for where that gesture came from is tidy and reassuring: two strangers extend empty right hands, each proving to the other that the hand most likely to hold a blade is holding nothing. It is a satisfying story, and the evidence behind it is surprisingly thin.
The empty-hand explanation appears almost everywhere the gesture is discussed, usually stated with a confidence the evidence does not quite support. Some versions add that the up-and-down motion was meant to shake loose any dagger tucked in a sleeve. The History channel’s own account calls this “one popular theory” and admits the origins are murky. That hedging is the honest part, and it tends to get lost in the retelling.
What the earliest evidence actually shows
Push past the just-so story and the documentary trail thins quickly. The Nimrud relief, now held in the Iraq Museum in Baghdad, places a recognisable handshake at least 3,000 years back. The Greeks left more: clasped hands appear on vases, reliefs, and funerary art, and Homer describes men grasping hands in the Iliad and the Odyssey, usually as a display of trust. Roman coins carried the motif too.
What none of this proves is why the gesture started. A relief tells you a handshake meant something to the people who carved it. It does not tell you what the first handshake was for.
The Dutch historian Herman Roodenburg, one of the few scholars to study the gesture seriously, put the difficulty plainly. In his work on the cultural history of gesture, he noted that the historian of gesture has fewer clues to work with than almost any other kind of historian. The record is sparse, and much of it arrives already interpreted by whoever recorded it. The classicist Walter Burkert offered the tidiest formulation of the peace reading, describing an agreement as something expressed in words but made effective only by a ritual gesture of open, weaponless hands reaching toward each other. It reads well. Burkert was describing what the gesture had come to mean by his reckoning, drawing on the same thin record as everyone else.
Why the empty-hand story is so easy to believe
Part of the appeal is that it feels like it must be true. Most people are right-handed, most weapons were held in the right hand, and offering that hand open is a clean piece of logic. The dagger-dislodging embellishment adds a satisfying mechanical detail. The trouble is that a story can be plausible and widely repeated and still rest on almost nothing.
Folklorists at the Smithsonian’s Center for Folklife and Cultural Heritage have pushed back on the idea that the handshake runs continuously through human history at all. In their reading of the evidence, greeting gestures wax and wane across cultures rather than descending unbroken from antiquity. They point out how quickly gestures appear and vanish, using the dap and the “call me” thumb-and-pinky sign as recent examples. A gesture that feels ancient and universal can be neither.
The account survives because it is useful. It hands a concrete origin to something we do without thinking, which is much of why it sticks.
A different reading, from a lab in Rehovot
In 2015, researchers at the Weizmann Institute of Science published a study in the journal eLife that approached the question from a completely different direction. Rather than asking where the handshake came from, Idan Frumin and colleagues, working in the laboratory of Noam Sobel, asked what people actually do with their hands afterwards.
The answer, captured on hidden camera across roughly 280 volunteers, was strange. People sniff their own hands a great deal, keeping a hand near the nose around a fifth of the time at rest. After being greeted with a handshake, the time spent with the shaking hand near the face more than doubled. Nasal airflow measurements confirmed the volunteers were sniffing, not just fidgeting. The pattern shifted by gender: after shaking hands with someone of the same sex, people sniffed the hand they had used, and after shaking with someone of the opposite sex, they sniffed the other one. The full paper frames the handshake as a possible channel for social chemosignalling, a socially acceptable way of sampling another person’s scent.
This is one study, not settled consensus, and its authors were careful about that. Asked what signals might travel through a handshake, Frumin said plainly that they could only speculate. The experiment shows that odour molecules transfer and that sniffing increases. It does not show what the brain does with that information, or that scent-sampling is why the gesture exists. It is a finding from this dataset, not a universal rule about everyone.
What we find useful about it is the reframing. The peace theory treats the handshake as a message sent outward, a visible declaration to another person. The Weizmann work suggests part of what is happening runs inward, a quiet act of gathering information that mostly stays below awareness.
The deep-time claim, and why to hold it loosely
There is a bolder hypothesis still. The palaeoanthropologist Ella Al-Shamahi, in her 2021 book The Handshake: A Gripping History, argues that the gesture is not merely prehistoric but older than our species, perhaps seven million years old, and that even Neanderthals likely shook hands. Her case leans on the handshake’s kinship with grasping and greeting behaviours seen in other primates.
It is a genuinely interesting case, and one built on inference rather than direct evidence. The Smithsonian folklorists, reviewing the same terrain, reached the opposite conclusion, holding that the handshake is a learned cultural habit that appears and fades across societies. Both cannot be fully right, and neither can be settled from the fragments available. The evidence simply does not reach that far back.
What we can actually say
The everyday handshake, the one exchanged at introductions and the close of deals, is younger than any of this. Historians often credit the 17th-century Quakers with popularising it as a plain, egalitarian alternative to bowing or doffing a hat, and by the 1800s etiquette manuals were fussing over grip and duration. That part of the history is comparatively firm.
The origin is not, and it is worth being comfortable with that. We can say the gesture is at least three millennia old in recognisable form. We can say it has long carried meanings of trust, alliance, and good faith. When people shake hands today, something measurable passes between them, chemical as well as social. What we cannot claim, with the assurance the familiar telling invites, is that it all began with a wary stranger proving his hand was empty.
The clasp still says what greetings alone cannot. We simply know more about what it does than about where it came from.
