Over the two decades that archaeologist Gus Van Beek excavated Tell Jemmeh, an Assyrian settlement inhabited from around 3,800 to 2,200 years ago, he recovered so many objects, it took the Smithsonian 40 years to catalogue them all. There were coins. Scarabs. Amulets. And an amount of pottery so vast, some of it later would have to be discarded.
But for Van Beek, the site – in what is now modern-day south-west Israel – yielded a discovery that was "among the more enigmatic objects recovered": 17 small, rounded discs – some made of chalk, some of stone, but most upcycled from potsherds – with two deliberate holes in the centre.
Van Beek wasn't the first archaeologist to discover objects like these. Nor was he the last. They've been found at sites across Japan, Egypt, India, and the Americas, among others. Three were found in New York City at the site of a British army camp during the American War of Independence, one fashioned from a coin. Others found elsewhere date back 4,000 years.
Some archaeologists believed they were buttons. For others, they were loom weights, perforated pottery, or simply "miscellaneous objects". But they reminded Van Beek of something else. "I remembered playing in my early years with a similar object," he remarked. Thread string through the holes, then stretch and relax the string, and the discs spin. He called the objects what they were known as when he was a child – "buzzes" – and went so far as to experiment with creating one himself.
While a handful of earlier scholars also suspected these were toys, others were sceptical. Relying on one's own childhood memories, and projecting our own, modern experience onto a distant society, seemed unacademic at best.
The mystery of the ancient buzzes is just one of many archaeological puzzles related to children's play, and one that highlights many of the pitfalls of studying it. We know that children played. We know they often played with objects. But other questions, such as which objects, and in what way, have remained stubbornly difficult to pinpoint. So difficult, in fact, that they've inspired archaeology's version of a "dad joke": an archaeologist finds a small object. "Hey, what's this?" he asks. "I dunno," says another. "Must be a toy… or a religious object!"
But understanding how children played is important, not least of all because it gets to the heart of a decades-old debate: what childhood really meant to past generations – if it meant anything at all. In the 1960s, the amateur French historian Philippe Ariès promulgated the theory that for most of history, with child mortality too high for parents to invest much in the way of sentiment or resources in their offspring, children were treated as mini-adults. (Read more about whether we really live longer than our ancestors). That extended to play. After the age of infancy, Ariès wrote, children no longer had toys and games specifically made for them. Instead, they played with the same objects as adults.
A relief dating to the 9th Century BC in Turkey depicts two men playing with spinning tops (Credit: Getty Images)
While academics have dismantled much of Ariès's theory, many of his beliefs persist. But archaeologists, particularly those who study childhood, are pushing back. And one of the biggest tenets of their argument has to do with their findings around children's play.
"It has been said too often that there was no feeling for childhood – that childhood was a time in life that had to be passed as quickly as possible to become an adult, and then you fully 'exist'," says Véronique Dasen, professor in classical archaeology and art history at Switzerland's University of Fribourg, leader of the EU-supported Locus Ludi project on Greco-Roman games and co-editor of a forthcoming volume on ancient play which publishes in September 2022.
"But that's not true. There is something special in children, and this special value is revealed by their fondness for play. And adults recognised that."
One problem is that, historically, childhood has been ignored by academia. "The child's world has been left out of archaeological research," archaeologist Grete Lillehammer wrote in her seminal 1989 work A Child is Born: The Child's World in an Archaeological Perspective.
"Few archaeologists have looked into the subject or given it attention, less ever thought of it as the main field of interest."
But that doesn't mean children weren't a valued part of the community – or that there weren't specific activities and objects primarily aimed at children. We even have etymological evidence: the ancient Greek word for "child" means "someone who plays". And some philosophers describe childhood as a specific life stage that centres on playing.
Small Ancient Greek jugs like this one from 430 BC, which are called choes, often show children playing (Credit: Alamy)
"Both Plato and Aristotle [talk] about the importance of play, about how it's good for children's development," says Maria Sommer, co-author of the book Care, Socialisation and Play in Ancient Attica. "They actually write to the parents, 'You need to let your children play'. And very interestingly, you didn't start at school in Ancient Greece before the age of seven. Until then, you have free play." (Read more about the benefits of starting education later).
But determining exactly how children played 2,000, 5,000 or 25,000 years ago – and with what – requires both intrepid research and some calculated guesswork.
For one, most playthings likely were made from natural materials like wood or straw, meaning they are unlikely to have survived: think dolls made from reeds, or games using knucklebones. But even in the face of more durable archaeological evidence, challenges remain.
Take one of the most important clues that archaeologists use to determine what an object is and how it's used: context. If a cup is found in a part of a home where there are also plates and spoons, they might hypothesise that it was used for serving or consuming drink. But if the same cup is found in a tomb alongside jewellery and amulets, it might have been made for decorative or ritual purposes.
Was this miniature horse on wheels, found in a 2,100-year-old grave, a toy, a funerary offering to the gods, or both? (Credit: Kerameikos Archaeological Museum in Athens)
With toys, context can be even more slippery. Children play everywhere, not just in predefined areas. (Although there may have been the ancient equivalent of child-centred playrooms, too). Just because an item is excavated in a context we associate with adults doesn't mean it wasn't played with. Some toys may have been adult items that were also used by children. Think of giving a toddler pots and pans to bang. If an archaeologist found those items 2,000 years from now, they might identify them as cooking tools, not as items a two-year-old spent countless hours happily banging.
On the other hand, even when it is excavated in a context associated with children, like a child's tomb, that doesn't mean it always was played with: the object could have been ceremonial or religious.
Complicating matters further is that past cultures differed a great deal from our own – so much so that even our question of "was this a toy, or a sacred object?" might be virtually meaningless.
Various female figurines have been found all over the world, including these, dating back 1,500 years, in Israel – but were they meant for children? (Credit: Getty Images)
Take dolls. Like buzzes, miniature female figurines have been discovered around the world. Some ancient writers also seem to describe girls playing toys that could correspond to our modern-day dolls. Plutarch, remembering his daughter who died aged two, says she would ask her nurse to give food to her "objects and toys" and would invite the toys to her table.
One particular doll type has been found in Ancient Greek and Roman sites, either in religious sanctuaries or buried in girls' tombs. With articulated limbs and elaborate details, including on-trend hairstyles and gendered adult characteristics like breasts (no baby doll from antiquity ever has been found), most surviving examples were made of terracotta in Greece, and bone or ivory (with one astounding example made of amber) in Rome. In Greece, t†erracotta dolls were so popular that they were mass produced with moulds. In Rome, the dolls were produced by specialised manufacturing centres for bone and ivory objects.
That doesn't mean, however, that these were the same dolls Plutarch's daughter would have played with, says Dasen, who is planning an exhibition on dolls at Switzerland's Yverdon Museum in 2024.
"We take for granted that it looks the same as a Barbie, so it is a Barbie. But no." Most of the Greek dolls that have been discovered are made of terracotta, too delicate to withstand the rough and tumble of much play. And many of the manufacturing moulds have been found at religious sanctuaries – which points to more of a sacred function.
This ivory Ancient Roman doll with articulated arms and legs was found in a 4th-Century child's grave (Credit: Archaeological Museum of Albacete Spain)
Instead, most researchers today agree that these types of dolls were used for specific ceremonial functions: dedicated to the sanctuaries of deities protecting girls and women, like Artemis, Demeter and Core, as part of the rite of passage before marriage or during the wedding ceremony, for example. In Ancient Greek, Dasen points out, korê means both "doll" and "unmarried girl".
Then again, the same doll might have had dual uses. "We have the tendency to divide sacral and normal. But they didn't back then," Sommer says. "It was all integrated. There was no segregation between those two worlds." Even in contemporary times, some cultures have similar overlap: anthropologists have noted that children in the Andes often make, and play with, miniature houses, which are later offered to the gods in the shrines.
Piecing play together
So how do archaeologists and historians have any real idea what children played with?
In some cases, there is a literary record. As one woman named Diogenis wrote to her brother 1,700 years ago in Oxyrhynchus, Egypt, "Many greetings to little Theon. Eight toys have been brought for him by the woman you told me to greet and these I have sent you." There are writings about Agesilaus II, the king of Sparta 2,400 years ago, who liked to ride a horse made of sticks with his son, and Octavian Augustus, Rome's first emperor, who played marbles with children.
Then there is iconographic evidence: the illustrations shown on vases, tombstones and reliefs, for example. Sommer shows me one, an Ancient Greek stelae depicting a child sitting and playing with a ball. "So that is not even up for discussion, is it, that this ball is actually something they play with."
One Ancient Greek stelae shows a child sitting and playing with a ball (Credit: Maria Sommer/The National Museum of Athens/Hellenic Ministry of Education and Religious Affairs)
She pulls up a different image. "Look at this boy – he's playing with a rattle. And we find identical rattles in the archaeological record." Ancient Greek rattles were so popular, they even had a manufacturing centre on the island of Cyprus.
This underscore another clue, one relied on by Van Beek in his interpretation of buzzes: what children play with today.
One sociocultural anthropologist, for example, found dozens of parallels between toys of children in contemporary Africa and in Greek and Roman antiquity. If we assume that children today are like children 3,000 years ago, the parallels might shed light on some archaeological mysteries, like the terracotta animal miniatures found in children's graves in Ancient Greece and Rome, often interpreted to be symbolic. In North Africa, children make their own miniature toy animals out of clay and play make-believe games with them.
But we must be careful not to overuse this approach, cautions Dasen. In one study, she references an Ancient Greek object that looks for all the world like a modern-day yo-yo. It appears on vases, being dangled in the air by children. Identical objects have been found in archaeological excavations. But those that have been found are made of fragile terracotta, and they're often decorated with motifs of seduction. Instead of a yo-yo, it could be an iynx: a disc that was twirled to try to attract luck in love.
An Ancient Greek vase depicts what looks like today's yo-yo – but probably depicts something else (Credit: Staatliche Museen zu Berlin, Antikensammlung/Johannes Laurentius)
Even rattles have been contested. While they've been found around the world – Siberia, where the 4,000-year-old clay toy is shaped like a bear cub's head; Turkey, where the black and white decoration is similar to the high-contrast colour schemes we use for infants today; ancient Greece, where even bronze rattles have been found in tombs of particularly wealthy infants – some of them have been too big, and made of material too fragile, to have been convincingly used by small children.
"But look at all these broken rattle pieces that no one's paid attention to. They're in houses, they're in streets, they're in places you would find children," says Kristine Garroway, an associate professor who focuses on children in ancient Israel and Mesopotamia at Hebrew Union College in Los Angeles, California.
"So maybe there's a broader context that has been overlooked, simply because children have been overlooked. And maybe it's the grown-ups or an older child that are shaking a rattle to keep the child quiet."
Objects that look, and sound, like rattles, such as this one from Greece 2,500 years ago, date back millennia (Credit: Maria Sommer/Benaki Toys Department, Phaliro, Athens
In the past, the only way to hypothesise if an object was created by a child was to gauge how "crudely" it was made, but today we have more scientific methods. While scholars often have interpreted miniature vessels to be votives, for example, at Israel's Tel Nagila site, which dates back 3,500 years, researchers used fingerprint analyses to determine that children made a number of them. If play was (and is) inherently educational – a "means of developing life-skills", as Garroway puts it, and a way to foster "a child's development as he moves into the adult world" – then these vessels were the result of play. (In her discussion of miniature axes, pots and arrowheads that have been found in Danish Bronze and Iron Age sites, Lillehammer similarly concluded that the items were likely both for play and education: small arrowheads found at a Mesolithic child burial in Skateholm, Sweden, for example, could have been used by children for training. More recent research, including of child-sized 1,700-year-old spear-throwing tools found in modern-day Oregon, has reached similar conclusions).
Another way to find out if a child could have made an object is to set up an experiment. Thirty years after Van Beek's own attempt to make a buzz, Garroway had a different take. Perhaps buzzes not only were played with by children, but made by children – a process that was both play, and a way for children to learn important skills around craft production. To test her hypothesis, she recruited 22 children to break pottery, then try to make their own buzzes. The standout, she says, was one child who took the sherd of a flowerpot which already had a drain-hole, jammed a pencil in the hole, and made a spinning top. "He understood the assignment in a different way," Garroway says, laughing. (Indeed, some archaeologists believe that one use of discs that have been found with a single hole, versus the double-holed buzzes, could have been spinning tops).
Although experts are continuing to chip away at the mystery of how past children played, many questions remain. There are puzzles that we may never solve – which may make archaeology's "dad joke" relevant for some years to come.
Still, Garroway points out, this isn't only true of child archaeology. "As much as we don't know for sure with children, we don't know for sure with adults about how they were using things," she says wryly. "We make educated guesses, a lot."
* Amanda Ruggeri is a senior journalist for BBC Future. You can find her at @amanda_ruggeri on Twitter.
Join one million Future fans by liking us on Facebook, or follow us on Twitter or Instagram.
If you liked this story, sign up for the weekly bbc.com features newsletter, called "The Essential List" – a handpicked selection of stories from BBC Future, Culture, Worklife, Travel and Reel delivered to your inbox every Friday.