Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 17 January 2020

Toys, Dolls, and the Material Culture of Childhood

Abstract and Keywords

When is a toy not a toy? The material culture of childhood in the past is notoriously difficult to locate but is most often identified in terms of playthings. This chapter looks at the problems of examining this body of material in terms of the life and death of the child. Dolls and miniatures, in particular, are considered because they appear to have a fascination for the modern reader and museum visitor. In funerary contexts these items have multiple possible meanings: Do they have a ritual significance? Are they placed in the grave by loving parents as representative of the lost child or the loss of the potential adult? Were they toys played with in life by the deceased child? None of these interpretations are mutually exclusive, but their range highlights the complexity of examining the material culture of childhood.

Keywords: Amulets, baby, babies, baby doll, balls, board games, dolls, games, gender roles, grave goods, material culture, play, toys


Children in the Roman world played with toys and games of many kinds: some were specially constructed with the child in mind; others were no doubt items to hand that inventive children used in various ways. However, behind these apparently rather self-evident statements lies a complexity of interpretation governed by the context in which “toys” are found and that raises questions such as when is a toy not a toy or when is it so much more than just a toy.1

In contemporary modern life the most obvious sign of the presence of a child in a household is the vast array of usually colorful toys and playthings and a panoply of specially designed equipment such as cots, high chairs, and reduced-size chairs and tables and whole rooms given over as special spaces for children in nurseries, bedrooms, and playrooms. The presence of a child is immediately visually and materially obvious, and the types of playthings and specialized furniture might even give an indication of the age and gender of that child and the socioeconomic status of the family. Children themselves, however, interact with this material world as a whole, not just the parts of it that are deemed created for them by their parents and other adults, and a child’s experience of the world is not curtailed by these “child-friendly” areas. Likewise, a child will turn anything into a plaything: a broom becomes a horse, and even if used to sweep in (p. 323) imitation of adult activity it will not perform in the same way as in an adult’s hands. A child’s everyday interaction with the material world is not controlled and mediated by adults at all times, and, even in the modern world, time outside adult space and control is hard to record. For antiquity this is only the start of the problem; while—as this volume demonstrates—there is a wealth of information about children, tracking how they engage with the material world is difficult.2

Toys and Playthings

Extensive research in the past three decades has given us new insights into the social use of Roman domestic space and the rhythms of daily life as seen in a variety of families in the empire, but we still find it very hard to locate children in the physical space of the household. This raises the matter of where, with what, and when children played. Small finds discovered in domestic contexts that have been identified as games and toys might have belonged to children and testify to their presence, but adults also played with balls, knucklebones, board games, dice, and the like. Moreover, the lack of defined space in the house also says something about the interrelationship between children and other inhabitants and may be read in many ways: it may imply that children are marginalized in the family and not considered full members of society until they reach an age approaching adulthood; it could suggest that they are fully integrated members of the household who range freely in the space and are allowed access to all areas; or it could simply reflect archeological or historical invisibility. The reality for the upper classes is possibly a mixture of all these, given the daily rituals that dictated the use of the domestic space of the home and the role of nonparent carers in the upbringing of young children.3 For less well-off families the situation would be different again: in the smaller dwellings or apartment blocks of larger towns it is hard to locate space that would fit with any modern sense of family life—for instance, rooms that might allow for privacy. Here the space must have been multipurpose and served as sleeping, working, eating, and general living quarters. Children here may have had less time to play and spent more time generally helping the household survive. In all these spaces, and on the streets and in the fields, children presumably played. They would have made use of things to hand, sticks, stones, broken pottery, and earth perhaps (Epictetus 3.13.18), in even the poorest homes. They may also have made use of household objects as playthings: for instance, small (p. 324) spindle whorls would fit neatly into little hands and can be rolled or spun, loom weights have potential as building bricks, and both make good noises when knocked together.

Numerous surviving artifacts generally labeled “toys” and playthings and a fair amount of literary and iconographic evidence provide apparent vignettes of children playing. But as with most things in the ancient world, the material culture and literary and visual evidence do not neatly dovetail to create a homogenous picture. How do we identify the material culture of childhood that was either chosen by adults or created by the child? How might it be different from the material culture of adulthood? Using such generic terms as childhood and adulthood is also problematic: there are many childhoods and many adulthoods. Looking for the material culture of childhoods in the past requires us to be very circumspect about what we claim belongs specifically to the realm of the child. This issue is further problematized by the fact that artifacts apparently securely associated with children (as opposed to childhood per se) are, in the main, found in identifiable burial contexts. Even here, however, we cannot be certain if grave goods belonged to the child or were chosen by the commemorators to represent the child or to serve a particular purpose in the afterlife, or for the funerary rite itself, or had some cultic association; still less can we tell if these artifacts were treasured by the child and marked his or her own view of their material world. This makes any grave goods open to a number of interpretations, which are not necessarily mutually exclusive. Furthermore, it is very difficult to sex juvenile skeletal remains (under the age of twelve), so in older excavation reports the grave goods have sometimes been used as an indicator of the sex of the deceased, with no other supporting evidence. This creates a circular and gendered argument that is not helpful. The dolls discussed in this chapter are associated with graves for which either the body or an identifying inscription survives, unless otherwise stated.4

Not all the items discussed here were found exclusively in burial contexts associated with children, but the probability of them being part of a child’s life is high enough for them to be included and to be relatively securely accepted as part of the child’s interaction with society and materiality. The different phases of childhood—infancy, toddler, middle childhood, and early adolescence—need also to be taken into account, and we need to be alert to the danger of reading modern assumptions about age stages into the material.5 As children grow, their demand and need for particular types of playthings evolves. Tiny babies may be amused by movement and noise but unable to maintain hold of an item to amuse themselves. As children grow in dexterity they can first hold, shake, and bang little things that might distract or please them with sound; they can push and hold balls and chase after them, even if on all fours. A very young child does not need much in terms of specially fabricated material to be amused. Once a child is mobile a new range (p. 325) of playthings is accessible. Push and pull-along items take on a new dimension once they can be moved across space by a child who can also begin to create an imaginary world for them to inhabit. It is hard to think of a plaything that does not have some educational aspect: some are more implicit in that a child will learn dexterity and calculations when at play (e.g., knucklebones), and others are more explicit (building bricks with letters on them or dice); at every stage of childhood such toys play different and varied roles in the life of the child. Dolls and miniature copies of daily life objects can open up a world of imaginative play, while at the same time they can inculcate social values and traditional gender roles and also perhaps subvert them in the hands and minds of imaginative children.6

Toys, Dolls, and the Material Culture of Childhood

Figure 16.1 A pipe-clay rattle in the shape of seated mother and child. Campania H. 16.8 cm ca. 100–200 CE. This could also be an example of sigillaria.

(Courtesy British Museum.)

Rattles of various descriptions are found throughout the Roman world. These come in a variety of sizes and shapes and are made from various materials, including silver, bronze, terracotta, and plaster. Some are simply spherical or oval, but others are like small busts, like the mother and child illustrated here (Figure 16.1) or are zoomorphic in form with painted or incised details in the shapes of pigs, dogs, chickens, cocks, and other birds as well as some hybrids, for example, the body of a bird with the addition of stag’s antlers.7 Some are shaped to fit neatly in a small hand, while others have a handle of some description. The noise seems to be made by putting tiny pebbles or clay balls inside the hollow body of the rattle, and, as today, they were used to amuse and distract babies (p. 326) and very young children (Mart. 14.54; Lucr. 5.228; Quint. Inst. 9.4.66). Similar in shape and size are whistles in the shape of tiny birds; one particularly delightful find of five terracotta chickens from Compiègne is now in the Musée des Antiquités Nationales de St-Germain-en-Laye.8 Like many toys, such items might continue to amuse older children but be played with in different ways. Some toys, however, are designed for particular age stages. Quintilian suggests that ivory letters be played with as the first stage of learning to read and write (Quint. Inst. 1.1.26).9

Toys, Dolls, and the Material Culture of Childhood

Figure 16.2 A terracotta group of young women playing knucklebones. Campania ca. 330–300 BCE.

(Courtesy British Museum.)

Sculptors use the motif of playing with nuts, knucklebones, and balls to encapsulate the idea of childhood (Figure 16.2). Several sets of knucklebones survive; some are literally the knucklebones of sheep or pigs, whereas others were made of far more precious material such as ivory (Mart. 14.14; Seneca De Constantia 12.2).10 Similarly, nuts would have been easy to find and play with, but terracotta models of nuts have also been found; a young woman from Vetralla was found buried with one carved from rock crystal.11 (p. 327) This last example is far more likely to belong to the category of ritual object (crepundia) than toy, but the fact that it is in the form of a plaything is worth noting. Literary and iconographic evidence suggests that both boys and girls enjoyed playing with these easily available objects. A sarcophagus dating to the third century CE and now in the Vatican Chiaramonti Museum (inv. no. 1304) has a relief on its front face showing young girls playing with nuts in one third of the space and boys playing and fighting in the rest of the space. The girls play beneath a canopy or curtain, perhaps suggesting they are inside, while the boys’ play is far less constrained and more vibrant.12 Martial, Horace, and Persius all describe playing and gambling with nuts (Mart. 14.19 [18]; Hor. Sat. 2.3.171–4; Pers. 1.10)—a habit that could last into adult life.

Children also played with small figures, animals, and wheeled horses, carts, and mounted soldiers. Tiny copies of birds and animals made in terracotta or plaster are common, as are little figures mounted on horses made with a hole in the legs through which a wooden spindle could be threaded and wheels mounted on each side, allowing the figure to be rolled across the floor (Figure 16.3).13 Little lead figures like the horse and rider shown here have also been found (Figure 16.4).

Toys, Dolls, and the Material Culture of Childhood

Figure 16.3 Painted wooden wheeled horse from Roman Egypt. H. 7.6 cm, L. 11 cm.

(Courtesy British Museum.)

Toys, Dolls, and the Material Culture of Childhood

Figure 16.4 Small lead figure of a horse and rider. H. 4.318 cm. Roman, 1st–3rd century CE.

(Courtesy British Museum.)

Toys, Dolls, and the Material Culture of Childhood

Figure 16.5 Small hollow balls made of linen and decorated with reeds. These have stone chips inside so could also have served as rattles and probably as ritual objects. Roman Egypt.

(Courtesy British Museum.)

Balls were also part of the ancient toybox. As might be imagined, they came in all sizes and weights: colored wool balls, linen, and reed balls (Figure 16.5); and leather skins stuffed with either feathers (paganica: Mart. 14.45) or air (follis, folliculis: Mart.12.82, 14.47) or that were small and very hard (trigonalis: Mart. 14.46). Playing with balls is recommended as good exercise for all ages (Galen De San. 8.1; Mart. 14.47). Balls rarely (p. 328) survive in the archeological record, but many representations of children of all ages show them holding or playing with balls. A relief in the Louvre (inv. no. 120) from the mid-second century CE shows three girls following each other, one holding a ball and another throwing one in the air. At the other end of the same panel, a group of four boys is playing a game that involves rolling the balls down a board. A similar scene is (p. 329) illustrated on a sarcophagus in the Museo Nazionale, Rome (inv. no. 67612), which shows three chubby erotes playing with what looks like a push-along wheel. Mirroring them on the other side of the central inscription are four little figures who play at rolling circular objects with holes in the middle (not unlike loom weights) down a board.14

In this category of more active play we might also include hoops and push/pull-along carts. Hoops are mentioned by Ovid (Ars Am. 3.381), Horace (as something Greek, Odes 3.24.57), Martial (14.168, who suggests his gift is a wheel), and Propertius (3.14, a Spartan reference), but they rarely occur in iconography and not at all, as far as I am aware, in the archeological record.15 Carts and pretend chariots are known from iconography showing children using the ancient equivalent of the baby-walker and riding in vehicles pulled by animals. On an early second century sarcophagus from the Museo Nazionale Romano collection (inv. no. 65199), a small, chubby, naked child is shown learning to walk by pushing a small cart. More popular are images such as the famous relief on the mid-second century biographical sarcophagus of Cornelius Statius in the Louvre (inv. no. MA659), which encapsulates a stage of childhood in an image of a little boy standing in a small pretend chariot pulled by a ram. (See Larsson Lovén in this volume for scenes of children’s play on sarcophagi).16

Spinning tops also seem to have been common playthings. There are small boxwood examples of these that were clearly made to be spun by hand, with a conical pivot on the underside and a slightly raised center on the top for gripping.17 Others seem to have been made for whipping, like the one Vergil describes flying back and forth under the whip in front of a group of amazed boys (Aen. 7.376–87).18


Dolls are probably the most well-documented “toys” of antiquity.19 They have a fascination for modern authors that other toys seem not to have and have been far more discussed in modern scholarship than any other plaything and almost any other grave good (p. 330) commonly found in child burials, with the exception of weapons.20 Dolls as tiny human (usually female) figures survive in great numbers from antiquity. In 1987 Michel Manson had recorded 493 and recent discoveries have probably upped this figure to just over 500.21 Manson defines a play doll as being any figurine that could be identified as not coming from an exclusively cultic or religious context, was of a size and weight that could be manipulated by a child, and whose structure and articulation would allow the illusion of real life in play.22

Toys, Dolls, and the Material Culture of Childhood

Figure 16.6 Rag doll from Roman Egypt, made of linen and stuffed with rags and papyrus.

(Courtesy British Museum.)

Toys, Dolls, and the Material Culture of Childhood

Figure 16.7 Ivory doll from Via Valeria, Tivoli. Dated to late 2nd century CE by Julia Domna hairstyle. (Drawing J. Willmott; reproduced with permission of the University of Sheffield.)

Most surviving Roman dolls are made of cloth, wood, terracotta, bone, or ivory (Figure 16.6).23 The most frequently illustrated and discussed are those that date from the mid-second to early fourth century, all found in female (and for the most part young female) funerary contexts. The burial environment, the workmanship, and accoutrements of these dolls suggest that they belonged to girls or young women of the upper classes.24 The dolls were all found without clothes, although some have jewelry attached. All reproduce the body type of young or mature women rather than babies. They share, to varying degrees, small but clearly shaped breasts (some with defined nipples), wide (p. 331) hips, slightly protruding bellies (some with clearly delineated navels) and buttocks, and carefully marked-out pubic triangles (Figure 16.7). Their arms are jointed at the shoulder and elbow and their legs at the hips and knees, thus allowing for a wide range of potential movement. Unfortunately, I have never seen one of these dolls outside of a museum case so I have no idea how loose or stiff the joints might have been in antiquity or whether the dolls could hold a pose or were more like puppets and needed to be held in position. On some of them the facial features and hairstyles are very skillfully and carefully carved. Traditionally the dolls have been dated, like female sculpture, by the imitation of their hairstyles to imperial models. Those that survive intact range from 15 to 24 centimeters in height. They have been found across the western empire. The fact that these are not baby or even child dolls means they have been viewed not only as preparing young girls for their future role as a wife25 but also as inculcating ideas of sexual identity rather than maternity.26 (p. 332)

When dealing with rag dolls that could more easily be imagined as babies, Janssen (1996) also took the view that the dolls could be seen as playing a part in preparation for the future role as mother.27 Some of the dolls have been found with small items such as miniature furniture or toilet implements that have been identified as doll accessories.28 Dolls’ clothes and a woolen ball for a doll, dating to the fourth century CE, were found at Behnasa in the late nineteenth century.29 As with clothing for adults, the dry conditions of Egypt have allowed the preservation of the doll and the clothes. Unfortunately the articulated dolls have not been found with clothes, but literary evidence suggests that they were dressed (Anth. Pal. 6.280). This is the single reference to dolls’ dresses, and the Palatine Anthology is difficult to date. The nudity of the dolls has not often been remarked upon, presumably because of the assumption that they were originally dressed. Janssen comments that “naked dolls are appealing to little girls as they could dress them themselves” (p. 239). However, the detail on the bodies of the articulated dolls, in not only shape but also the defined nipples, navels, and pubic triangles, deserves more attention. Not all dolls are female shaped, however. An enigmatic little torso of a Roman soldier was found in the grave of ten-year-old Claudia Victoria in Lyon. This has holes at the shoulders suggesting that it too could have been articulated.30 This find also reminds us not to use dolls to assume the gender of the deceased. The quality of some of the articulated dolls both in terms of material and craftsmanship suggests that they probably belonged to the wealthier classes, but it must be remembered that it is imagination that makes a doll more than just an inanimate object. Dolls can be created from almost any material: a wooden spoon with a face drawn on it or straw bound and tied to resemble a human figure.

In recent research Fanny Dolansky (2012) worked with a sample of eighteen well-preserved, articulated dolls, including cloth dolls, ranging from the early second to mid-fourth centuries CE. Until Dolansky’s article, the role such dolls played in inculcating social norms had been relatively unproblematized, but still dolls have rarely been considered as simply toys or even as items of special affection for the deceased. The fact that they form part of the grave goods of girls and young women has led to a series of readings: (1) that the dolls might mark out girls who died before fulfilling their potential as mothers;31 (2) that the dolls signify the unmarried status of these girls and are a symbol of the ritual wherein girls laid aside their dolls before their wedding day;32 or (3) that they have some hitherto unknown religious meaning.33 (p. 333)

Dolansky (2012) made a far more complex case for the role played by dolls in the inculcation of gender and status ideologies by focusing on their potential for movement, their adornment in the form of attached or painted jewelry, and their likeness to imperial models.34 She placed the dolls with carefully carved hair and additional jewelry into the Roman debate on cultus. The notion and practice of cultus is a locus of contention among Roman authors, but the material culture associated with Roman women presents a far more positive image than the literature suggests.35 Upper-class Roman women were required to cultivate a look that would make them attractive partners for their husbands to display. This involved the right clothes, having the correctly arranged hair, a properly made-up face, and adornment of just the right amount of jewelry.

Latin authors make it impossible for women to get the balance between excess and moderation right: the weak female mind, seduced by expensive jewelry and exotic textiles, is a literary commonplace. The meretricious nature of cosmetics and wigs was a fundamental trope in almost all genres of literature. Moralists, satirists, elegists, and historians alike all played with this idea to suit their own ends. On the other hand, the material culture of women’s lives is often displayed as made up entirely of the cosmetic pots, jewelry, hairpins, tweezers, spatulas, and other implements required to engage fully with the management of cultus. In fact, women were so proud of this self-fashioning that it became part of their self-representation in death, on tombstones, and in grave assemblages.36 Dolansky (2012) argued that young girls, especially as they reached marriageable age, had to learn to negotiate the tightrope of the problem of personal adornment.37 In the graves of at least two of the girls who were buried with their dolls, miniature toilet implements and accoutrements were also found. If these can be argued to form part of their toy box in life, then there is an unambiguous message about learning to use the tools of cultus, either on themselves or on their dolls.

Dolansky (2012) reprised current notions of how the articulation of the dolls might lead to certain types of play: imitating the life of the child or those of adults she observed being dressed and adorned; or learning decorum and body language. Refreshingly, Dolansky pushed the potential of articulation further by using ethnographic studies of modern girls interacting with their Barbie dolls as a springboard to present ways a child might be more active herself in choosing the roles her doll might play. Barbie dolls come with dozens of manufacturer-sanctioned personae that plenty of girls extend to fit their own environments or subvert to create Barbie types of their own making. In modern research and society Barbie has become a rather complex figure. She has been seen as a role model (and body shape) to despise and reject, while at the same time she is continually remade, often in role models that those who denigrate her would want girls to aspire to or that might sit outside or beyond the social expectations of the players and (p. 334) their families. Dolls can both constrain and set free the imagination. Dolansky posited that the articulation of the dolls might have allowed little Roman girls to create a play world that took them outside their safe family environment.38 This seems to be a justifiable hypothesis, that a creative young girl could imagine the doll into the exoticism of the streets and the theaters and perhaps, given that most of these dolls looked like empresses, to an imagined imperial palace. Indeed, since most of these dolls would have belonged to upper class owners, a little girl might use the world of her father and brothers and imagine her dolls in the forum, the law courts, and the senate house. The same child, as Dolansky said, could at other times play at domesticity and dressing up and use the doll to practice the arts of cultus.

Dolansky’s (2012) final point was that the dolls, by virtue of their hairstyles, could be associated with empresses. She made a case that there are those with a resemblance to Faustina the Elder and Younger and presented a thesis that these dolls in particular became by association part of the symbolism of the imperial ideologies promoted by the empresses and their husbands. These ideologies, especially through the establishment of alimenta programs (on which see McGinn in this volume), stressed ideals of fertility, prosperity, and generosity and the concomitant virtues of the good wife and mother. The dolls, by association, became complex symbols, and their owners, while recognizing that they could never attain the positions and potential for euergetism expressed by empresses, might yet consider the virtues of good matronae while at the same time might “play with power and question established gender roles.”39

A Tiny Problem: The Case of Miniatures

I have spent some time on Dolansky’s (2012) reading of dolls as she made the multiple functions of inanimate objects clear; this returns me to the original question: when is a toy not a toy? While dolls may have complex interpretive baggage, finds of small items or miniatures are particularly problematic when dealing with the material culture of children.40 Miniatures survive in a range of materials from clay, tin, amber, and rock crystal to silver and gold and replicate all manner of things from household furniture to tiny animals and fabulous creatures. Particular kinds of miniatures, such as stamped pottery items known as sigillaria, were given as presents at the Saturnalia (Aul. Gell. 2.3.5; Sen. Ep. 12.3.5); the rattle shown in Figure 16.1 may also be classified as one of these (p. 335) sigillaria. Small objects, and sometimes the material they were made from, also came to have a number of amuletic, religious, and cultic associations. Consequently, in the surviving material culture, it is often hard to decipher which, if any, association should be dominant. Small charms or amulets were frequently carried or worn on a string around the neck by all classes of people in the Roman world. These items, which often also made a noise by jangling together, were known as crepundia. In Plautus’s Rudens, a long-lost daughter is recognized by such crepundia: Palaestra’s box contained a miniature gold sword inscribed with the name of her father, a two-edged ax of gold with the name of her mother, a little silver knife, two little hands linked together, a little sow and a golden bulla,41 given to her on her birthday (Rudens 4.4. 1151–9). Small items are, of course, very attractive for small hands, and toys made especially for children are small for a reason. However, amulets are also small, so there is a conundrum here for the modern viewer: when are crepundia toys? This is an equally vexing issue whether the item is found in a secure burial context or a particularly religious or cultic one, as toys were also dedicated at sanctuaries and played a part in some rites.42 A baby with a string of charms around his or her neck would presumably play with them, regardless of their amuletic associations, and I am tempted to say a bulla would also distract a baby, especially one that was suckable!

Numerous small artifacts survive from across the Roman Empire, but even in child graves some objects retain an ambiguity and multiplicity of function. We cannot simply assume that miniatures are toys or, conversely, that they have solely cultic connotations.43 In the tomb of fifteen-year-old Julia Graphide from Brescello, a collection of tiny tin domestic objects (ranging between 2 and 4 centimeters high) was found; it consisted of a set of plates, drinking vessels, and tiny furniture. These have been interpreted variously as the playthings of the child when alive or as toys for dolls or a sign of parental grief and loss: a child’s favorite toys buried with her to keep her company in the afterlife or as crepundia.44 The tiny objects would certainly suit small hands and encourage imaginative play, which would also inculcate traditional gender values. More recently Stefanie Martin-Kilcher (2000) argued that they should be seen only as crepundia and suggested that they were of too poor a quality for play and more likely to be votives that a young girl would dedicate to the gods before her wedding; their presence in the grave is a marker of the “unattained wedding.”45 This view means that the tiny (p. 336) domestic objects serve to represent, like the dolls, a life not fully lived, a maturity never reached. Martin-Kilcher did, however, think that another miniature dinner service, this one of silver, found in the “Tomb of a Girl” was a “proper toy,” one that would have been dedicated to the gods on the night before the wedding but, unlike Julia Graphide’s collection, would have been played with in life. She argued that, because the tin set of Julia Graphide is too poorly made and tin has apotropaic powers, it is not a toy; the other silver set, however, is.46

The interpretation of miniature objects is complex; however, it is worth asking how a child might have engaged with these items, and the answer might depend on the age of the child. Small-sized objects have a fascination for child and adult alike, and it is easy to imagine them being provided, especially by wealthier parents, for their children as playthings. In her catalog Shumka labels some of her selection of small objects as “doll accessories,” and it is not hard to make this link although the provenance of many of her items is unknown. Shumka’s collection includes a ladle, goblet, and four little amphorae of varying sizes as well as tiny lamps, a silver footstool, and candelabra all dating to around the first century CE.47 Other examples could be added, for instance, a small dinner service of a dish and five drinking vessels of various shapes from Germania.48

In the thirteen burials investigated by Martin-Kilcher (2000), eight were identified as containing miniature implements; of these, six contained amulets. She categorizes both amulets and miniatures as crepundia. These collective crepundia consist of the domestic objects of Julia Graphide; a number of amber miniatures from a young woman’s grave from Vetralla including a tiny caped figure (cucullatus), a model of a tortoise, a little rock crystal skyphos and other vases, and a tiny scallop shaped container for cosmetics alongside other objects of mundus muliebris;49 toilet items from the tomb of Crepereia Tryphaena;50 and amber shell-shaped pots, amber vases, and amber ladles from a tomb on the Via Cassia.51 From Gabriella Bordenache Battaglia’s collection of grave goods from the Museo Nazionale at Rome we could add a similar collection of miniature items: a little shell-shaped amber pot, a little amber cup with a handle, little amber ladle with a long handle, all found in the tomb of the mummified eight-year-old girl from Grottarossa (second half of second century CE), who also had an ivory doll;52 thirteen appliqué figures made of thin, translucent ivory from Ariccia (late first century CE),53 (p. 337) and the small model of a boat made of cut glass and white ivory found with the cremated remains of Laetilia Gemella who died at age twelve (mid-first century CE).54 The common link among many of these miniatures is that they are made of amber or rock crystal, both materials that had apoptropaic powers in antiquity (Pliny NH 37.9, cited in Martin-Kilcher, p. 69). Close examination of the items to show wear might tell us if these were bought new to go in the grave or if, like those of Palaestra, they had been used in life. And if used in life, especially by small children, would these objects have played the role of toys, even of toys with special powers?


Artifacts defined as toys and playthings are part of the material culture of societies living in the Roman Empire. The toys I have dealt with here could conceivably have been used by a child. This does not exclude their use by adults, and many items that form part of a child’s experience might also be used in adult life; however, the balls and dice that adults played with might not be considered toys in the same way.55 For those who have read this volume thus far or who have some knowledge of Roman childhood, it requires a series of interpretive gymnastics to let the material evidence speak for itself. We tend to bring to the interpretation of material culture all that we know about a society and read it in ways that support those assumptions; an artifact per se generally takes a passive role in attempts to slot it into reconstructions of social behavior in the past. Artifacts defined as playthings or toys are no exception and, as we have seen, can often have multiple or ambiguous functions. But it seems apparent from all the evidence that Roman parents and society assumed that some children, at least, could and should have time to play.


Becker, M. J. (2006). “The Archaeology of Infancy and Childhood: Integrating and Expanding Research into the Past.” American Journal of Archaeology 110(4): 655–8.Find this resource:

    Bertelli, C., L. Malnati, and G. Montevecchi (2008). Otium: L’ arte di vivere nelle domus romane di età imperiale. Milan: Skira.Find this resource:

      Bordenache Battaglia, G. (1983). Corredi funerari di età imperiale e barbarica nel Museo nazionale Romano. Rome: Edizioni Quasar. (p. 338) Find this resource:

        Bradley, K. (1991). Discovering the Roman Family. Oxford: Oxford University Press.Find this resource:

          Calvi, M. (1974–75). “La barchetta vitrea del Museo di Treviso.” Aquileia Nostra 45–46: 479–86.Find this resource:

            Chioffi, L. (1998). Mummificazione e imbalsamazione a Roma ed in altri luoghi del mondo romano. Opuscula Epigraphica 8. Rome: Edizioni Quasar.Find this resource:

              Coulon, G. (1994, revised ed. 2004). L’ enfant en Gaule romaine. Collection des Hespérides: archéologie, histoire. Paris: Errance.Find this resource:

                ——. (1995). “L’enfance en gaule romaine.” Les Dossiers de l’Archéologie 12: 19–34.Find this resource:

                  ——. (2003). “Jeux et jouets.” In D. Gourevitch, A. Moirin, and N. Rouquet (eds.), Maternité et petit enfance dans l’antiquité romaine. Bourges: Editions de la ville de Bourges : 201–3.Find this resource:

                    Crawford, S., and G. Shepherd (2007) eds. Children, Childhood and Society. Oxford: Archeopress.Find this resource:

                      D’Ambra, E. (2007). Roman Women. Cambridge, UK: Cambridge University Press.Find this resource:

                        Dasen, V. (2003). “Les poupées.” In D. Gourevitch, A. Moirin, and N. Rouquet (eds.), Maternité et petit enfance dans l’antiquité romaine. Bourges: Editions de la ville de Bourges: 198–9.Find this resource:

                          Denzey, N. (2007). The Bone Gatherers: The Lost Worlds of Early Christian Women. Boston: Beacon Press.Find this resource:

                            Dolansky, F. (2008). “Togam virilem sumere: coming of age in the Roman world.” In J. Edmondson and A. Keith (eds.), Roman Dress and the Fabrics of Roman Culture. Toronto: University of Toronto Press: 47–70.Find this resource:

                              ——. (2012). “Playing with Gender: Girls, Dolls and Adult Ideals in the Roman World.” Classical Antiquity 31(2): 256–92.Find this resource:

                                Durand, A. (1991a). “Les jouets bruyants.” In Jouer dans l’antiquité. Catalogue d’exposition. Marseille: Réunion des Musées Nationaux : 50–3.Find this resource:

                                  ——. (1991b). “Les jeux de balle.” In Jouer dans l’antiquité: Catalogue d’exposition. Marseille: Réunion des Musées Nationaux: 92–9.Find this resource:

                                    ——. (1991c). “Moblier et dinette” and “Les animaux.” In Jouer dans l’antiquité. Catalogue d’exposition. Marseille: Réunion des Musées Nationaux: 64–9.Find this resource:

                                      Eilberg-Schwarz, H., and W. Doniger (1995) eds. Off with Her Head: The Denial of Women’s Identity in Myth, Religion and Culture. Berkeley: University of California Press.Find this resource:

                                        Elderkin, K. (1930). “Jointed Dolls in Antiquity.” American Journal of Archaeology 34: 455–79.Find this resource:

                                          Fittà, M. (1997). Giochi e giocattoli nell’antichità. Milan: Leonardo Arte.Find this resource:

                                            Going, C. J., M. Green, C. Duhig, and A. Taylor (1997). “A Roman Child Burial with Animal Figurines and Pottery, from Godmanchester, Cambridgeshire.” Britannia 28: 386–93.Find this resource:

                                              Gourevitch, D., A. Moirin, and N. Rouquet (2003) eds. Maternité et petit enfance dans l’antiquité romaine. Bourges: Editions de la ville de Bourges.Find this resource:

                                                Hales, S. (2003). The Roman House and Social Identity. Cambridge, UK: Cambridge University Press.Find this resource:

                                                  Harlow, M. (2012a). “Dressing to Please Themselves: Clothing Choices for Roman Women.” In M. Harlow (ed.), Dress and Identity. Oxford: Archaeopress: 37–48.Find this resource:

                                                    ——. (2012b) ed. Dress and Identity. Oxford: Archaeopress.Find this resource:

                                                      ——. (2012c). “Death and the Maiden.” In M. Carroll and J. P. Wild (eds.), Dressing the Dead in Classical Antiquity. Stroud: Amberley: 148–57.Find this resource:

                                                        Harlow, M., and R. Laurence (2002). Growing Up and Growing Old at Rome: A Life Course Approach. London: Routledge.Find this resource:

                                                          ——. (2010a). “Betrothal, Mid-Late Childhood and the Life Course.” In L. Larsson Lovén and A. Strömberg (eds.), Ancient Marriage in Myth and Reality. Newcastle upon Tyne: Cambridge Scholars: 56–77.Find this resource:

                                                            ——. (2010b) eds. A Cultural History of Childhood and the Family. Oxford: Berg. (p. 339) Find this resource:

                                                              Harlow, M., R. Laurence, and V. Vuolanto (2007) “Past, Present and Future in the Study of Roman Childhood.” In S. A. Crawford (ed.), Children, Childhood and Society. Oxford: Archeopress: 5–14.Find this resource:

                                                                Hersch, K. (2010). The Roman Wedding: Ritual and Meaning in Antiquity. Cambridge, UK: Cambridge University Press.Find this resource:

                                                                  Huskinson, J. (1996). Roman Children’s Sarcophagi. Oxford: Oxford University Press.Find this resource:

                                                                    Janssen, R. (1996). “Soft Toys from Egypt.” In D. Bailey (ed.), Archaeological Research in Roman Egypt. Ann Arbor, MI: Journal of Roman Archaeology: 231–9.Find this resource:

                                                                      Jouer dans l’antiquité (1991). Catalogue d’exposition. Marseille: Réunion des Musées Nationaux.Find this resource:

                                                                        Lanciani, R. (1892). Pagan and Christian Rome. London: Macmillan.Find this resource:

                                                                          Lavaniouk, O. (2007). “The Toys of Dionysos.” Harvard Studies in Classical Philology 103: 165–202.Find this resource:

                                                                            Manson, M. (1975). “Le droit de jouer pour les enfants grecs et romains.” Recueils de la société Jean Bodin 39: 117–50.Find this resource:

                                                                              ——. (1987). “Le bambole romane antiche.” Ricerca Folklorica 16: 15–26.Find this resource:

                                                                                ——. (1991). “Les poupées antiques.” In Jouer dans L’antiquité. Marseille: Musée de Marseille : 54–9.Find this resource:

                                                                                  ——. (1992). “Les Poupées Antiques.” Les Dossiers d’Archaeologie 168: 48–57.Find this resource:

                                                                                    Martin-Kilcher, S. (2000). “Mors Immatura in the Roman World—A Mirror of Society and Tradition.” In J. Pearce, M. Martin, and M. Struck (eds.), Burial, Society and Context in the Roman World. Oxford: Oxbow: 63–77.Find this resource:

                                                                                      May, R. (1991). “Les jeux d’osselets.” In Jouer dans l’antiquité. Catalogue d’exposition. Marseille: Réunion des Musées Nationaux : 100–5.Find this resource:

                                                                                        Nevett, L. (2010). Domestic Space in Classical Antiquity. Cambridge, UK: Cambridge University Press.Find this resource:

                                                                                          Parkin, T. (2010). “The Life Cycle.” In M. Harlow and R. Laurence (eds.), A Cultural History of Childhood and the Family. Oxford: Berg: 97–114.Find this resource:

                                                                                            Poplin, F. (1992). “Les jeux d’osselets antiques.” Les dossiers de l’archéologie 168: 46–7.Find this resource:

                                                                                              Rapinesi, I. A. (2004). “Il lusso a Roma.” In D. Candilio (ed.), Moda, costume e bellezza nella Roma antica. Rome: Electa: 33–40.Find this resource:

                                                                                                Rawson, B. (2003). Children and Childhood in Roman Italy. Oxford: Oxford University Press.Find this resource:

                                                                                                  Richlin, A. (1995). “Making Up a Woman: The Face of Roman Gender.” In H. Eilberg-Schwarz and W. Doniger (eds.), Off with Her Head: The Denial of Women’s Identity in Myth, Religion and Culture. Berkeley: University of California Press: 185–213.Find this resource:

                                                                                                    Ricotti, E. S. (1995). Giochi e giocattoli. Rome: Edizioni Quasar.Find this resource:

                                                                                                      Rinaldi, M. R. (1956). “Richerche sui giocattoli nell’antichità a proposito di un’iscrizione di Brescelo.” Epigraphica 18: 104–29.Find this resource:

                                                                                                        Rossi, F. (1993). “Deux poupées en ivoire d’époque romaine à Yverdon-les-Bains VD.” Archéologie Suisse 16(4): 152–7.Find this resource:

                                                                                                          Rouvier-Jeanlin, M. (1991). “Réflexions sur les jouets en terre cuite de la Gaule romaine.” In Jouer dans l’antiquité. Catalogue d’exposition. Marseille: Réunion des Musées Nationaux: 72–3.Find this resource:

                                                                                                            Scheuer, L., and S. Black (2000). Developmental Juvenile Osteology. London: Academic Press.Find this resource:

                                                                                                              Shepherd, G., and S. Crawford (2007). “Children, Childhood, and Society: An Introduction.” In S. Crawford and G. Shepherd (eds.), Children, Childhood and Society. Oxford: Archaeopress: 1–4.Find this resource:

                                                                                                                Shumka, L. J. (1993). Children and Toys in the Roman World: A Contribution to the History of the Roman Family. M.A. diss., University of Victoria. (p. 340) Find this resource:

                                                                                                                  ——. (1999) “A Bone Doll from the Infant Cemetery at Poggio Gramignano.” In D. Soren and N. Soren (eds.), A Roman Villa and a Late-Roman Infant Cemetery: Excavations at Poggio Gramignano, Lugnano in Teverina. Rome: L’Erma di Bretschneider: 615–8.Find this resource:

                                                                                                                    ——. (2008). “Designing Women: The Representation of Women’s Toiletries on Funerary Monuments in Roman Italy.” In J. Edmondson and A. Keith (eds.), The Fabrics of Roman Culture. Toronto: University of Toronto Press: 172–91.Find this resource:

                                                                                                                      Sofaer Derevenski, J. (1997). “Engendering Children, Engendering Archaeology.” In J. Moore and E. Scott (eds.), Invisible People and Processes: Writing Gender and Childhood into European Archaeology. London: Leicester University Press: 192–202.Find this resource:

                                                                                                                        ——. (2000) ed. Children and Material Culture. London: Routledge.Find this resource:

                                                                                                                          Swift, E. (2012). “The Archaeology of Adornment and the Toilet in Roman Britain and Gaul.” In M. Harlow (ed.), Dress and Identity. Oxford: Archaeopress: 47–58.Find this resource:

                                                                                                                            Trifilò, F. (2011). “Movement, Gaming and the Use of Space in the Forum.” In R. Laurence and D. Newsome (eds.), Rome, Pompeii and Ostia. Oxford: Oxford University Press: 312–31.Find this resource:

                                                                                                                              Wallace-Hadrill, A. (1988). “The Social Structure of the Roman House.” Papers of the British School at Rome 56: 43–97.Find this resource:

                                                                                                                                ——. (1994). Houses and Society at Pompeii and Herculaneum. Princeton, NJ: Princeton University Press.Find this resource:

                                                                                                                                  ——. (1996). “Engendering the Roman House.” In D. Kleiner and S. Matheson (eds.), I Claudia: Women in Ancient Rome. New Haven, CT: Yale University Art Gallery: 104–15.Find this resource:

                                                                                                                                    Wardle, D., and K. Wardle (2007). “The Child’s Cache at Assiros Toumba, Macedonia.” In S. Crawford and G. Shepherd (eds.), Children, Childhood, and Society. Oxford; Archaeopress: 29–44.Find this resource:


                                                                                                                                      (1) I have to thank both Leslie Shumka and Fanny Dolansky for their generosity in sharing their research with me. Leslie sent me a copy of her 1993 MA thesis “Children and Toys in the Roman World: A Contribution to the History of the Roman Family”; Fanny allowed me to read a draft copy of her article, “Playing with Gender: Girls, Dolls and Adult Ideals in the Roman World,” Classical Antiquity 31(2): 256–92. This chapter owes much to their research and ideas and is undoubtedly the better for it. Any errors remain my own.

                                                                                                                                      (2) For an excellent introduction to the problems of children and material culture see Sofaer Derevenski 2000; see also the chapter by Langdon in this volume.

                                                                                                                                      (3) There is now a large bibliography of the social use of domestic space. See, for example, Wallace- Hadrill 1988, 1994, 1996; Hales 2003; Nevett 2010. Shumka 1993: 58–62 directly addressed the problem of recreational space. On nonparent carers see Bradley 1991; Rawson 2003: 130–3; see also the chapter by McWilliam in this volume.

                                                                                                                                      (4) For discussions of the problem of artifacts belonging to children and different ranges of interpretation see Ricotti 1995: 20; Sofaer Derevenski 2000: 6–12; and other papers in Sofaer Derevenski. For the problems with aging skeletons, see Scheuer and Black 2000; chapter 2 deals specifically with skeletal development and aging).

                                                                                                                                      (5) On the ancient life course see Harlow and Laurence 2002; Parkin 2010.

                                                                                                                                      (6) For a selection of images of toys in the Roman world see the catalogs of Jouer dans l’Antiquité and Bertelli et al. 2008.

                                                                                                                                      (7) Shumka 1993: 105–9, cat. nos. 71–83; Coulon 1994/2004: 76–9; Ricotti 1995: 18–24 (where crepundia of all types are included in the section on rattles).

                                                                                                                                      (8) Coulon 1994/2004: 80; Ricotta 1995: 25; Durand 1991a: 50–3 (with images).

                                                                                                                                      (9) Jerome uses Quintilian’s advice in his late fourth-century letter (128.4) on how to raise a little girl to a life of asceticism.

                                                                                                                                      (10) Poplin 1992, May 1991: 100–5, 108–9 (with images).

                                                                                                                                      (11) Bordenache Battaglia 1983: 71; Shumka 1993: 98, cat. no. 51.

                                                                                                                                      (12) Huskinson 1996: 17, cat. no. 1.37.

                                                                                                                                      (13) Coulon 1994/2004: 108–9 (with images); Ricotti 1995: 25–9 (with images); Rouvier-Jeanline 1991: 64–73 (with images).

                                                                                                                                      (14) For more on balls see Shumka 1993: 120–3, cat. nos. 1–6. For images see Ricotti 1995: 29; Durand 1991b: 92–100 (with images).

                                                                                                                                      (15) Shumka 1993: 123–5, cat. nos. 54, 55 has only the four literary references and two iconographic examples, one of which, the mosaic from Istanbul, is unclear. For illustrations of hoops in Greek vase paintings see Ricotti 1995: 33 and the chapter by Oakley in this volume.

                                                                                                                                      (16) See Huskinson 1996: 14–5, pls. 1.3, 2.1, her cat. nos. 1.23, 1.29.

                                                                                                                                      (18) The progress of the top is a simile for the furor of Amata; these lines are also rather overused as one of the two examples of children actually being placed within the confines (here the atrium) of a house (the other is Lucr. 4.400).

                                                                                                                                      (20) Sofaer Derevenski 2000: 6–10 and her references for the problems of interpreting weapon deposits.

                                                                                                                                      (21) Manson 1987: 19; Dolanksy 2012: 257, n. 2 lists a number of new finds.

                                                                                                                                      (22) Manson 1987: 15.

                                                                                                                                      (23) On rag dolls from Roman Egypt dating from first to fourth centuries CE see Janssen 1996.

                                                                                                                                      (24) There is a potential problem with the evidence here. As noted in note 4, it is impossible to sex a skeleton under the age of about twelve. In the past dolls, especially female figurines, have been used to assume the grave is one of a female child when this might not be the case. Most of those dealt with here come from a context where the skeleton has been sexed, or an inscription identifies the deceased. The sample in Dolansky 2012 includes a wider range, and her article deals essentially with dolls as girls’ toys.

                                                                                                                                      (26) Shumka 1993: 168–9.

                                                                                                                                      (27) Janssen 1993: 239.

                                                                                                                                      (28) Shumka 1993: 91–4, cat. nos. 31–40. For a tiny silver table service from the “Tomb of a Girl” near Rome, see Martin-Kilcher 2000: 69.

                                                                                                                                      (29) Janssen 1993: 237–9. She also shows an illustration of doll’s clothing from Qasr Ibrim dating from the Ottoman period.

                                                                                                                                      (30) Shumka 1993: 131, cat. no. 18; Claudia Victoria: CIL 13.2108.

                                                                                                                                      (32) Martin-Kilcher 2000; Harlow 2012c: 154–6. On prenuptial rituals and dedication of dolls and toys see Hersch 2010: 66–8; Harlow and Laurence 2002: 67–72; Dolanksy 2008: 47–70.

                                                                                                                                      (34) Dolansky 2012: 259.

                                                                                                                                      (37) Dolansky 2012: 274–5.

                                                                                                                                      (38) Ibid.: 278–81 (with bibliography).

                                                                                                                                      (39) Ibid.: 288. I find this point harder to process: I can understand the association with empresses and conscious (and unconscious) links with imperially presented virtues; however, while the point is undoubtedly well argued, I am not yet convinced that this would be the impetus behind any questioning of gender roles.

                                                                                                                                      (40) The problem of miniatures is also stressed by Sofaer Derevenski 2000: 7 (with bibliography).

                                                                                                                                      (41) A bulla is an amulet given to freeborn children; this is a rare example of a girl possessing one. On bullae, see Larsson Lovén in this volume.

                                                                                                                                      (43) For differing interpretations of miniature objects see Wardle and Wardle 2007 (Aegean) and Green in Going et al. 1997 (Roman Britain).

                                                                                                                                      (44) Rinaldi 1956: 105–14. See Durand 1991c: 64–6 for images.

                                                                                                                                      (45) Martin-Kilcher 2000: 69. This hypothesis fits with Martin-Kilcher’s overall argument on the grave goods belonging to a group of young girls whose mode of commemoration suggests that they had died before being married. I think Martin-Kilcher overestimates the literary evidence on premarital rituals for girls, which is scanty and antiquarian; see Harlow and Laurence 2002: 67–72; Dolanksy 2008: 47–70.

                                                                                                                                      (46) Martin-Kilcher 2000: 69. In editing this chapter, Judith Evans Grubbs suggested that it makes more sense to let a child play with a set made out of cheaper material rather than the more expensive silver one. Given that we know that Julia Graphide was the foster child of a freed couple, it may also be that the lead set was as good as they could afford.

                                                                                                                                      (47) Shumka 1993: 91–4, cat. nos. 31–40.

                                                                                                                                      (48) Römisches-Germanisches Museum de Cologne, in Coulon 1994/2004: 101.

                                                                                                                                      (50) For Crepereia Tryphaena see Lanciani 1892; Chioffi 1998: 81–4; Martin-Kilcher 2000: 64, 69; Denzey 2007: 2–9; Harlow 2012c, and for images see Bertelli et al. 2008.

                                                                                                                                      (51) Bordenache Battaglia 1983: 118–21, 115–7 on the ivory doll also found in this tomb.

                                                                                                                                      (52) Ibid.: 118–23; for the mummification and death mask of this little girl see Chioffi 1998; Dasen 2010: 138–42.

                                                                                                                                      (53) Bordenache Battaglia: 35–38.

                                                                                                                                      (54) Ibid.: 29. There is no external hint as to why this artifact was chosen to go in with the ashes. It has been variously interpreted as something that the child treasured, a reflection of her favorite activities, and something that would help her deal with the dangerous crossing to the underworld (Calvi 1974–75 cited in Bordenache Battaglia 1983: 29), or as having a symbolic meaning, as an allusion to the last journey to the underworld, particularly as traditional Roman gender roles would make the association of Laetilia Gemella with marine activities unlikely.

                                                                                                                                      (55) For gaming in the forum see Trifilò 2011.