Context

Terracotta figurines depicting interactions between adults and subadults represent a relatively recurring category within the material culture of the ancient Mediterranean [1, 2, 3, 4, 5]. These objects typically portray an adult figure holding, carrying, or otherwise interacting with a subadult and have often been discussed in relation to themes such as motherhood, caregiving, and childhood protection. Recent approaches inspired by the framework of Lived Ancient Religion (LAR) [6, 7] have encouraged scholars to examine religious practices through the material and relational experiences of individuals and communities. Within this perspective, figurines depicting adults with subadults can be approached as material expressions of caregiving relationships and social interactions, rather than simple representations of biological maternity. Such an approach resonates with the concept of othermothering [8], which highlights the collective and relational nature of childcare. This concept involves not only mothers but also other caregivers, whether belonging to the same family or not. However, despite the frequent appearance of terracotta figurines in archaeological publications and museum collections, evidence for the iconographic group of interacting adults and subadults remains widely dispersed across excavation reports, regional corpora, or individual object studies. This fragmentation has made it difficult to examine these figurines through large-scale or comparative approaches.

The project M/OTHERING was developed to address the dispersion of such figurines [9]. A primary aim of the project is that of creating a structured digital corpus dedicated to terracotta figurines representing adults with subadults. The first part of the project focused on the Italic peninsula and resulted in the compilation of a database of 2,725 figurines dated between the late seventh century BCE and 27 BCE (beginning of the Augustan period.) [1, 2, 3]. That dataset, composed primarily of votive material from sanctuary contexts, provided the first systematic collection of this iconographic category. Building on this initial corpus, the project was extended to the Greek world in order to investigate the broader distribution and uses of these figurines across the ancient Mediterranean. Figurines depicting adults with subadults are widely attested within Greek coroplastic production and are distributed across a broad geographical area including mainland Greece and the Ionian and Aegean islands. In contrast to the Italic material, however, these objects are more rarely associated with votive deposits. Greek examples occur in a wider range of archaeological contexts, including sanctuaries, funerary assemblages, domestic deposits, and secondary contexts documented in museum collections or antiquarian records. This contextual diversity makes the Greek material particularly relevant for examining the different settings in which representations of adult–subadult relationships circulated and were deposited. Moreover, the Greek corpus shows a comparatively broader constellation of caregiving figures, including nurses, pedagogues, and other attendants [10]. Yet, existing studies have generally approached this type of evidence through regional corpora, site publications, or thematic discussions. These are listed in the bibliography resource of the M/OTHERING_Gr database. No comprehensive dataset has previously brought together terracotta figurines depicting adults with subadults from the Greek world [4].

The dataset presented in this article represents the second phase of the M/OTHERING project. By systematically collecting and standardising published evidence for terracotta figurines representing adults with subadults, the database provides a consolidated digital resource that allows the material to be explored at a larger spatiotemporal scale. The dataset includes 730 figurines from the Greek world, although in several cases the number of preserved objects cannot be determined with certainty. The dataset is intended to facilitate future comparative research between the Italic and Greek corpora and to support broader investigations into the archaeological contexts and relational representations documented by these figurines.

Spatial coverage

The spatial coverage of M/OTHERING_Gr includes the historical regions of Aetolia, Attica, Boeotia, Locris, Phocis, Thessaly, Macedonia, and the Peloponnese, as well as the islands of Astypalea, Crete, Delos, Euboea, Kalymnos, Kos, Kythnos, Lesbos, Melos, Naxos, Nisyros, Rhodes, Samos, Samothrace, and Thasos. Additional evidence comes from Korkyra, Lefkada, and Illyria. Regional labels follow historical-geographical conventions commonly used in classical archaeology. Overall, these historical regions encompass modern Greece, although the site of Epidamnos (Illyria) belongs today to Albania. Figure 1 shows the distribution of the sites included in the database.

Temporal coverage

Data collected in M/OTHERING_Gr cover a period spanning from approximately 700 BCE to 1 BCE, with a single entry extending to 50 CE. This object has been included because its iconography remains consistent with the Hellenistic tradition. This temporal range corresponds broadly to the Archaic (700–480 BCE), Classical (480–323 BCE), Hellenistic (323–31 BCE), early Roman (31 BCE–50 CE) periods.

Steps

M/OTHERING_Gr compiles a comprehensive array of iconographic, typological, historical, archaeological, chronological, and georeferenced data on published terracotta figurines depicting adults with subadults from ancient Greece. This database follows the compilation of M/OTHERING, focusing on pre-Roman and Roman Republican Italy [3], and adopts a largely consistent data structure and metadata schema. The present dataset extends this research to the ancient Greek city-states and compiles published terracotta figurines depicting adults with subadults from historic Greece and the Aegean islands. The information was gathered through extensive desk-based research that identified a wide range of bibliographic sources relating to childhood, caregiving relationships, and coroplastic production in the Greek world. These sources include scientific publications in the form of articles, book chapters, excavation reports, and monographs. In addition, information was verified through research visits to museums in Greece and other European or North American institutions that curate these figurines, whenever possible.

Since the relevant scholarship has been published in several languages, particularly Italian, English, French, German, and, above all, modern Greek, the database provides an important tool for bringing together dispersed information and facilitating broader access to this material. The systematic compilation of these data also contribute to addressing significant gaps in the study of figurines representing relationships between adults and subadults in the ancient Mediterranean.

Sampling strategy

Bibliographic sources were identified based on prior knowledge and through web search engines such as Google Scholar and library catalogues. Additional references were traced through the bibliographies of consulted works. The research also made extensive use of digital cultural heritage repositories and museum databases, including the Greek cultural portal (https://www.searchculture.gr/aggregator/portal), which provides access to digitalised collections and catalogues of Greek museums and cultural institutions.

Alongside published literature, particular attention was devoted to identifying relevant material preserved in museum collections. In several cases, additional information was obtained through contacts with foreign archaeological schools based in Athens, particularly the British School at Athens and the École française d’Athènes. Direct communication with museums, curators, and collection managers also made it possible to identify figurines that remain unpublished or only partially documented in the scholarly literature. As a result, the dataset incorporates not only previously published material but also evidence that had remained largely inaccessible or insufficiently studied, significantly expanding the available documentation for the study of figurines depicting adults with subadults in the Greek world.

The representativeness of the dataset is inevitably impacted by the uneven state of publication, documentation, and accessibility of coroplastic material in museum collections and excavation archives. The database has been conceived as an expandable resource, open to future additions as new material is identified, published, or made accessible.

Quality Control

To ensure accurate reporting of data, systematic cross-reference and verification of bibliographic sources was conducted. Original publications are identified both within the main spreadsheet (Author Year, page) and in a separate file containing the full references, allowing users to verify information directly from primary sources.

Whenever possible, bibliographic data were cross-checked through the direct examination of artefacts in museum collections or through communication with museum curators, collection managers, and specialists. This process helped clarify issues related to provenance, chronology, and iconographic interpretation.

Additional quality control measures included the verification of geographical information and coordinates, as well as the standardisation of terminology used to describe the relationships between adults and subadults and their iconographic features. These procedures were adopted in order to ensure the internal consistency of the dataset and facilitate its reuse for comparative and quantitative analyses.

Constraints

Limitations that affect the dataset are typically reported in archaeological databases. For ancient Greece, a significant portion of the available literature, particularly excavation reports, dates from the nineteenth and early twentieth centuries. During this period contextual documentation was often incomplete or entirely lacking. Many excavations were conducted without standardised archaeological methods, and coroplastic materials were frequently considered minor artefacts and therefore insufficiently recorded. As a result, numerous figurines may have been omitted from excavation reports, dispersed, or entered the antiquities market. This issue is even more dramatic for funerary contexts. Illegal excavations, tomb robbing, and the modern history of the antiquities trade have resulted in a large number of retrieved terracotta figurines lacking confident archaeological provenance. Major Greek sanctuaries are relatively better documented through systematic excavations and detailed publications. However, figurines depicting adults with subadults usually represent only a very small fraction of the overall coroplastic assemblages from these sites. This limits meaningful analyses of sanctuary contexts, associated cults, or the broader composition of votive assemblages.

Several fields in the dataset contain missing values, particularly for variables such as geographical coordinates, precise findspots, chronology, and iconographic details. These absences reflect the limitations of the available archaeological and museum documentation discussed above, especially the frequent lack of secure provenance for many objects.

Missing iconographic data may result from the fragmentary condition of the objects or from surface erosion that obscures diagnostic features. Other types of information, such as chronology or the identity of the buried individual, are often absent because they are not reported in the existing literature. Empty fields in the dataset therefore indicate that the information is either not available in the source material, cannot be determined with sufficient reliability, or is not applicable to the specific entry.

Additional limitations concern the granularity of specific information that could not be systematically included. For example, traces of pigmentation were occasionally noted in some figurines. Yet, colour is not recorded as a formal field in the database, since most published illustrations are limited to black and white photographs or drawings. Despite these limitations, the data recorded in the database follow as closely as possible the information provided in original publications and museum records, even when these may contain inaccuracies or imprecisions. This approach was adopted in order to avoid introducing additional interpretative bias during the compilation of the dataset.

Object name

The dataset consists of a single table available as a spreadsheet file (M_OTHERING_Gr.13.07.2026.xlsx) and in CSV format (M_OTHERING_Gr.13.07.2026.csv), a file compiling a list of reference sources from which data were retrieved available in .docx and .rtf (M_OTHERING_Gr-bibliography.docx; M_OTHERING_Gr-bibliography.rtf), and finally a file with the description of typologies and sub-typologies used to group the figurines available in .docx and .rtf (M_OTHERING_Gr-typologies_subtypologies.docx; M_OTHERING_Gr-typologies_subtypologies.rtf) [1, 2, 3]. Typologies and sub-typologies (see below) were consistently assigned to figurines following a general overview of the collection to facilitate the grouping and future re-analysis of such figurines.

The database (https://www.doi.org/10.48493/vswy-ks92) is made available via “Hebe: data repository for the study of past childhood” [11] (https://pandoradata.earth/organization/hebe-data-repository-for-the-study-of-past-childhood), a data community accessible via the Pandora data platform (https://pandoradata.earth/).

Data type

The dataset is structured into a series of fields that record archaeological, historical, spatial, chronological, typological, and iconographic information. To improve readability, these variables are grouped below into thematic categories.

• ID and Number Information
  Each record is assigned a unique “Database ID”, consisting of an acronym referring to the archaeological site of discovery followed by a sequential number. In some instances, the original source describes multiple figurines or artefacts collectively. In such cases, a single record is created in M/OTHERING_Gr, and the total number of items is specified in the “Number of Items” field.
• Geospatial Information
  M/OTHERING_Gr is designed as a georeferenced dataset. For each entry, the location of the archaeological site is documented through the following fields: ‘Historical Region’, ‘Latitude’, ‘Longitude’, ‘Exact Location’, and ‘Estimated Uncertainty Radius (km)’. Coordinates are expressed in decimal degrees according to the WGS84 standard. When precise coordinates are not provided in the original publications, they are reconstructed using Google Earth in combination with the closest identifiable archaeological or environmental feature. This is indicated in the ‘Exact Location?’ field (Yes/No; ‘y/n’). The ‘Estimated Uncertainty Radius (km)’ field specifies the degree of spatial approximation.
  In cases where the provenance is broadly known but cannot be assigned to a specific site, standardised labels have been adopted to indicate different levels of spatial uncertainty. When attribution is possible at the regional level, labels such as UOA (Unknown Origin Attica), UOB (Unknown Origin Boeotia), UOCr (Unknown Origin Crete), UOM (Unknown Origin Macedonia), and UOP (Unknown Origin Peloponnese) are used. In cases of more general uncertainty, where no specific region can be identified, the label UOG (Unknown Origin Greece) is employed. In all such cases, precise coordinates cannot be provided.
  In the case of funerary material, the exact position of individual tombs is often unknown, particularly when not reported in excavation records.In such circumstances, and in the absence of more precise information, an approximate uncertainty radius of 3–4 km from the associated settlement has been adopted. Archaeological sites are organised according to broad historical regions conventionally used in the study of the Greek world (e.g. Attica, Boeotia, Thessaly, Phocis, Locris, Aetolia, Macedonia, the Peloponnese, Crete, Euboea, the Ionian Islands, and the Aegean islands). These regional designations follow widely accepted scholarly and cartographic conventions (e.g. the Barrington Atlas of the Greek and Roman World), although their boundaries are not fixed and may vary depending on chronological and historiographical perspectives. In a few cases, where the identification of the ancient site is uncertain or where modern toponyms are more commonly used in the literature, the modern place name is indicated in parentheses in order to facilitate identification.
• Contextual Information
  The interpretative value of figurines depicting adults with subadults is closely linked to their archaeological context. For this reason, the dataset records both the ancient ‘Settlement’ and the specific findspot (‘Site’). The ‘Context Description’ field classifies the depositional environment using a controlled set of categories, including religious, funerary, domestic, and other contexts. More specifically, contexts are classified into sanctuaries or votive deposits, tombs, caves, domestic spaces, workshops, and harbour contexts.Additional variables refine the spatial setting by distinguishing between urban areas, urban acropolis zones, urban agoras, the Pnyx in Athens, areas near city gates, suburban zones, suburban areas close to city gates, and extraurban contexts.
  This classification framework allows the analysis of figurines in relation to their depositional environments, which are closely connected to their function and meaning. In the case of religious contexts, it also enables the positioning of sites within broader urban, suburban, or extraurban ritual landscapes, sometimes with further topographical specification (e.g. acropolis areas or locations near city gates).
  Despite the inherent difficulties in defining precise boundaries of ancient settlements, this system provides a useful tool for exploring patterns of religious practice and issues of accessibility for different social groups, including pregnant women [12, 13, 14, 15].
  The dataset also records the presence of water in proximity to the findspot (‘Presence of Water?’), as springs, rivers, lakes, and marshes are frequently associated with female divinities and fertility-related cults in the ancient Mediterranean [16, 17].
  This information is derived either from archaeological descriptions or from the modern topography of the site. Finally, although recognising the polytheistic nature of ancient religion, the principal deities associated with each sacred context are documented where possible (‘Main Deities’).When multiple deities are attested within the same sanctuary, they are listed together following the terminology adopted in the original sources.
• Chronological Information
  Chronological information is recorded as reported in the original sources and expressed as a possible range (‘Min. Year (BCE)’; ‘Max. Year (BCE)’). Such ranges often encompass multiple centuries, reflecting the difficulties encountered in the archaeological literature when attempting to establish precise production dates for terracotta figurines. Since this is a fully numerical field, BCE years are indicated using a minus (“-”) sign. Positive numbers correspond to CE dates. In the present dataset, a single entry extends beyond the BCE range and is recorded as 50.
• Burial Information
  When figurines derive from funerary contexts, the database records information about the individual buried in the tomb whenever such information is available in archaeological literature. The field ‘Buried Person’ documents the biological profile of the deceased, including age category and sex when these can be identified. A controlled set of standardised categories is used to describe these individuals, distinguishing between adults, infants or adolescents, and, when sex is reported, sex-specific categories such as female adults and female infants or adolescents. Information on burials belonging to male individuals is also collected, although currently none is present in the dataset.
  It is important to stress that the attribution of sex and age does not derive from original analyses conducted within this project, but reflects the identifications proposed in the archaeological publications. In many cases, these attributions are based on the composition of the funerary assemblage rather than on osteological analysis. These identifications should therefore be considered with an appropriate degree of caution.
• Typological Information
  To classify the figurines, the typological system developed by Pedrucci [1, 2, 3] has been adopted. This system organises figurines depicting adults with subadults into primary typological categories.
  The two main types represented in the present dataset are:
  
  1. Kourotrophos (KT). A woman breastfeeding a subadult.
  2. Kourophoros (KP). A woman carrying one or more subadults.
  
  In a very small number of cases, the dataset also includes Double Kourophoros (DKP) representations, in which two adult female figures are depicted together in association with subadults. In the Greek material currently recorded, these cases are extremely rare and consist of two groups of women, one of which includes a woman holding a subadult.
  Additional typological subcategories developed within the broader M/OTHERING project, through the study of both the Greek and Italian material, are described in detail in the supplementary file M_OTHERING_Gr-typologies_subtypologies.docx.
• Technical Information
  The manufacturing process in the creation of a terracotta figurine is identified in the ‘Manufacturing Technique’ field. This distinguishes between mould-made, handmade, or mixed techniques. Recording this information is important for understanding technological practices and workshop traditions associated with coroplastic production. While mould-made figurines may reflect the use of standardised models, the available documentation for the Greek material does not necessarily indicate large-scale serial production comparable to that observed in some Italic regions [1, 2]. Handmade figurines, on the other hand, often introduce a greater degree of individual variation.
  Object dimensions are reported with reference to height through the ‘Reported Max. Height (cm)’ field. Whenever multiple artefacts share a single ID, the maximum height is recorded. The field ‘Complete Object’ (Y/N) assesses the completeness of each figurine, noting whether it survives intact or in a fragmentary state.
• Stylistic Information
  M/OTHERING_Gr records a series of variables describing the visual and iconographic characteristics of figurines depicting adults with subadults. These fields capture key features that enable comparison across different sites and regions.
  Basic descriptive variables include the number of adult figures (‘Adults’ Number’) and their gender as expressed through clothing or anatomical features (‘Adult’s Gender’). The presence of breastfeeding is also recorded (‘Breastfeeding?’), together with the specific breast used (‘Breast Used to Breastfeed’).
  The relative position of figures within the composition is documented through the fields ‘Adults’ Position’, ‘Number of Subadults’, and ‘Subadults’ Position’. These variables help reconstruct spatial relationships between figures and highlight degrees of hierarchy, interaction, and naturalism within the representations.
  The dataset further includes variables capturing specific iconographic markers. These comprise the presence of bird-like facial traits (‘Presence of Bird-like Face?’), as well as the identification of particular figure types such as actors (‘Is he an actor?’), Silenus-like figures (‘Is he a Silenus-like figure?’), and Pataikos-type figures (‘Is he/she a Pataikos?’). Additional fields record the presence of twins (‘Presence of twins?’) and possible identifications with mythological figures (‘Are they supposed to be Aphrodite and/or Eros?’). In this latter case, the identification is primarily based on the presence of a winged child, which is often automatically interpreted in the literature as Eros in association with Aphrodite. However, this attribution remains interpretative: the presence of wings may more broadly allude to an otherworldly or liminal dimension, rather than necessarily indicating a specific mythological pair.
  These categories are particularly relevant for the Greek corpus, where they point to the role of theatrical imagery and mythological narratives in the creation and reception of figurines depicting adults with subadults. They also intersect with the representation of grotesque or servile figures, especially nurses and pedagogues, and mark a significant difference from the Italic corpus, in which such theatrical and mythological dimensions appear to be much less prominent.
  More generally, mythological and theatrical subjects are significantly more represented in the Greek corpus than in the Italian dataset. These include figures such as Silenus, the nymph Tyro or the Dioscuri in some representations of twins, and Pataikos-type figures. In addition, theatrical imagery is particularly prominent, especially in representations of stock characters such as the nurse and the pedagogue. These figures are almost absent in the Italic material, whereas they occur in the Greek corpus, likely because they are more commonly associated with funerary contexts rather than with votive practices.
  Further variables document visual attributes associated with social roles or stylistic conventions. These include the presence of grotesque or non-normative facial traits (‘Presence of Adults’ Grotesque or Servile Traits?’), which in coroplastic representations are often associated with figures interpreted as nurses and, in the Greek material, also with pedagogues responsible for the care of subadults mentioned above [10, 18]. The presence of jewellery (‘Presence of Jewellery?’), sacred objects (‘Presence of Sacred Objects?’), and animals or toys (‘Presence of Animals or Toys?’) is also recorded.
  Finally, the dataset records selected bodily and stylistic characteristics of the subadult figures, including clothing (‘Subadult Clothes’), which may also reflect age-related iconographic conventions, biological sex when identifiable (‘Biological Sex of the Subadult’), and the presence of adult female nudity (‘Presence of Adult Female Nudity?’), a relatively rare feature in Classical iconography [19].
  Together, these variables aim to document what is visually represented, including the number and arrangement of figures, bodily and visual attributes, and specific iconographic markers. This structured description supports the analysis of caregiving practices, social roles, and the symbolic or mythological dimensions of these representations. Beyond their descriptive function, these variables also contribute to the analysis of how age and sexual difference are visually constructed through coroplastic imagery, offering further insight into the cultural encoding of bodies and roles.
• Relational Information
  The M/OTHERING_Gr database includes variables describing the relational and interactional aspects represented in figurines depicting adults with subadults. A central element of caregiving is breastfeeding, and this behaviour is recorded whenever present (‘Breastfeeding’). The database also notes whether the left or right breast is used (‘Breast Used to Breastfeed’), allowing the identification of potential stylistic or compositional patterns.
  The dataset further records signs of care directed toward the subadult (‘Care toward the Subadult’), inferred from protective gestures and facial expressions. These observations are complemented by the registration of eye contact (‘Eye Contact between Adults and Subadults’) and body contact (‘Body Contact between Adults and Subadults’), which serve as indicators of interaction and nurturing behaviour.
  This type of information documents observable interactions between figures (such as breastfeeding, touching, or close bodily positioning) which help interpret the relational dynamics represented in the figurines. These visual cues may provide insight into caregiving practices and emotional bonds between adults and subadults within the represented scenes.

Format names and versions

.xlsx; .csv; .docx, .rtf.

Creation dates

Records created between 2024 and 2026.

Dataset Creators

Carlo Cocozza and Ricardo Fernandes designed the metadata structure; Giulia Pedrucci collected and verified the data.

Language

English

License

Creative Commons Attribution-ShareAlike.

Repository location

https://www.doi.org/10.48493/vswy-ks92

Publication date

06/08/2026