Ghost Children: Delayed Personhood and Culture-specific Models of Infancy in Western Anatolia

Bestattungspraktiken zu entschlüsseln. Durch unsere anatolische Fallstudie liefern wir sozio-anthropologische Erklärungen für die Aufbewahrung von „ghost children“, die in der Nähe von Häusern aufgrund einer verzögerten Persönlichkeitsentwicklung begraben wurden. Wir spre-chen uns daher für die Konstruktion kulturspezifischer Modelle des Säuglingsalters auf der Grundlage der archäo-logischen Belege in Anatolien und darüber hinaus. Abstract: Most societies view infants and partially children, up to a certain age, as not fully human beings and/ or persons. This paper takes a longue durée perspective to examine built spaces shared by the living and dead infants during the four millennia (seventh to third millennium BC) in Anatolia. Evidence of infant burials within and around houses in several prehistoric periods and sites is analysed through a child-centred approach to mortuary remains, which does not equate adults with subadults or fully human with not fully human beings. This allows us to gain new perspectives of how age, age groups and infancy or childhood were perceived in prehistory. By perceiving houses as social spaces where ritual and non-ritual mimesis is embodied in shared practices and beliefs, where the material and social collide, rather than simply as signifiers of social units, we are better able to grasp subadult identities and decipher the personhood of infants and children through mortuary practices. Through our Anatolian case study, we provide socio-anthropological explanations for keeping the ‘ghost children’, buried close to houses, due to delayed personhood. We argue for con-structing culture-specific models of infancy based on the archaeological evidence in Anatolia and beyond.


Introduction
In response to Ian Kuijt's1 influential article on remembering and forgetting the dead in the Pre-Pottery Neolithic Near East, Chris Fowler2 suggested that instead of exclusive biographical individual identity of the dead, "a blurred and more complex relationship between the aspects of the person might be expected where death and life are integrated". The evidence for the integration of life and death was not only common during the Pre-Pottery and Pottery Neolithic in the Near East, but also during Late Chalcolithic and Early Bronze Age in western Anatolia. In the latter contexts, exclusively child and infant burials have been found interred within settlements, close to the houses, walls and hearths. By taking a step away from child mortuary practices informing us about status and demographic trends, examining the spatial distribution of child burials linked with concepts of personhood may help us in addressing complexities and emotions underlying children's deposition3. As we will show, by remaining critical of our own socio-cultural biases, ethnographic examples and cross-cultural studies of personhood may continue to help us interpret child mortuary practices in prehistory and enable us to construct new, culture-specific models of infancy, based on the archaeological evidence. During the later Chalcolithic and Early Bronze Age in western Anatolia, communal beliefs in delayed personhood that implied burial of 'ghost children' within settlements and close to houses, co-existed with complementary beliefs of burying children with adults in communal burial grounds, as it is recorded at Ilıpınar, Demircihüyük-Sarıket and Karataş-Semayük.

Child-centred Approach
This contribution discusses the topic of infant burials in archaeological contexts from a cross-cultural perspective based on the evidence from Anatolia. The Neolithic period in Anatolia (tenth to sixth millennium BC) is marked by enormous variability in the treatment of adult and child bodies, as well as the location of burials. This diversity is evident in distinct regions as well as on a micro-regional or local scale. Placing dead adults and children in burial grounds and keeping the infants inside the settlements appears to be the rule, especially in the later Chalcolithic and Bronze Age periods (fifth to third millennium BC). For discussing the crucial question of why infant burials, in particular, remain intramural, their dead bodies linked with houses and settlements, a child-centred approach is needed. As we argue, children and infants were not able to 'leave' the house and be on their own as they were not full persons yet. Their 'ghost' or spirit was reabsorbed by the 3 Barba 2020. house or household in which they were buried. The 'ghost children' refers to not fully human persons, mostly infants and children that were buried intramurally.
In the following, we use 'intramural' to refer the co-occurrence human burials and residential architecture in archaeological contexts4. In archaeological literature, intramural burial context refers to either the one or the other, meaning to burials within houses or within settlements. For example, Ward5 defined intramural burials as "those graves located in a domestic structure (or an associated feature) within a settlement, or in a new unit constructed primarily for death and/or interment a short distance away"6. We use the terms 'infant' for "individuals younger than 24 months" and 'child' for those "aged between two and twelve years of age"7. To the latter age group we refer as 'subadults' or 'children' interchangeably to explore non-modern, non-ethnocentric definitions of age classes (based on osteological evaluation) and their association with funerary practices in our analysis8.
As we understood that adults and subadults underpinned the past and future within 'history houses' at Çatalhöyük9, we should then pose a question of how can we explain the shift towards the differential treatment of infants in comparison to adults and children -especially during Chalcolithic and Early Bronze Age in western Anatolia. This contribution does not study burial treatments in their own right, but considers transitions in the life course of individuals (following Joyce 2000) and the fact that "children are the key to understanding the links between the past and the future"10. By understanding child and infant burials as "one of the significant locales where social and cultural reproduction took place, and where we can locate social and cultural change"11, does it mean that during the Chalcolithic and Early Bronze Age period we witness a stronger in-group distinction regarding age, between adults, children and infants than during the Neolithic in Anatolia? How could this shift towards the differential treatment of adults, children and infants be explained through children's life cycles, which are of primary concern to adults, since children link the past with the future?
If Neolithic Çatalhöyük was "as much a cemetery as a settlement"12, why were the Early Bronze Age settlement sites such as Bakla Tepe, Çukuriçi Höyük and Troy primarily settlements with the exception of also being places for deposition of infants? Moreover, why were infants and children at Ilıpınar, Demircihüyük-Sarıket and Karataş-Semayük buried within communal burial grounds whereas at Bakla Tepe, Çukuriçi Höyük and Troy, infants and children were buried within settlements? We will address these questions through an ethnographic analogy and a wider socio-cultural anthropological approach towards the understanding of children in non-state, sedentary communities. For if we understand that "the elementary forms of kinship, politics, and religion are all one"13, then subadult burials should be analyzed through all three lenses simultaneously, rather than detached from one another. The latter point has already been highlighted in several archaeological and socio-cultural anthropological studies of mortuary practices14. Moreover, qualitative dimensions, such as personhood, practice and emotions, may be also fruitful in addressing social complexity, which has been recently shown through children's burial practices during the Egyptian Predynastic period15.
The transformation of children into adults and the socialization of bodies into girls or boys have stood at the centre of multiple ethnographic observations among non-state sedentary communities in the past century16. The call to "go beyond seeing the human remains discovered in settlements as unusual/atypical/non-funerary discoveries" and instead consider them as "traces of complex multi-stage funerary practices, which contributed to the creation and manipulation of collective identities"17 has been recently made about Balkan (E)Neolithic evidence of intramural burials. By bridging this existing ethnographic knowledge with the archaeological evidence of child and infant burials, we propose that deposition of children and infants within settlements during the Anatolian Chalcolithic and Early Bronze Age Anatolia are not examples of 'deviant burials' but rather 'delayed personhood'18.
As we will show below, based on a wider regional comparison, deposition of infants within settlements was a norm and not an exception. This evidence speaks against infant burials being 'deviant burials' if we understand 'deviant burials' as those that most archaeologists define as "different from the normative burial ritual of the respective period, region, or cemetery"19. It has already been proposed that "deviant life and deviant death does have an effect on treatment and deposition of the corpse" and its investigation in prehistoric times should be undertaken in a "cautious manner"20.
It is widely attested that children in a large number of societies were in comparison to adults treated differ-ently21. Children and infants buried within houses were common across Anatolia during the Chalcolithic and Early Bronze Age periods22. Burying children and neonates intramurally was a practice that seemingly ensured an undisturbed reproduction of the household and the local community. Children and neonates deposited within settlements at these sites were therefore not only a concern to the house or the household within which they were buried but the local community as well, as it will be shown below.
The above-presented argument will not be based on the principle of "add children and stir" that is "no more insightful"23 than the famous Tringham quote in Households with faces, "add women and stir"24. Instead, in a longue durée perspective, the archaeological evidence of child and infant burials may help us understand shifting practices between these interrelated and co-dependent bodies of evidence. Furthermore, through the archaeological evidence of child and infant burials, we may construct culture-specific models of infancy with the help of new sexing methods recently developed for subadults25. How ethnographic insights already have been and remain valid for complementing our understanding of the ways of depositing subadults deposited in the systematic archaeological contexts will be summarized below.

Philosophical-religious explanations of mortuary practices
Based on a cross-cultural survey, Lewis Binford26 showed that age commonly determines the burial location. He claimed that in egalitarian societies, subadults were oftentimes buried under the house or in private household spaces, or close to the settlement. In contrast, adults in egalitarian societies, according to him, would be commonly buried within the public areas of the settlement. Binford explained this practice by the lack of communal engagement concerning the deposition of subadults.
Binford's interpretation somewhat resembles more recent interpretations of subadult burials in the wider Aegean basin. Regarding the record from the Greek Neolithic, it has been argued that "funerary rituals seem to have occurred within the family and were not yet a means of integrating the whole community" in the seventh and sixth millennium BC27. In this case, the authors argued that both dead adults and subadults were exclusively the concern of the household or the family they belonged to.
Fowler's28 systematic study on Neolithic mortuary practices agreed that the intramural burials at Neolithic Lerna, Dimini, Agia Sophia, and Pevkakia, all located on the Greek mainland, may indeed be an indicator of the family's concern. However, he simultaneously acknowledged that it would be "a leap to conclude that individual families were responsible for the mortuary treatment, had control over ritual practices, and that the wider community had no influence over the procession of the scale of rites"29. Here, it is important to stress that Fowler started the sentence with "a leap to conclude"30. Himself inspired by Hertz's writing on death31, the Hertzian triangle32, and Carr's33 important cross-cultural study based on Human Relations Area Files (HRAF), Fowler34 could not accept that family alone could be responsible for burying either subadults or adults within or close to houses, countering Binford's35 narrative. Therefore, we should accept that even in the case of intramural deposition of subadults within houses or 'private spaces', it is not only the household or the family but also the local community that is involved, as will be elaborated below.

Age as well as local beliefs matter
Two decades after Binford's36 influential study,Carr's37 Mortuary Practices: Their Social,Circumstantial,and Physical Determinants extended Binford's initial interpretation regarding the location of burials. Based on the HRAF cross-cultural sample, Carr38 identified the four most common determinants of local mortuary practices in egalitarian societies. Instead of no community involvement in subadult burials, as initially argued by Binford39, Carr's study showed that the four most influential determinants are: (1) beliefs about the soul's nature, (2) beliefs about the universal orders, (3) beliefs of the nature of the afterlife, (4) age. Difficulties to examine the first three determinants with archaeological data persist and therefore, most archaeological interpretations of mortuary practices are mostly based on age. Age, however, cannot be treated as the only determinant. Carr40 summarized the three complementary non-age determinants of mortuary practices under philosophical-religious reasons for burying people in specific locations or in specific ways. Philosophical-religious reasons, therefore, need to be simultaneously addressed alongside age that is more "social in nature"41 to fully grasp the multiplicity of reasons behind a certain location and type of burials.
Carr's42 philosophical-religious determinants fall within what an influential socio-cultural anthropologist of our time, Maurice Godelier, called "political-religious relations"43, a term he coined to describe what holds societies together. In his response, he denies that it is kinship (either fictive or biological) that makes a set of kin groups into a society. As he argues, it is political-religious relations that unite and provide means for the overall reproduction, including access to land, which is of prime importance in more or less sedentary farming communities. As Godelier puts it: "To exist as such, a society must exist as a whole that unites all of the groups that form it and at the same time encompasses them, because this whole lies at another level, the level of political-religious relations, which cement its unity in a largely (for us) imaginary and symbolic manner and ensure, by means that are not imaginary or symbolic (e. g. warfare, access to hunting grounds, etc.) its overall reproduction"44.
Godelier's remarks on political-religious relations may have fruitful implementation within archaeology. For an archaeologist, the first step is then to accept that no tell site, no village site, or other types of settlements, would come into being without political-religious relations to encompass these seemingly independent units such as buildings, houses and households. If political-religious relations unite groups of households, adults and children, men and women, how are we to understand the role of children in these non-state, non-literate societies? What distinctions are there between children and adults in such societies? And finally, what types of political-religious relations govern the deposition of children in these societies, if they leave this world prematurely? To address these questions, we first need to address how children come into being. Based on a qualitative, cross-cultural study of how children in diverse societies come into existence, Gode-lier45 argues: "Nowhere, in any society, do a man and a woman alone suffice to make a child. What they make together … is a fetus, but never a complete, viable human child. For this, other agents are needed, who are more powerful than humans, present in the vicinity but normally invisible, and who add what is lacking for the fetus to become a child. What is lacking is what customary we call a soul, a spirit, in short a usually invisible component but one which is not necessarily immaterial, since the soul can reappear after death in the form of a 'ghost,' which has a material, visible form but is usually intangible"46.
Based on this understanding of a child as a unit of mother's and father's contribution and the supernatural component, be that a spirit or a 'ghost', we can postulate that not only at birth but also at the death of individuals, both humans and spirits are involved. At the birth of a child, the spirit is necessary for the creation of a truly social person, a person that is "enrolled in a social whole (tribe, ethnic group, religious community) and a cosmic whole that extend beyond the universe of kinship rela- tions"47. At death, the process is reversed. To dis-enrol from a social and cosmic whole, the body and the spirit inhabiting the body must follow particular dis-enrolment rites that enable the group's undisturbed reproduction and reinforce the social whole. "The child is at the heart of kinship"48, as it is born from a union of human and the supernatural that allows society's reproduction, binding the past and present with future. Therefore, the child-centred perspective in understanding mortuary practices are particularly important if we accept that one does not bury one's enemies, but that burials attest to at least two things: "a) The existence of ties of kinship or friendship between those who were buried and those who buried them b) The existence of beliefs that death is not the end of life, that something of the deceased person lives on after death"49.
Following these initially established predispositions -regarding care for deceased, humans being shaped through political-religious relations, and death is not the end of human life -we can return to the archaeological record. While some scholars understand the burial ground's location being determined by cosmology and beliefs in the afterlife50, others understood that a type of descent group determines post mortem location51. It has already been pointed out that infants and children under house floors would allow a continuous relationship between living members of a household and its dead souls52. All these aspects taken together can be summarized under political-religious relations53 that govern children's coming into being, their integration into the local social whole, and their disintegration at the death of individuals. But what seemingly unites children in non-state societies cross-culturally, is a shared belief in delayed per-sonhood54, a shared feature or belief that may furthermore govern a differentiated mortuary deposition of children in comparison to adults.

Ethnographic evidence of child burials and delayed personhood
Before we dive into a specific ethnographic case of how delayed personhood might affect the deposition of children, let us first briefly highlight a diversity of conception and child mortuary practices in different sedentary, nonstate societies. This is necessary for the understanding of delayed personhood, which argues that personhood does not begin with conception but only after a full integration into the social whole55. The use of the 'simple' analogy56, applied in the following section of our contribution, allows us to compare properties in similar cultural forms that follow a common determining structure. Better known as a controlled comparison within socio-cultural anthropology57, it supports selective, qualitative parameters in choosing suitable ethnographic cases for comparison. The shared cultural forms and determining structures applicable to the ethnographic examples described below and to the Chalcolithic and Early Bronze Age settlements in western Anatolia include non-state, non-literate, sedentary farming communities, residing in villages or proto-urban settlements. Since a 'simple' analogy58 "can be a powerful tool to tackle questions for which no empirical answers exist"59, let us now provide a few examples where empirically attested deposition of subadults supports a shared belief in delayed personhood.
Among the Okinawans, a proper child burial was denied to everyone below seven years of age60. Children below this age limit do not require the attendance of people other than immediate relatives61. Among the Gikuyu in the central region of today's Kenya, it has been observed that child death is strongly associated with sorcery and witchcraft, and to a lesser extent ancestral spirit. It is believed that a murogi (sorcerer or witch) may cause a woman to be infertile or miscarry, or her children die in infancy62. Among Gikuyu, belief in witchcraft as causing the death of a child has slowly replaced the belief in ancestors and spirits causing misfortune63. Due to the spread of Christianity in the region, the same replacement of the belief -from blaming ancestors for child death to ascribing it to witchcraft for misfortune -has also been observed among the Bugisu of Uganda64. Among the Banyoro people in modern Uganda, if both mother and child died during labour, the foetus would be removed and buried in a separate grave65. Among the Nuer, at the birth of the first child, a cow would have been paid by a father to his wife's mother. This practice, known as the yang joghni, translates as "cow of the spirits"66. In case the child died, it was believed that the father's spirit should be calmed67.
Among the LoDagaa68 in modern Ghana, children up to the point when they can walk and talk were not treated like human beings69. If an infant died before that, being roughly around the age of three, the infant was buried "under a mound of earth beside the first crossroad on the path leading to the mother's home"70. The Akkan and Dagara's shared perception of children belonging to the spirit world71, has been recently documented also among the Kayan, a village and a language group on the north coast of Papua New Guinea72. According to Kayan cosmology, "persons did not necessarily come into being at birth. Newborns were believed to be still part of the spirit world. And only when it became clear that an infant would survive and not be taken back by the spirits, would locals -especially men -permit themselves to be emotionally attached and make preparations to introduce him or her to adult life"73.
A few of these ethnographic insights point towards a differentiated treatment of children in comparison to adults. Whereas in these societies adults received a complete burial treatment, linked with mourning, ceremo-63 Ibid. 64 La Fontaine 1963, 218-219. 65 Roscoe 1923, 250. 66 Howell 1954. 67 Among the Nuer and many Nilotic people, including Dinka and Shilluk, sexual intercourse between a pregnant or breastfeeding woman and her husband is strictly forbidden. If this taboo is breached, the woman or her child will be harmed. An example of such an occasion has been documented among the Dinka, where a man had sexual intercourse with his breastfeeding wife. As the child died, the child's father needed to pay a cow as compensation locally known as puk de thiang for breaching the law and causing the death of his child to mother's father (Howell 1954, 119 nies, and primary and secondary treatment of the dead, dead children were commonly associated with sorcery, witchcraft and an alternative way of depositing bodies, commonly with little or no mourning. This might require a separate cemetery, away from the settlement. Among the Marshallese people, Carucci74 observed that people build houses at a distance from cemeteries, as they were perceived as 'full of spirits', and residing away from cemeteries was perceived as being 'safer'75. In Hawai'i, the established practice of burying dead infants was close to or behind a house76 as "sometimes [even] the spirits of infants can be highly dangerous [kaueitatta]"77.
That babies are not always persons has been recently shown through a rich cross-cultural analysis examining more than 200 ethnographic cases covering all subsistence patterns, from the Mesolithic to the present78. That deceased children were treated differently from adults until recent times can also be seen in an example from the House of Habsburgs. Unlike deceased adults who were immediately or as soon as possible transported 'home' for burial, children who died far away from home were, "interred in the next available crypt" of the kin group79. Therefore, it remains particularly valid that the immediate attachment of parents to their children is not universal but rather a very recent and culture-specific model of infancy80 that is today predominant in the global North. If we are to understand other culture-specific models of infancy, we should refrain from ethnocentric projecting of our notions of childhood and infanthood to the archaeological record. An important step in understanding the mortuary evidence of infants and adults is, therefore, to include delayed personhood81 as a tool, defined as "a firm foundation for building cultural models of infancy … for example, infanticide is excused on the basis that one is not disposing of a person. Chronic illness and failure to thrive can be explained away as the failure of body and spirit to fuse, to with the spirit drawn back to the other world … infants still have one foot in the spirit world, rendering them vulnerable to supernatural forces82".
Delayed personhood can also be translated into post mortem treatment of children. First of all, burial rites and mourning may be minimal or actively discouraged by the local community in the case when a child dies below 5 or 10 years of age83. That allows the family and the community not to focus on the dead child that was not a person yet, which is now the past, but onto the future, towards another birth. Such children are at death "mourned privately or not at all, and are interred discreetly, without ceremony"84. However, the belief in delayed personhood is shared, which brings us back to Carr's85 conclusion that both age and local beliefs determine post mortem treatment and burial location. The signature of delayed per-sonhood86 is a 'private', small-scale burial, a practice that can be well translated to the archaeological record. Before we proceed with a detailed ethnographic case of what delayed personhood means in practice and how it affects the burial treatment of children, let us summarize a few important points relevant to the child-centred approach to mortuary practices based on our review: -family or a household as well as the local community are (in)directly responsible for and involved in intramural deposition of children within houses87 -mortuary deposition of subadults is not only determined by (1) their (young) age but also (2) beliefs about the soul's nature, (3) beliefs about the universal orders, and (4) beliefs of the nature of the afterlife88; local political-religious relations govern the conception and deposition of children89 -both humans and spirits are involved in the deposition of dead individuals -kinship or friendship exists between those who are buried and those who bury them90 -delayed personhood: infants and children are not persons and therefore, are deposited differently than adults, commonly without a ceremony, in 'private' spaces91 -examining the mortuary record through 'delayed personhood'92 allows us to build culture-specific models of infancy

'Ghost children' and delayed personhood reflected in mortuary practices among the Akan of western Africa
Ethnographic accounts of burying children among the Akan of present-day Ghana and Ivory Coast provides a more detailed example of delayed personhood. Among the Akan, the birth ritual was associated with the ancestral spirits93. Birth did not grant membership in the commu-nity94. Only after a succession of rites did the child become accepted into the clan. A child in this transitional period, between birth and the name rituals, was known as 'ghost child'95. During the transitional 'ghost child' period that lasts for eight days after birth, the child was treated rather negligently. It was placed on an old rag or mat while the locally inferior baha fiber was tied as a charm around the child's neck and wrists. The child was not denied mother's milk but hardly encouraged to feed on its mother's breast. Both mother and child stayed these first eight days indoors, as the mother was considered unclean during this period. During these days, the infant was not allowed to be washed with water that was previously boiled, as this would mean killing it. During these first anxious eight days, when the ties binding the child to the earth were developing, the child must have received all possible help from both terrestrial spirits and abosom, meaning gods, to survive and stay on the earth96. During the first eight days, the child was scarcely considered a human being but more like a wandering ghost97. If a child died before the completion of the naming rites that took place on the eighth day after its birth, the child returned to the land of spirits98. In fear of avoiding the long-term sterility of its mother, dead child was beaten and buried in the village midden-heap99. The beaten body of a ghost child would be placed in a pot with sharp elephant grass (Pennisetum sp.) and buried near the women's latrine. The child's body was also occasionally mutilated, having a finger cut off before being wrapped in the sharp grass, then placed in a pot and buried in the village midden 93 For a detailed discussion and the argument that birth of a child is always linked with supernatural (be that spirit, soul, or a ghost), see Godelier 2011, 229-298. 94 For a study of delayed personhood, which shows that cross-culturally, children and infants are not considered as persons, see Lancy 2014. 95 Clarke 1930, 463;Manoukian 1950, 51. 96 Rattray 1923 Ibid. 60. 98 Manoukian 1950, 51. 99 Clarke 1930 heap, which was formerly the women's latrine100. Multiple observers reported parents of a ghost child expressing joy rather than mourning after the 'ghost children' left the earth101. The ghost child's parents would dress in holiday attire, enjoy the ground-nut soup, a marker of a joyful feast, and retire to their house lying together102. This practice discouraged the 'ghost children' to ever return and endanger the mother.
When an Akan child survived the eight-day period, there was the Nteatea rite on the eighth day after its birth, when a child was named by its senior relative. On this day, the child was for the first time regarded as a member of the human family and no longer as a ghost103. The baha fiber, the material used for the first adornment worn by the 'ghost child' was generally used for sanitary purposes but its use at the birth of a child was twofold: "First it satisfies that innate desire to protect the little stranger by the use of charms, which all necklaces, bracelets, and such-like originally were, and secondly the use of such inferior material with which to bind these, is to avoid any semblance of making a premature or too open claims to this new young being, who is regarded at this particular stage … nothing more than some 'ghost child' which has no intention of remaining long in this world"104.
Even when the child survived the initial eight days after its birth, and received a name, it has been reported that "still, the link with the land of spirits is not yet severed absolutely; the child grows up and lives in a kind of borderland between the world of men and women and the world of ghosts"105. It was not until puberty when the bonds with the spirit world completely disappeared, and when a 'ghost child' or the so-called 'pot child' became a man or a woman. In puberty, children gained the status of an adult, a grown mortal that in case of his or her death was now entitled to a proper burial and funeral rites106.
Among the Akan, children who died before puberty were buried on the village midden heap, rather than receiving ordinary funerary rites. They were classified similarly to the 'ghost children' who did not survive eight days after birth, as nkuku mma, meaning 'pot children', named after the pot into which the body would be placed for burial107. As Robert Sutherland Rattray understood the deposition of 'ghost children' in kitchen middens, he ordered an Ashanti man to dig one of them108. He dug out several skeletons, not only of infants in pots but also bones of older children. Rattray submitted a few of these bones for examination, which resulted in the identification of two individuals: an infant and a child approx. five years old109. This confirmed the observation that also older children were buried the same way as 'ghost children': "in former days children dying before puberty were considered to be 'ghost children' and treated in the same way; to accord funeral rites to such children would cause the mother to be barren"110.
Based on the above-detailed description of the Akan child burials, we again emphasize that the death of a child is not exclusively a mother's, household's, or community's affair; the affair affects all of these. The death of a child matters to both mothers and the household and at the same time, the wider community, be that village group or descent group, simultaneously. Whereas the community would disapprove of burial within the communal burial ground, the mother or possibly other members of a household wanted to keep the infant away from the communal burial grounds as not to violate the local beliefs. Therefore, they buried them close to them, within houses, not only because they would be emotionally attached to these children, that were no more than wandering spirits, but more likely not to violate the living community's beliefs and ancestral spirits. From both the communal and the mother's perspective, it was not a child who died but a spirit that was not yet human. Therefore, although the wider community would reject the infants' burial, the household needed to respect communal rules in burying off these not yet human individuals, possibly 'ghost children', away from the ancestral burial grounds.
Following this initial overview of a child-centred approach to mortuary practices, cross-cultural evidence of delayed personhood111, and a specific ethnographic example from the Akan deposition of 'ghost children' 111 Lancy 2014.

Archaeological evidence of infant burials from the site Çukuriçi Höyük
Today, the multi-period site Çukuriçi Höyük is located at the central western Anatolian coastline south of the ancient city of Ephesus. Based on a reconstruction of the prehistoric coastline, Çukuriçi Höyük was a coastal site in prehistoric times. The mound consists of 8.5m of occupation layers, dating from the seventh to the beginning of the third millennium BC, from the Neolithic to the Early Bronze Age period (for detailed information about

The Late Chalcolithic burial
The intramural Late Chalcolithic stone cist burial of an approximately 12 month old infant113 was discovered close to the northern border of the settlement mound in trench N3 ( fig. 2). Although the northern and eastern stone slabs of the cist (length: ca. 72cm) were not preserved, the burial was intact. The stratigraphic sequence indicates that the grave pit was dug into a Late Chalcolithic occupation level (ÇuHö VII). Due to the size of the trench, no nearby architectural remains were discovered. Inside the pit, the body of the infant was placed upon a layer of sandy clay. The skeleton was east-west orientated and lying on the right

The Early Bronze Age burial
The intramural Early Bronze Age 1 burial115 was discovered inside room 19 of settlement phase ÇuHö IV in the southern trench S2. The burial was situated in the southwestern corner, next to the wall underneath a floor level. In contrast to the Late Chalcolithic burial, the remains of a neonate were buried in a tripod cooking pot with burn marks from the outside (fig. 4). The vessel represents a typical Early Bronze Age shape of the settlement phases ÇuHö IV-III. In this case, the vessel lost its primary function and was secondarily used as a burial container. The    vessel contained no further determinable organic material ( fig. 5). The infant was placed on its right side in the southeast-northwest direction with a strongly contracted flexed position and the face to the north ( fig. 6). The burial can be dated by several radiocarbon dates from the settlement phase ÇuHö IV within the timeframe 2950/2900-2850 calBC116.
The two well-dated burials from Çukuriçi Höyük fit the prevailing tradition of burying infants in intramural contexts in western Anatolia during the Late Chalcolithic and Early Bronze Age period. Although there are different burial contexts -a stone slab cist and a cooking pot -this is not an unusual mortuary practice in the Aegean and western Anatolian during these times117. The following section provides additional support through the evidence of infant and child burials within settlements in Late Chalcolithic and Early Bronze Age western Anatolia.

Chalcolithic and Early Bronze Age: Spatial separation between places for the living and the dead
The detailed study of burial customs at the cemetery of Demircihüyük-Sarıket suggested that age is a determining factor in how members of the community were treated in mortuary practices118. Children and neonates between one and eleven years of age were commonly buried in small jars within the communal cemetery at Demircihüyük-Sarıket119. In contrast, individuals below one year of age were rarely buried in extramural cemeteries. Instead, they were normally placed underneath house floors, without any grave goods120. From the fourth millennium BC, intramural burials of neonates below one year of age are frequently found in western Anatolia121.  This rule can also be attested based on the evidence from Çukuriçi Höyük and other Late Chalcolithic and Early Bronze Age sites. Several subsequent studies have shown that older children and adults were mainly buried outside settlements, whereas foetuses, neonates and infants are commonly found within Chalcolithic and Early Bronze Age settlements122.
If the signature of delayed personhood123 is a 'private', small-scale burial, then the same signature corresponds to the archaeological record of infant burials at Çukuriçi Höyük and other archaeological cases from the western Anatolian Chalcolithic and Early Bronze Age. In our case study, it appears that delayed personhood was limited to one year of age124, when the child begins to crawl and may stand, handles finger-foods, uses spoons and cups, responds to a name, explores the environment, and interacts with other children, among other activities125. Only after this developmental stage could children be buried in extramural cemeteries. The intramural infant burials (tab. 1) could be seen as liminal, placed between the spirit world and the world of the living, have already been proposed for Chalcolithic Anatolia126. In the case of Late Chalcolithic Çadır Höyük, the concept of infant mortuary liminality within houses has been directly translated from Moore's127 study of the Roman-British period, where houses, corners of the rooms and walls, were perceived as liminal places in which infants were buried. In contrast, we aim to extend the concept of infant mortuary liminality by looking at different perceptions of infant and child identities through delayed personhood, which complements our understanding of burying infants within settlements and liminal spaces for the dead. If delayed personhood allows us to build culture-specific models of infancy, then based on the above presented archaeological data (tab. 1 and fig. 7), we can make a first attempt for the Chalcolithic and Early Bronze Age in western Anatolia. Nevertheless, micro-scale and site-specific beliefs in delayed personhood may have varied, as it will be further discussed in the last part of our contribution.

The Neolithic: Interweaving spaces of the living and dead
It has already been argued that the Neolithic, as a new mode of life, promoted a broad range of ways of dealing with the dead, indicating little intra-group distinction in terms of age and gender, as well as a great diversity of human depositional practices in different regions of Anatolia through time128. A short overview of mortuary diversity based on 'burial location' and 'age' during the Neolithic and Early Chalcolithic in southeastern and central Anatolia (tenth to seventh millennia BC) and western Anatolia (seventh to sixth millennia BC) serves as a contrasting example to the later Chalcolithic and Early Bronze Age practices that are the focus of this contribution.
Examples of rich and diverse Neolithic mortuary practices within or between houses can be seen from the Neolithic (tenth to seventh millennium BC) in southeastern Anatolia129 at sites such as Çayönü130, Nevalı Çori131 and Çatalhöyük132. At these southeastern and central Anatolian Neolithic sites, the Neolithic adults, children and infants were buried within houses, below the house floor, in pits, without any strict division based on sex or age of individuals133. These examples imply that spaces for the living and the dead mostly overlapped during the Neolithic. While the living inhabited house floors from above, 128 Lichter 2016. 129 Regarding the Middle Pre-Pottery Neolithic (PPNB) in the Near East, it has been observed that skull removal was applied to infants and adults (Kuijt 1996;. In contrast, Croucher (2012, 170) reports that at 'Ain Ghazal, only infants after 15 months of age were decapitated, which clearly depicts a difficulty in using terms such as 'child' and 'infant' without clearly defined age classification. Moreover, the example from 'Ain Ghazal supports the evidence that although dead adults and subadults could have inhabited the same spaces, deposition of their bodies followed a different pattern during the PPNB. This could be a suitable evidence for a differentiated mortuary of treatment of children in comparison to adults in a large number of societies (Weiss-Krejci 2011, 87 Coşkun et al. 2010, 19;Erdal 2015, 3. An equal burial treatment of children and adults within or close to houses has been also documented outside Anatolia, such as at the Neolithic intramural cemetery of Gomolava in nowadays Serbia, of the Vinča culture, dating to the fifth millennium BC. There, male children were buried alongside adults, which "suggests that Neolithic people considered them equal to adults, at least to a certain degree" (Stefanović 2008, 98). the dead inhabited them from below134. Moreover, the practice of burying the dead within or close to houses has already been identified as one of several markers of the Neolithic expansion from Anatolia to the Balkans135.
Çatalhöyük provides insight into the diversity of distinct mortuary practices within and between sites and Anatolia's micro-regions during the Neolithic and Early Chalcolithic. Considering the deposition of the dead only through intramural vs. extramural archaeological context, scholars have argued that the archaeological evidence from Çatalhöyük indicates equal treatment of individuals of both sexes and all ages136. In contrast to this interpretation of mortuary practices based on the 'burial location' and 'age', ways of depositing the dead and their spatial distribution within a house provide new insights on the deposition of infants and children at Çatalhöyük.
A few examples of ritual child burials from this site have been interpreted as evidence of child sacrifice137, pointing towards 'deviant burials' of infants at Neolithic Çatalhöyük. Overall, dead within houses at Çatalhöyük were spatially segregated between the spaces for the adults and those for subadults. Dead adults were mostly buried in the north and northeastern part of a house whereas the infants and young children were buried within the south and southwestern part of the house138. Moreover, the organic material in which the dead body was wrapped and subsequently buried within the house appears to vary based on age. Whereas mats for adult burials at Çatalhöyük were made of sedges, infant burials were placed in baskets that were made of different plant materials139.
Furthermore, a recent publication on the Neolithic burials at Çatalhöyük proposed that "genetic relatedness may not have played a major role in the choice of burial location … at least for subadults"140. This may indicate that a heterogeneous group of subadult individuals belonging to the same group based on age -and not based on genetic relatedness -were buried together. As our contribution shows, interpretation of subadult burials, including the beliefs and delayed personhood should be taken into account when interpreting ancient genomes from adults and subadults.  Yaka et al. 2021. In sum, although all age groups were buried within houses at Çatalhöyük141, the spatial division between adults and subadults and their different mortuary treatment within houses strongly indicates an existing difference in ways of treating dead infants and young children compared to adults. Although the dead did inhabit the house floors from above and below, they inhabited them differently. Dead bodies within houses at Çatalhöyük clustered based 141 Hodder/Pels 2010; Boz/Hager 2013. on their age and mortuary deposition that could be furthermore explored through delayed personhood142.
In our research area of western Anatolia, mortuary practices differed at the regional, micro-regional, and local scales during the seventh and sixth millennium BC. In northwestern Anatolia, mortuary practices comprise burying dead adults on the edges or peripheries of settlements while burying children within the settlement or 142 Lancy 2014. houses, such as at Ilıpınar, Menteşe and Barcın Höyük. Neo lithic settlements in northwestern Anatolia were places for the living and the dead, but their burial-related spatial position became more nuanced.
Based on burial location and age, a difference can be observed at Aktopraklık in northwestern Anatolia. During the Late Neolithic period, dead adults and children were buried in pits beneath the floors of huts of the Aktopraklık C settlement143. In contrast, during the subsequent Early Chalcolithic period, the settlement area shifted to Aktopraklık B, located 100m to the south of Aktopraklık C, and adults and children from Aktopraklık B were buried in a communal burial ground on top of the abandoned Aktopraklık C144. The case of Aktopraklık B, therefore, indicates a clear separation between the settlement and burial ground during the Early Chalcolithic period.
Heterogeneous evidence of mortuary practices is also observed from the Neolithic Lake District and coastal western Anatolia145. At Bademağacı, burials of children and adults were discovered inside the settlement area. Mainly subadult burials (60 % of children from which 90 % are below 2 years of age) were discovered in intramural contexts146. However, due to the small number of adult burials within the settlement, it can be assumed that the majority of the dead were buried in extramural burial ground(s), which have not yet been discovered147. Burials are also known from Hacılar148, Höyücek149 and Kuruçay Höyük150. In addition, no burials have been discovered inside settlements and cemeteries are equally unknown in coastal western Anatolia, with the exception of a single adult male burial under the floor from Ege Gübre151.
Similarly, in the entirety of the Lake District, no Neolithic "extramural graveyard or intramural collective graves have been determined"152. Twenty-two burials are known from Hacılar, including two burials with a child attached to its mother from Hacılar IIA153; seven from Kuruçay Höyük including two children (sk. 3-4) found close to the entrance of the Neolithic enclosure and underneath an attached bastion154 and one burial of a newborn discovered next to a wall outside of a house at Höyücek155. These examples from western Anatolia raise the question of whether the mortuary practice of burying the adults and children in extramural burial grounds was already in use during the Late Neolithic period in coastal western Anatolia and the Lake District. In any case, this evidence demonstrates that mortuary practices varied at multiple scales across and within particular periods and regions.
Based on this combined evidence, we could also distinguish between different practices in incorporating histories. It has already been argued that houses served as the main loci of incorporating history at Çatalhöyük156. If we treat adults as the 'carriers of history', then histories in the western Anatolian Neolithic period were more often than not moved away from houses to the edges of settlements, since adults were commonly buried on the edges of settlements, and the house and the settlement area only vaguely incorporated history. In the case of Aktopraklık, it was a communal burial ground with both adults and children that incorporated the site's history. In this case, it seems likely that both children and adults were perceived as the incorporators of history, as their dead bodies were placed in the same space, outside the settlement. Nevertheless, it has been recently pointed out that at Neolithic Aktopraklık, the lack of infant bones is stunning and therefore, the authors hope that "the sector where babies were interred will be detected one day"157. This could point towards the earliest evidence for a differentiated treatment of infants in comparison to adults during the Neolithic period in western Anatolia.
If we focus on children, then they were part of incorporating histories during the Neolithic, within either 'history houses'158 or communal burial grounds. In contrast, the case of adults buried on the edge of the settlement whereas children were buried within or close to houses points towards a differentiated treatment of dead children for incorporating history. Even after their death, children remained associated with the house, (e. g. at Barcın Höyük) which was not the main loci of incorporating communal history. Following the outline of con-siderable diversity of burying the dead and incorporating history during Neolithic and Early Chalcolithic periods in comparison to the later Chalcolithic and Early Bronze Age in Anatolia, let us now discuss a few insights stemming from our contribution.

Discussion of the archaeological evidence of infant burials and delayed personhood in western Anatolia and the east Aegean
During the Chalcolithic and Early Bronze Age in western Anatolia, a tradition of diverse mortuary practices based on the location of burials and the age of individuals buried -previously attested for Neolithic -was not maintained. Spatial separation between settlements and burial grounds, initially evident from Late Neolithic/Early Chalcolithic Aktopraklık, became the norm. Unlike at Aktopraklık, where dead children, neonates and adults were buried together in a common cemetery, burial grounds elsewhere in Late Chalcolithic and Early Bronze Age western Anatolia were more often than not reserved for the adults and older children.
Already K. Bittel mentioned that especially infants were buried in intramural contexts159 pointing to a general spatial separation which cannot be supported in all cases. The Late Chalcolithic cemetery of Ilıpınar provides a piece of rare evidence that infants and children were placed within pottery vessels in extramural burial grounds160. The Early Bronze Age burial grounds of Karataş-Semayük161 and Demircihüyük-Sarıket162 confirm that both adults and children were buried extramurally, within the communal burial ground, spatially separated from the settlement. However, age seems to be a crucial determinant for where the infants and children were buried -intramurally or extramurally. The separation between child and adult burials within communal, extramural burial grounds is not evident at Karataş-Semayük, where both adults and children were buried in individual pithoi or jars within a communal burial ground163. However, M. J. Mellink mentioned 'intramural' burials164 including children from hab-159 Bittel 1934, 30. 160 Roodenberg 2001, 352;2008, 317;324 fig. 3. 161 Mellink 1964, 272;1965, 241;1967, 256;cf. Stech Wheeler 1974. 162 Seeher 2000Wittwer-Backhofen 2000;cf. Massa 2014. 163 Stech Wheeler 1974, 78. 164 Mellnik 1967 itation areas, which indicate that subadult burials were also found within the settlement of Karataş-Semayük. A similar case is also known from Demircihüyük-Sarıket, where subadult burials are recorded from the cemetery165, and remains of skeletons from infants, newborns and foetuses were discovered within settlement layers166. Therefore, age seems to be a determining factor for subadult burials during the Early Bronze Age period at both of these western Anatolian sites, although that cannot be accepted as the only determining factor.
Age-related determinants and local beliefs both contribute to answering the question of why infants under one year of age were buried differently at these sites in comparison to older infants, children and adults. Based on our analysis of infant burials in a wider region of western Anatolia (see tab. 1 and fig. 7) and other detailed studies of cemeteries167, it appears a common rule that infants below one year of age were buried within settlements and houses, whereas older infants, children and adults were buried in extramural cemeteries. The striking similarity in the treatment of dead infants up to one year of age, who were during these periods commonly buried underneath house floors168 cannot be exclusively ascribed to the age of these individuals. Following our analysis and widely shared notion that both age and local beliefs determine the deposition of adults, children and infants169, then not only age but also shared belief in 'delayed personhood'170 determined intramural deposition of infants within houses. Based on differentiated mortuary practices in western Anatolia, 'pot children' or 'ghost children' as not fully human beings could only be placed in communal burial grounds alongside adults as fully human beings -after the infants turned one year old and transformed into persons.

Infant burials as remnants of 'history houses'?
If we now look closer at Anatolian Neolithic sites, we can also distinguish a few important differences between the treatment of the dead and the inhabitation of spaces for the living and the dead. At Çatalhöyük, a few elaborate houses with multiple burials concentrated under house floors have already been described as 'history houses'171. These long-lived 'history houses' have been defined as "buildings in which Çatalhöyük people accumulated more transcendent knowledge and symbolic capital than in others"172. 'History houses' at Çatalhöyük differed from other buildings in the number of human burials (commonly more than 10) and their elaboration in terms of wall paintings but not in the size of a building173. Hodder and Pels174 proposed that heads and other body parts removed from individuals in 'history houses' were possibly placed in other houses, which may have established an alliance between history and non-history houses. 'History houses' point towards the emergence of division of labour between houses, in which a few houses at times became a central focus of the settlement, by 'specializing in modes of incorporating 'history'175. But whereas some houses incorporated more history than others, based on the number of burials and building's elaboration, it is important to observe that incorporated history within these houses included dead adults, children and infants alike. Therefore, it seems that infants, children and adults all played an important, if not an equal role, in generating and incorporating communal histories at Çatalhöyük. Moreover, delayed personhood, as seen from the Akan and other cross-cultural ethnographic cases176 may be further explored at Çatalhöyük. Since dead infants and children were inhabiting houses below floors differently than adults based on the 'burial location' and 'age'177, a more nuanced explanation for the differentiated treatment of adults, children and infants while placing all of their dead bodies within houses should be possible, although beyond the scope of this contribution.
Neolithic mortuary practices evident from northwestern Anatolian sites such as Ilıpınar, Menteşe, and Barcın Höyük provide a different example. History at these sites was not evident in 'history houses'178, but on the edges of settlements and open areas, if we understand adults as full members of the local communities and incorporators of history. A house in Neolithic western Anatolia, therefore, was not necessarily the only loci for incorporating full members' histories. Instead, a separation of the adult individuals from the house and their incorporation to the site's periphery after death is evident from these sites. Deceased children and infants, who cannot be treated as ancestors, however, remained tied to the house also after their death, a practice that was continued from the Neolithic to Early Bronze Age in western Anatolia. This evidence shows that because children and infants were not able to 'leave' the house and be on their own as they are not yet full persons, their 'ghost' or spirit can be reabsorbed by the house or household, and therefore, the so-called 'ghost children'.
There is no evident similarity with Çatalhöyük-style Neolithic 'history houses' during Chalcolithic and Early Bronze Age periods in western Anatolia. Therefore, it can be presumed that history during later periods was incorporated within communal burial grounds that mostly comprised adults. Burials of young children and infants were mostly absent within these communal burial grounds, suggesting delayed personhood179 at these sites. The assumption of a different social position of children and infants within the community was already raised by biological anthropologist U. Wittwer-Backhofen during her examination of the burials from Demircihüyük-Sarıket180. In addition, whereas delayed personhood appears to predominate in most cases during the Chalcolithic and Early Bronze Age in western Anatolia, burial grounds at Ilıpınar and Karataş-Semayük should nevertheless be considered. At the latter two sites, besides indications of intramural burials, adults, children and infants were buried together in the burial grounds. This suggests the possibility that beliefs in delayed and non-delayed personhood co-existed not only at different sites, but also within sites during the Chalcolithic and Early Bronze Age. Clear evidence of delayed personhood could be ascribed to sites such as Aphrodisias, Bakla Tepe, Barcın Höyük, Beycesultan, Çukuriçi Höyük, Demircihüyük, Gavurtepe Höyük, Hanay Tepe, Kuruçay Höyük, Perge, Heraion on Samos, and Thermi on Lesbos, where infants were buried in pits, pots or stone cists within settlements.

'Ghost children' -a multilayered archaeological approach
It could be that we are dealing with different types of 'ghost children' who were deposited more or less similarly in pits, pots and stone cists, close to the houses that could reincorporate their 'ghosts' or spirits. In these cases, the setting within the settlement remains important. Whereas no other individuals were found buried within Çukuriçi Höyük's settlement either during the Late Chalcolithic or Early Bronze Age period, one thing is certain: infants and children below a certain age -commonly younger than one year -were the only ones allowed a burial within the house from whence they came. Burying infants within settlements was not a deviant act. Rather, it was the expectation of the wider social group, be that the village community or the lineage. Infants had no place within communal burial grounds as they are not humans but 'ghost children' that would disturb the ancestors' spirits within the burial ground, should they be buried alongside them. Infant mortuary liminality within settlements, observed at Late Chalcolithic Çadır Höyük181 can now be extended as a common infant mortuary practice not limited to Late Chalcolithic central Anatolia but also Chalcolithic and Early Bronze Age in western Anatolia.182 These 'ghost children' can then be divided into two groups. First are those with or without a name who died as infants below one year of age and were buried within settlements, within or close to houses (e. g. Aphrodisias, Bakla Tepe, Barcın Höyük, Beycesultan, Çukuriçi Höyük, Demircihüyük, Hanay Tepe, Heraion on Samos, Kuruçay Höyük, Thermi on Lesbos, Troy, see tab. 1)183. If we translate the age to developmental traits of a child, then these 181 Yıldırım et al. 2018. 182 Further evidence is also known from the central Anatolian site Çamlıbel Tarlası (Schoop 2008, 150-151 fig. 53;2009, 56-57 figs. 48-49;2010, 195-196 figs. 53-54;200;2011, 61-62 fig. 12). In the phases ÇBT II-III, neonates and young children were buried in jugs and older children in pits. On the one hand, the burials were found below the floors, inside of houses, on the other hand, next to the dwellings outside. 14 of 20 subadult burials belong to infants and children below six years (Erdal/Erdal 2017, 345-347). 183 Due to the unspecific information about the detailed age (just are all children that were not able to crawl and stand, handle finger foods, use spoons and cups, respond to a name, explore the environment, and interact with other children184. The second group are those with a name, above one year of age, who would be able to tell stories, engage in social interaction and role-playing, ask questions about the meaning of words, and understand distinctions such as 'male' and 'female', yet without being fully initiated into the social sphere as adults. They would also possibly be denied burial in the communal burial ground and be instead buried within or close to houses (e. g. Perge, Kuruçay Höyük and Troy, see tab. 1).
In both cases, a 'ghost child' was, however, not fully a human, and therefore was denied the 'standard' mortuary deposition that was accorded to adults. There was no space for 'ghost children' in the communal burial ground. Those 'ghost children', however, were cared for not only by mothers and members of the immediate household group but also the members of the wider (village) community, who denied burial of infants and in some cases children in communal burial grounds. The possibility that those 'ghost children' were to be kept close to the household so that they could enter the next body that was produced, thereby not being lost to the household or the larger community, should remain a possibility185. A belief in delayed personhood, therefore, was a feature that was commonly shared not only at specific sites in western Anatolia but encompassed a wider region. However, the Chalcolithic and Early Bronze Age 'outliers' in terms of shared belief in 'ghost children' and delayed personhood, may be in- Varies (up to 5-15 years of age of delayed personhood) Age limit to subadults buried differently from adults?

Burial location
Private, small-scale, incomplete Private, small-scale, incomplete, intramural, others? Important insights onto child mortuary treatment Babies aren't persons (Lancy 2014) Osteological age and delayed personhood matter for the construction of culture-specific models of delayed personhood

Late Chalcolithic & Early Bronze Age western Anatolia
A shared belief in delayed personhood is commonly limited to 1 year of age (but older children may be included, see tab. 1) dicated by the data from sites of Ilıpınar and Karataş-Semayük, which points towards a co-existence of different beliefs of what makes a child and when does a 'ghost child' become a person during these periods in western Anatolia.

Child matters to household and local community
The question that remains to be addressed is whether social groups who bury their dead within houses only belong to households and not to larger social groups, such as the clans and lineages or temporary kindred networks. Certainly not, if we look at a few cases, among which is also the Tikopia186. Although they belonged to a chiefdom and followed a conical clan structure, Tikopia buried their ancestors inside houses, under the house floor, similar to Neolithic dwellers at Çatalhöyük. Therefore, those buried underneath house floors at Çatalhöyük most likely belonged to wider social groups than households, as has been previously proposed187.
Household activities and food sharing practices point toward an interpretation that during both the Late Chalcolithic and the Early Bronze Age 1, adults at Çukuriçi Höyük belonged to groups other than their households188. This further implies that adults in such communities followed not only their household heads, but also the possible village council, the head of the village group, or other powerful actors beyond the household. Following these actors means following and respecting the local rules. It means that to become a member of a certain group or a village itself, one's children and their adult members should learn and follow some pre-existing communal rules that also affect the deposition of dead infants and children. The emerging members (infants and children) of these groups should be given a name and be initiated to become full members of these groups that cooperate in work, sharing of meals, festivities, work activities, and other reproductive activities. Children and infants would be considered 'incomplete humans' and 'non-persons'189 until completion of such 'rites of passage'190. Therefore, initiation is not only about how to initiate a man or a woman. It is the making of men and women into what is expected from them within the society191 so that society can live on.
What can be concluded based on the pit, pot and stone cist infant and child burials within the Chalcolithic and Early Bronze Age settlements in western Anatolia, is that in most cases, the local beliefs included delayed person-hood192. Because of shared belief in delayed personhood, infants younger than one year of age and older children, up to 13 years of age, at death, were buried intramurally, unlike adults (cf. tab. 1). Therefore, the age limit of delayed personhood193 seems to vary between or within sites, which could furthermore explain the evidence of burying children not only below one year of age but also older (cf. Kuruçay Höyük, Perge, Troy) within the settlement. Goody194 noted that "there are sound sociological reasons for expecting different types of burial customs side by side in the same society", a claim that has been supported in several other follow-up studies195. The mortuary evidence of 'ghost children' from Chalcolithic and Early Bronze Age Anatolia therefore also supports different types of 'ghost child' burials (in pots, pits and cists) of different ages at death that may co-exist within the same society, since personhood is more often than not negotiated and contested within communities (tab. 2)196.

Conclusion
When we started to think about the intramural infant burials at Çukuriçi Höyük during the Late Chalcolithic and Early Bronze Age 1 period, our first assumption was that following its singular occurrence in each phase, it was the mother or household who deposited the infant. Our initial interpretation, that only mothers or households cared for these dead bodies, is therefore most likely mistaken. Late Chalcolithic and Early Bronze Age dwellers at Çukuriçi Höyük were also united through death, through a differentiated deposition of 'ghost children' in comparison to adults as full members of the local community, considering that no adults were found buried within the settlement. A similar argument has been previously made regarding Late Chalcolithic mortuary practices in Anatolia: "housing areas were mostly used for sub-adults and older children; adult males and females were possibly interred to the extramural graveyards"197. A practice that could resemble Rattray's description of 'pot children' or 'ghost children'198 commonly deposited in a midden heap among the Akan is a reminder that child's death and birth in non-state, sedentary societies are a combination of both human labour and the supernatural. Neonates and young children do not have personhood at birth but delayed personhood, the age limit of which is often culture-specific and can vary greatly. If we agree that "nowhere, in any society, do a man and a woman alone suffice to make a child"199 then also nowhere, in any non-state sedentary society, do a man and a woman alone suffice to deposit their child. For how are they to deposit a body if they do not get rid of the spirit that may turn into a ghost, return to earth in the shape of a human, but leave prematurely? In conclusion, the later Chalcolithic and Early Bronze Age in western Anatolia, communal beliefs in delayed personhood co-existed with complementary beliefs of burying children and neonates with adults in communal burial grounds, as it is recorded at Ilıpınar, Demircihüyük-Sarıket and Karataş-Semayük.
From a longue durée perspective, it is possible to conclude that a longstanding tradition of burying dead children and infants within settlements has persisted from the Neolithic to the Early Bronze Age period in Anatolia. Their dead bodies remain tied to hearths, house floors and the places for the living, which was not a deviant but an established and widely shared practice at many sites across the region. A differentiated deposition of 'ghost children', as not fully humans, from those of the adults, most likely enabled undisturbed reproduction of the house and/or the household. To fully understand people's choices to bury children and infants requires a child-centred approach. This does not equate a fully human being with a non-person, e. g. an adult and a child/infant, but considers subadult burials as possible indicators of delayed personhood. Through such an approach, we could better understand how age, age groups and childhood were understood in prehistory and build culture-specific models of infancy rather than ethnocentric interpretations of infant and child burial practices.
for support with the illustrations and to R. Rebay-Salisbury for English editing. The excavation at Çukuriçi was financed by the Austrian Science Fund (P-19856; P-25199; Y-528) and the European Research Council (263339). Authors thank B. Horejs for using data from Çukuriçi and enabling OA by the 'Prehistoric Phenomena' research group at the Dpt. of Prehistory & WANA Archaeology. This contribution was funded by the Austrian Academy of Sciences Post-Doc-Track grant (85076) and completed within the fellowship program of the OeAW at OeAI Athens (both S. Cveček).