The publication of this significant deposit in the Agora in Athens has been a long time coming: the project began 80 years ago in 1938, when the well was excavated by Dorothy Burr Thompson. The well and its contents received only cursory attention in the succeeding years and, contrary to expectation, the delay turned out to be beneficial, as changes in the field since the 1930s are what made this rigorous and interdisciplinary study possible (p. 139).
The publication itself is as significant as its subject: it represents a model for archaeological publications, careful as the authors are to evaluate and integrate multiple lines of evidence and inquiry, consider different interpretations for the formation of the deposit, and engage in differential diagnosis, as it were, for analyses and interpretations. The authors are transparent about the quality and nature of the data and they explain both where their interpretations are securely grounded in the evidence and where they are more open to alternative readings. Finally, the authors present the data and their interpretations through a combination of clear and precise prose, high quality black-and-white photographs, and well-conceived diagrams and charts, making the book more accessible to non-specialists.
The deposit that is the focus of the book includes the remains of at least 459 infants; a wealth of faunal material, including skeletons of at least 150 dogs; industrial and manufacturing debris; a wide range of ceramic vessels; an ivory chape from the scabbard of a sword; and a marble herm. It likely accumulated over a fairly short period of time, from ca. 165 to 150 BCE (p. 8), and, as the authors anticipate in the first chapter but discuss more fully in the fifth, is best understood as a burial place for newborn infants who had died naturally before being formally integrated into the family, along with a few older individuals who, for various reasons, were not granted conventional funerary rites. The stated goal of the study is not to exhaust all possibilities for explaining the deposit, “but rather to present the data in full, place it within its social and historical context, and offer a plausible suggestion about how, when, and why this remarkable assemblage of humans, animals, and artifacts came to be deposited in the well” (p. 1). In this the authors succeed.
The first chapter discusses the well and its neighborhood in 2 nd century BCE Athens. The authors detail not just the history of the Classical water system of which the well was a part, but also the history of its excavation and interpretation. Although the deposit was recorded hastily, the authors do a remarkable job reconstructing the stratigraphy, adding thoughtful visualizations (including top plans, section schematics, illustrated visualizations of depositions, and schematic diagrams) to aid the reader (especially on pp. 4, 6, and 7). They describe the nature of this area of Athens, on the northern slopes of the Kolonos Agoraios, and note that “the most important fact that emerges from study of the topography of this area is its isolation” (p. 10). The authors present the chronological development of the specific buildings associated with the well when it was in use (Buildings 3 and 4). Finally, they consider Pausanias’s description of the Sanctuary of Aphrodite Ourania and its potential relevance for the interpretation of the well.
The second chapter presents the human skeletal material. The bones “were recovered as commingled remains, together with many animal bones, and with no documentation of individual bodies or articulation of the remains” (p. 25). Nevertheless, careful analysis made it possible to isolate four individuals. An adult man (AA 24) is represented by a nearly complete skeleton that preserved evidence for a potential diagnosis of hereditary hemochromatosis. This case is an excellent introduction to the authors’ approach to the human remains, as they are explicit about how they arrived at this diagnosis and what alternatives were excluded and why. The same careful attention is found in the presentation of the other skeletons, including the older child (AA 25), who may have had brucellosis, and the two older infants, one of whom (AA 26A) unhappily presents the earliest case of child abuse reported archaeologically worldwide, while the other (AA 26B) provides evidence that considerable efforts were expended to care for children with debilitating conditions (in this case, hydrocephalus).
Finally, the authors present the 13,019 bones representing the remains of at least 459 infants and fetuses (collectively AA 26). Most of the remains range between 30 weeks in utero to 42 weeks, though the median age at death is one-week post-partum, and almost all died before they were eight days old. After a brief discussion of the potential for sexing the remains, the authors discuss identifiable pathologies and causes of death, including premature birth and low birth weight, trauma, infection and hemorrhage, developmental defects, and cleft palate. In sum, the authors argue that there is abundant reason to see the deposit largely as the result of natural mortality.
The faunal material is presented in the third chapter. As with the human remains, the animal bones are regarded as a single, undifferentiated deposit. In addition to the domestic dog assemblage were 339 non-canid animal bones, largely belonging to sheep/goats and pigs, but including three rabbit/hare bones, 51 domestic chicken bones representing at least three animals, the bones of a white stork, and, minimally, fish bones. The authors do not discuss whether selection protocols biased the sample toward larger animals, but they argue that the non-canid assemblage reflects common types of food refuse and many bear evidence of butchering (including the stork). Some larger bones of cattle, horses, and perhaps turtles represent industrial and manufacturing debris. At least seven cattle ribs and seven cattle scapulae, for example, were modified quickly to act as tools. Finally, the authors present the remains of the more than 150 dogs that were commingled with the fetal and infant human remains. The dogs, mostly adult (65-70% of the assemblage) and of a size similar to a modern beagle or small hound, appear to have entered the well whole and articulated. The evidence suggests that the dogs were likely free-roaming urban strays and the authors argue that the dogs’ sex, age, or physical attributes did not play a major role in their selection for inclusion in the deposit.
The artifacts in the assemblage are given full attention in Chapter 4. The authors begin with an informative and interesting discussion of 1930s standards of recovery, storing, and conservation before introducing the artifacts and classifying them by material and type. This section is aided by a complete catalogue at the end of the book. The ceramics are discussed according to type (fine wares and plain wares) and then subdivided according to basic functional classifications. Other clay and non-clay objects are then discussed, as well as a mass of bronze debris. This chapter is especially strong in its determination of which objects can be safely associated with the human skeletal remains and how, and which cannot. The discussion of the herm, which represents Andrew Stewart’s contribution, provides an interesting insight into the world of midwives, which receives more attention in the final chapter.
In the fifth and final chapter, the authors demonstrate the value of this study for the field as a whole: the evidence in Chapters 1 through 4 is here considered holistically, addressing how each piece fits into the larger story of pregnancy, birth, life, and death in 2 nd century BCE Athens. After discussing earlier interpretations of the deposit, the authors outline useful archaeological parallels: two wells in Eretria (one from the 3 rd century BCE, another from the Roman period) a Hellenistic well in Messene, and a Late Roman sewer in Ashkelon. They also discuss other examples of infants buried within settlements and in specialized infant cemeteries. These comparanda serve to situate the Agora Bone Well and suggest that it is not quite as unusual as it may seem; it is, instead, part of a larger and longer practice of differential treatment of the infant dead, which happened for many different reasons. The authors progress through potential interpretations for the deposit, including famine and epidemic; infanticide; and natural mortality. In this third case, they consider whether the population at Athens and, specifically, those living in the area around the well at the time, could have supported this number of natural infant deaths, as well as who may have been included in the population that contributed to the well deposit, whether citizens, slaves, metics, or visiting foreigners. They emphasize that, in most cases, no single explanation can account for a deposit like this, noting that insistence on a single cause “ignores complexities of biology and culture” (p. 138). The authors’ consideration of potential interpretations is an excellent example of academic honesty, as they embrace a complicated picture and do not erase evidence that undermines their conclusion, namely, that most of the infants died from natural causes in the first few days of life.
The authors acknowledge the possibility that some of the infants were victims of infanticide, the presentation of which is my only real quibble with the book, one that is (admittedly) likely to resound with only a small number of readers. With a brief discussion of Plato’s Theaetetus and Soranus’ Gynecology, the authors state that while we have no reason to assume that most of the infants were killed or allowed to die, “the simple fact that infanticide is attested for Greek society, coupled with the presence of a small number of infants with visible birth defects, suggests that some were victims of infanticide” (p. 125). One cannot argue that infanticide categorically is not a factor, especially because of the literary evidence for the practice. I would argue, however, that it is dangerous to pick out specific individuals from an otherwise undifferentiated collection as being the most likely victims, especially when presenting no contemporary evidence and, indeed, with definitive proof from the same deposit that extraordinary efforts were sometimes expended to care for infants with additional needs. Plato’s Theaetetus, as the authors note, does not define criteria that would have marked an infant for exposure or infanticide, while Soranus’ Gynecology is much later and better understood against a backdrop of Roman values. Moreover, one of the older infants in the deposit, AA 26B, had hydrocephalus, a condition that may or may not have been visible at birth. As the authors say, “the child was cared for during a period when it would have become progressively more debilitated and more disturbing in appearance” (p. 38). That this infant was deposited in this well suggests that it had not achieved a status that would have earned it a formal burial. The parents’ extended care may therefore indicate that, at least in some cases, infants with defects could be cared for before they had achieved that formal status. The important point is that we cannot hold up any one group of infants as likely victims without evidence and, indeed, in the face of evidence to the contrary. This has already been successfully argued for female infants but has yet to stick for infants with birth defects, who are often still considered inherently obvious candidates for infanticide. If most of the infants in this deposit died naturally, we should not identify individuals who were intentionally killed unless it can be proven, as it was for the older infant AA 26A.
This book represents an invaluable contribution to our understanding of 2 nd century BCE Athens and infant death and deviant burials in ancient Greece. It is also a demonstration of what can be accomplished when specialists work together to understand a total picture, one comprised of pieces that would have been considered separately in earlier days. The authors present an honest approach to an imperfect data set, allowing readers to assess the data and its potential interpretations for themselves. It is a welcome update to our understanding of this deposit and will not, I think, be soon superseded in its broader conclusions.