article

Mortality Crisis at Akhetaten?

Amarna and the Bioarchaeology of
the Late Bronze Age Mediterranean
Epidemic
Gretchen R. Dabbs and Anna Stevens
Open Access

The question of whether the ancient Egyptian city of Akhetaten (14th century BCE;
modern Amarna) was affected by an epidemic has long been debated. Evidence such
as the deaths of several Amarna-period royals within a short period and the identifica-
tion of Egyptian soldiers as the source of the plague that affected the Hittite empire in
Muršili II’s plague prayers has been cited to support the idea. More recently, the pres-
ence of burials containing multiple individuals in the Amarna cemeteries, high disease
frequencies, and some unusual demographic characteristics have been used to support
claims of an epidemic. This paper surveys the debate and then discusses archaeological
and demographic hallmarks of epidemics identified by others at cities with historically
documented disease outbreaks. Through this lens of expectation, we integrate archaeo-
logical and bioarchaeological data from ongoing work at Amarna’s cemeteries with per-
tinent data from long-running settlement excavations at the site to assess whether there
is empirical evidence of epidemic disease at Amarna and the wider implications of this.
We conclude that when the evidence is considered as a whole, there is little to currently
suggest Akhetaten was affected by a mortal epidemic.1

Content warning: Readers are advised that this article contains a photograph of human
remains.

introduction
The study of ancient epidemics has often privileged written source ma-
terials over archaeological evidence, and there is a pressing need to include

1
This research was conducted through the Amarna Project (British Mission to Tell
el-Amarna, McDonald Institute for Archaeological Research, University of Cambridge)
and would not have been possible without the long-term guidance of the late Barry Kemp,
who initiated the cemetery project with Jerome Rose (University of Arkansas). We thank
the Egyptian Ministry of Tourism and Antiquities for permission to undertake the field-
work, and MoTA staff in Minia, Mallawi, and Amarna for their support. The study of Amar­
na’s cemeteries has been funded by the National Endowment for the Humanities (grant
RZ-51672-14), British Academy (grant SG121253), National Geographic, King Fahd
Center for Middle East Studies (University of Arkansas), American Research Center in
American Journal of Archaeology Egypt (Antiquities Endowment Fund/USAID), McDonald Institute for Archaeological
Volume 129, Number 4 Research, Amarna Research Foundation, Aurelius Trust, Thriplow Trust, Amarna Trust,
October 2025 Institute for Bioarchaeology (British Museum), Pasold Research Fund, and Southern
Pages 455–89 ­Illinois University’s School of Anthropology, Political Science, and Sociology. Any views,
findings, conclusions, or recommendations expressed here do not necessarily represent
DOI: 10.1086/736705
those of the National Endowment for the Humanities. We also thank the international
team of researchers and support staff who have dedicated their time and expertise to the
www.ajaonline.org study of Amarna’s cemeteries.

455
456 Gretchen R. Dabbs and Anna Stevens [aja 129
a wider range of data to assess these accounts.2 The While epidemics are notoriously difficult to iden-
purported Late Bronze Age epidemic, attested mostly tify in the archaeological record, in part because they
in Hittite plague prayers, offers a case in point.3 The need not cause large numbers of deaths or create large
Hittite prayers record a disease event in the late 14th death assemblages,9 archaeological work at settlements
century BCE that is said to have claimed many lives in and cemeteries associated with known epidemic out-
the Hittite empire over at least 20 years, and they name breaks (especially from medieval Europe) is providing
Egyptian prisoners of war as its source. To date, how- a series of commonly observed hallmarks, particularly
ever, there have been few attempts to assess these writ- where catastrophic mortality has resulted. From these,
ten accounts against archaeological data,4 and many it is possible to develop a basic set of archaeological,
questions remain about the described outbreaks. Was taphonomic, and demographic features that provide a
this an extended interregional epidemic, for example, starting point to survey sites for evidence of epidemics
or a series of smaller localized outbreaks that are not where historical records are patchy (as for Amarna) or
necessarily of the same disease? Was this a time of in- entirely silent. While caution is required in transposing
creased risk of disease, or is the clustering of written disease hallmarks across temporal and cultural bound-
evidence a result of chance preservation of sources? aries, Amarna itself has unusual analytical potential for
Despite these gaps in knowledge, the events described the study of disease. The city of Akhetaten had a main
in the plague prayers are often singled out as an influ- occupation period of only around 20 years and after
ence on wider sociopolitical affairs. For Egypt, they its abandonment saw relatively little overbuilding. It is
correspond approximately with the much-discussed now one of the best-preserved cities from the ancient
reign of Akhenaten (ca. 1352–1336 BCE),5 the king world. Here, we also explore how Amarna’s short-lived
who worshiped the solar god Aten as the sole state and well-preserved cemeteries and settlement areas
deity and founded the royal city of Akhetaten (modern provide the opportunity to integrate bioarchaeological
Amarna). It is often speculated that epidemic disease and urban data in the study of past disease, and in
influenced Akhenaten’s actions and may have struck doing so further introduce data from the ancient
the city of Akhetaten.6 Over the last two decades, exca- world into scholarship on the history of epidemics. We
vations at the ruins of Akhenaten’s city at Amarna have first survey the broader historical evidence for a Late
provided an assemblage of 889 interments7 that can Bronze Age epidemic in the Mediterranean and situate
be closely dated to the mid to late 14th century BCE Amarna relative to this, before introducing the Amarna
and provide a unique opportunity to introduce bio­ burial assemblage. We then assess this assemblage, and
archaeological data into discussions of the proposed Amarna’s urban footprint more broadly, for evidence of
Late Bronze Age epidemic.8 epidemic in terms of burial practices, the construction
and occupation history of the city, and its demography
as regards both the anticipated number of dead relative

2
Kennedy 2006, 87.
3
Panagiotakopulu 2004, 273. 9
The declaration of an epidemic today does not require an
4
In part due to a lack of available Hittite burials: Gestoso empirically large number of cases. Epidemiologists define epi-
Singer 2017, 233. demics simply as the occurrence of cases of a disease in excess
5
Egyptian regnal dates after Shaw 2000. of what is normally expected in a community or region during
6
E.g., Goedicke 1984, 92; Aldred 1988, 289. a particular period of time, a definition we follow here; Bonita
7
This is the current count; ongoing study of the human et al. 2006, 119. It is the case, nonetheless, that many epidem-
remains may reveal a small number of additional interments. ics in the past did cause substantial mortality and sometimes led
8
See also Smith-Guzmán et al. (2016), who propose that to significant social change. Examples include the Black Death
endemic malaria may have been transferred beyond Egypt by (bubonic plague) in 14th-century CE Europe, which caused
former residents of Akhetaten to cause the Hittite epidemic. 30–50% depopulation and over time contributed to the end
Sabbahy (2020) suggests a malaria epidemic for Amarna itself. of the feudal system; Platt 1996, 190; DeWitte 2014, 101. The
Bioarchaeological studies of the Amarna burials early in the introduction of smallpox to the Americas in the 16th century
cemetery project revealed some demographic anomalies that CE caused the deaths of up to 50% of the population in some
pointed to the possibility of an epidemic outbreak when con- communities; Jones 2006, 2125. The 1918 influenza pandem-
sidered in isolation of other data: Rose and Zabecki 2008; 2010, ic killed 50–100 million people and impacted the progress of
412; Zabecki 2008, 66; Zabecki and Rose 2010, 6–7; Dabbs and World War I, with more American soldiers lost to disease than
Zabecki 2014, 236–38. We revisit and update these results here. to enemy fire; Johnson and Mueller 2002, 105; Byerly 2010, 82.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 457
to occupation period and who among the population The plague prayers stand out for the duration of the
are represented in its cemeteries. outbreak and the supposed interregional nature of
its transmission. Egyptian prisoners captured by the
background: late bronze age epidemics? Hittites at a battle in the northern Levant at Amka
The Hittite plague prayers of Muršili II (ca. 1330– (Beqa Valley, see fig. 1) are described as its source, al-
1295 BCE or 1321–1295 BCE),10 found as part of the though whether this was the true origin of the disease
royal archives at ancient Hattuša (fig. 1), provide the is unknown.17
most explicit written accounts of disease outbreak in Other references to outbreaks of disease in the
the late 14th century BCE. They refer to an event that 14th century BCE occur in the Amarna Letters, the
appears to have caused sustained loss of life, includ- cache of diplomatic correspondence found at Ama-
ing the deaths of Muršili’s father, Šuppiluliumas I, and rna itself. The Amarna Letters are usually thought to
his brother Arnuwanda II.11 The reigns of these kings date from the reign of Amenhotep III through to the
likely correspond, in Egypt, to the latter part of the early years of Tutankhamun (ca. 1336–1327 BCE),
reign of Akhenaten’s father, Amenhotep III (ca. 1390– Akhenaten’s likely son and eventual successor.18 They
1352 BCE), into Akhenaten’s own reign and possibly rarely mention the Egyptian king by name, making it
those of his immediate successors.12 No symptoms difficult to fine-tune the dating of events they describe.
are described or specific diseases named in the Hittite Up to five of the letters reference disease outbreaks
texts. Instead, the term henkan is used, which has the that are given particular gravity by the ancient authors,
meanings of death, death sentence, and doom, and termed mūtānu, or possibly through metaphorical
here seems to imply an epidemic-like disease;13 the phrases implying punishment from a deity, particu-
term is often translated as “plague,” to mean a large- larly “hand of Nergal,” referencing a Mesopotamian
scale disease.14 The disease itself has not yet been con- god associated with death, disease, and warfare.19 In
clusively identified,15 and it need not have been caused Amarna Letter EA35, an oft-cited example, the king
by Yersinia pestis, the bacteria that causes bubonic and of Alashiya (modern Cyprus) apologizes to the Egyp-
other plagues. Muršili addresses the storm-god:16 tian king for a copper shipment that was not as large
as intended because the “hand of Nergal” was in his
You have allowed a plague into Hatti, so that Hatti has
country.20 Megiddo, Byblos, and Sumur are other lo-
been very badly oppressed by the plague. People kept
dying in the time of my father, in the time of my brother, cations where potential outbreaks are attested in the
and since I have become priest of the gods, they keep Amarna Letters,21 while EA11, a damaged letter from
on dying in my time. For twenty years now people have Burna-Buriash II of Babylon to Akhenaten, may refer-
been dying in Hatti. Will the plague never be removed ence the death of a royal wife of Amenhotep III due
from Hatti? to a “plague.”22
In positioning these records relative to Akhetaten, it
is important to note that none speak directly of disease
at this city, and there are potential time lags between its
10
Kuhrt 1995, 30; Bryce 2005, xv.
occupation and the outbreaks mentioned. The Battle
11
The prayers are translated in Singer 2002, 47–69. of Amka, for example, the tenuous origin of the Hit-
12
On the long-standing difficulties in precisely aligning the tite outbreak, is usually thought to have occurred after
Hittite and Egyptian chronologies, see Gautschy 2014. the death of Tutankhamun in response to the death of
13
Singer 2002, 47. a Hittite prince while enroute to Egypt to marry the
14
On problems that stem from imprecise terminology in de-
scribing disease, see Alfani and Murphy 2017. These issues are
commonly encountered in studies of the ancient world where
“plague,” in particular, is often used as a general term for large-
scale disease, as here, but also in its technical sense to denote 17
Singer 2002, 47.
bacterial infection by Y. pestis, which causes bubonic, pneu- 18
Moran 1992, xxxiv–xxxix; see also our discussion be-
monic, and septicemic plagues. In this article, we use the term low of changes in settlement patterns for the post-Akhenaten
“plague” to refer to infections by Y. pestis, except where the term chronology.
is quoted directly from other sources. 19
Gestoso Singer 2017, 223–29.
15
But see the discussion in Smith-Guzmán et al. 2016, which 20
Moran 1992, 107–8.
proposes malaria as an option. 21
Gestoso Singer 2017, 229.
16
Singer 2002, 57. 22
Moran 1992, 21–23.
458 Gretchen R. Dabbs and Anna Stevens [aja 129

fig. 1. Map of the Late Bronze Age eastern Mediterranean (modified from Cohen
and Westbrook 2000, xii).

king’s widow.23 By the time of Tutankhamun’s death, Age or at other times.25 The absence of clear textual
Akhetaten had not been the seat of the royal fam- evidence could, in part, be due to Egyptian reticence
ily for several years, although it was not fully aban- in recording negative events.26 There is a body of pes-
doned.24 Geographically, the location of Akhetaten simistic literature that describes famine, disease out-
some 400 km from the Mediterranean should also be breaks, and related crises from around the end of the
noted for the possibility this provided a buffer against Old Kingdom (late third–early second millennium
diseases affecting regions to the north. Nonetheless, BCE) that is of debated reliability in terms of its his-
as a major population and political center during the torical details but which could nonetheless reflect ex-
decades in which disruptive disease events are docu- perience of such events.27 Some medical papyri also
mented internationally, Amarna is a relevant site from describe disease symptoms that may be indicative of
which to consider the archaeology of these events and bubonic plague, although the symptoms mentioned
to reflect upon their potential interregional impact. are not unique to this disease.28 There has been little
For ancient Egypt, while diseases with the poten- systematic study of Egyptian human remains from the
tial to develop into epidemics are known to have been perspective of large-scale disease outbreaks,29 or like-
present, no conclusive textual, archaeological, or pa- wise of cemetery and urban data, although at the Nile
leopathological evidence has yet been identified for Delta site of Tell el-Dab’a the co-occurrence in the early
widespread epidemic outbreaks in the Late Bronze second millennium BCE of a concentration of shallow
graves, some with multiple individuals, and a reduced
settlement area, has been provisionally attributed to

23
The events are documented in the Deeds of Šuppiluliuma.
The widowed Egyptian queen is not directly named; for the 25
David 2017, 283–84; Habicht et al. 2021, 218.
possibility she was instead the wife of the deceased Akhenaten 26
Goedicke 1984, 91–92; David 2017, 284.
or of his coregent or successor Smenkhkare, see the discussion 27
Morris 2023.
in Gautschy 2014, 151. 28
Goedicke 1984, 93–97; Nunn 1996, 75, 168; Strouhal et
24
Kemp 1995b, 446–48; Kemp and Gabolde 2022, 33–34, al. 2014, 79–80.
36, fig. 10. 29
David 2017, 284.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 459
an epidemic.30 In terms of an epidemic in the late 14th those on elite monuments and the administrative com-
century BCE, some have pointed to Amenhotep III’s munications of the Amarna Letters. The foundation
prominent worship of Sekhmet, a goddess associated texts do allude to some form of unrest that preceded
with pestilence; evidence of a fire in the tomb work- the city’s construction, in brief and broken passages of
ers’ settlement of Deir el-Medina; and Amenhotep the format “regarding [the] . . . in Akhet-Aten, it was
III’s shift of the royal residence to the site of Malqata [or, they were] worse than those which I heard in reg-
at Thebes as attempts to counteract the impacts of dis- nal year 4.”34 Akhenaten’s motives for creating Akhet-
ease outbreaks.31 All are fairly tenuous lines of thought, aten have been much discussed, and while some read
however, and are influenced by the Hittite and other these passages to suggest he founded an isolated new
written accounts noted above. city to escape disease,35 a more common interpretation
is that he was distancing himself from the religious and
epidemic and amarna: current political establishment, particularly the priesthood of
perspectives Amun at Thebes.36
Against the backdrop of the Hittite plague prayers An epidemic outbreak has also been postulated to
and related texts, Akhenaten’s atypical reign in the late explain why Akhetaten did not remain substantially
14th century BCE has prompted considerable specu- occupied after it was abandoned as the seat of the
lation about the potential influence of an epidemic on royal family in the early years of the reign of Tutankha-
his actions and on the occupation history of Akhet- mun. This scenario presupposes that an epidemic
aten. Akhetaten was the seat of royal power for about was present at the city itself. Other than preliminary
15–20 years—a phase Egyptologists term the Amarna bio­archaeological data, which are discussed further
­period—before being abandoned by the royal court below, two main lines of thought are usually put for-
and most of its population. Today, the site of Amarna ward to support this. The first is a long-standing idea in
is one of ancient Egypt’s most extensively studied cit- Egyptological scholarship that there was an unusually
ies, the main source of information on Akhenaten’s high number of deaths in the royal family during the
religious reforms and their social impacts, and is an Amar­na period.37 Three of the six known royal prin-
important case site for the archaeology of past urban cesses (Meketaten, Neferneferure, and Setepenre), one
experience more broadly.32 of Akhenaten’s secondary wives (Kiya), and Akhenat-
Located on the east bank of the Nile River in middle en’s mother (Tiye) all disappear from historical and
Egypt, Akhetaten seems to have been purposefully visual records during the latter years of Akhenaten’s
sited almost equidistant from the major urban cen- reign (beginning around Year 12), possibly indicating
ters of Memphis (modern Cairo) and Thebes (mod- their deaths.38 There is no surviving written evidence
ern Luxor) (see fig. 1). Foundation texts on rock-cut for their causes of death, or health in general, but this
boundary stelae around the city’s perimeter describe is not unusual for ancient Egypt.39 Akhenaten himself
how it was built on virgin land: “not being the property probably died in his regnal year 17. Two individuals
of a god, not being the property of a goddess.”33 This (Neferneferuaten and Smenkhkare), who were ei-
claim has been supported by excavations, with no evi- ther coregents with or successors of Akhenaten, had
dence of substantial earlier occupation encountered to very brief reigns that may have ended in death shortly
date at the site, although the wider region was settled
long before Akhenaten’s reign. Currently, there is no
known written source that speaks directly to the pres- 34
Murnane and Van Siclen 1993, 166–68.
ence of a large-scale disease event at Akhetaten, allow- 35
Goedicke 1984, 92.
ing for the fact that the site itself has not yielded large 36
E.g., Redford 1984, 158–65.
numbers or a particularly diverse range of texts beyond 37
E.g., Helck 1962, 187; Goedicke 1984, 92; Redford 1984,
187; Aldred 1988, 289; Tyldesley 1998, 150–51; Laboury 2010,
326; Dodson 2020, 65–66.
30
For preliminary publications: Bietak 1984, 334–36, pl. 38
Gabolde 1998, 122–23; Laboury 2010, 317–22; William-
14; 1991, 38. As Mourad (2021, 6) notes, conflict or wider de- son 2015, 8.
mographic change could also be contributing influences here. 39
Nor do we know the full composition of the royal family at
31
Aldred 1988, 283; Kozloff 2006, 39, 42, 44; Norrie 2016, 22. this (or any) period of Egyptian history; secondary royal wives,
32
Kemp 2012. children born to these wives, and other such figures rarely made
33
See Murnane and Van Siclen 1993, 155–56, 166–68. it into the historical or visual records.
460 Gretchen R. Dabbs and Anna Stevens [aja 129
thereafter. Some see these possible coregencies as at- know whether Tutankhamun died of this disease, and
tempts to solidify the throne in the face of a disease certainly cannot extrapolate a wider malarial epidemic
crisis affecting the royal family.40 Tutankhamun then based on this DNA result.
took the throne, dying at around age 19 after a reign of Much has also been made of Panagiotakopulu’s
nine years. In total, this sums to nine possible deaths identification of human fleas (Pulex irritans) at the
of known royal figures in a space of about 15–18 years, Amarna Workmen’s Village,48 a tomb workers’ settle-
spanning the latter part of the occupation of Akhet- ment on the city’s outskirts (fig. 2). P. irritans itself is
aten and the years after the royal family abandoned known to be a poor vector for transmitting the bubonic
the city. The idea of temporally clustered royal deaths plague (Y. pestis).49 Panagiotakopulu sought to expand
does not withstand careful examination at the finer on the preservation of fleas generally at the Work-
scale. In the context of life expectancy in the ancient men’s Village, and highlight the potential of archaeo­
world, many of these deaths also have plausible non- entomology in studying disease, to hypothesize an
epidemic causes. Queen Tiye was likely around 50 Egyptian, rather than Central Asian, origin for Y. pes-
years old41 and could have died of conditions brought tis. Y. pestis is thought to be transmitted to humans
about by old age.42 Meketaten is possibly shown hav- most efficiently by the Oriental Rat flea (Xenopsylla
ing died in childbirth in a very rare representation of cheopis),50 whose main host is the black rat. Panagio-
a royal mourning scene,43 while the other two royal takopulu suggested that Y. pestis may have begun as a
daughters, Neferneferure and Setepenre, were both disease of the Nile rat, and been transmitted from it
probably around 5 years of age or under when they to the black rat by X. cheopis at some point during the
died.44 Their deaths are not unexpected in the con- pharaonic period, shifting from a disease that was rare
text of preindustrial life histories: approximately 30% in humans to one that had the opportunity to become
of children born in the premodern period died before epidemic.51 She noted that the Workmen’s Village,
the end of infancy (birth–5 years).45 DNA analysis with its wider evidence of ectoparasites (human fleas,
of Tutankhamun’s mummy indicated he carried the bedbugs) and other insect remains suggestive of unhy-
malarial parasite,46 and while several strains of malaria gienic living conditions, was potentially representative
can cause mortal and morbid epidemics, the fact that
malaria is documented in Egypt from the early third
millennium BCE onward, and in human remains from Egypt: Bianucci et al. 2008; 2015; Nerlich et al. 2008; Smith-
varying socioeconomic levels, suggests a sublethal Guzmán et al. 2016; Mitchell 2024. When diseases become
endemic in a population, there are often cultural and biologi-
strain was endemic by the Amarna period.47 We cannot
cal adaptations to their presence; Silk and Fefferman 2021. See
extensive literature on balanced polymorphisms, such as the
presence of malaria and a number of genetic anemias, sickle-cell
40
Dodson 2018, 39–40. anemia and thalassemia primary among them.
41
Hawass et al. 2010, 640, table 1. 48
Panagiotakopulu 2001; 2004.
42
A small number of royal mummies of the Amarna period 49
Both between humans and from rodents to humans: Mi-
do survive for study, including Queen Tiye’s. No conclusive arinjara et al. 2021, 13. Some have suggested P. irritans may still
statements, ancient or modern, about her cause of death are contribute to large-scale plague outbreaks, arguing large num-
available. Hawass and colleagues (2010, 643) did not find mo- bers can compensate for low transmissibility; Miarinjara et al.
lecular evidence of tuberculosis, bubonic plague, leprosy, leish- 2021, 13, and sources cited; but cf. Dean et al. 2018. It is difficult
maniasis, or malaria in the mummy commonly assigned to her, at present to know how prevalent P. irritans was at Amarna. At
suggesting none caused her death. While only noted in a table the Workmen’s Village, they were attested in what is described
without further comment, the morphological analysis via CT as “high frequency” (35 specimens) across the small number of
scan revealed struma (enlarged thyroid), which is linked with a samples studied; Panagiotakopulu 2001; 2004, 274. None were
number of potentially fatal conditions such as thyroid cancer, noted in samples taken during excavations of nine houses in the
thyroid storm, and untreated Hashimoto’s disease; Hawass et Main City at Amarna (Panagiotakopulu 2001, 499; 2004, 270;
al. 2010, eAppendix pg. 2. Panagiotakopulu and Buckland 2010), but the deposits here
43
Martin 1989, 37–48; for alternative readings of the scene: are not as dry as at the Workmen’s Village, which might affect
Gabolde 1998, 120–24. preservation. One individual cat flea (Ctenocephclides felis) was
44
Gabolde 1998, 123. also identified at the Workmen’s Village (Panagiotakopulu 2004,
45
For extensive reviews on paleodemography: Scott and Dun- 270), but this species has long been dismissed as not a transmit-
can 1998; Chamberlain 2006. ter of Y. pestis; Eisen et al. 2008; Bland and Hinnebusch 2016, 1.
46
Hawass et al. 2010, 643–44. 50
Eisen and Gage 2012, 67–68.
47
For more on the frequency of presence of malaria in ancient 51
Panagiotakopulu 2004, 272–73.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 461

fig. 2. Map of Amarna showing the location of the city’s cemeteries and other key sites (courtesy the
Amarna Project).
462 Gretchen R. Dabbs and Anna Stevens [aja 129
of the kind of settlement where epidemics may have new royal cemetery in a wadi (dried water course) in
taken hold in populations. As yet, however, while the the eastern cliffs, and many of the city’s officials had
Nile rat, black rat, and X. cheopis were all potentially elaborate rock-cut tombs carved in the cliff face (the
present in Egypt in the Amarna period,52 they have not North and South Tombs). The Workmen’s and Stone
been identified at the Workmen’s Village or at Amarna Villages had their own small cemeteries.58 Only a few
generally.53 Panagiotakopulu’s work has been cited as human remains survive for study from the royal and
evidence for plague at Amarna itself,54 but it is impor- elite tombs and from the village cemeteries. The city’s
tant to note that she only explores here a potential main cemeteries were located on the desert outskirts
mechanism for the transmission of Y. pestis and does near the two groups of officials’ tombs. Here, there
not provide direct evidence for Y. pestis at Amarna. were four large burial grounds for the wider popula-
For now, while Y. pestis has been found in third mil- tion of the city. It is these cemeteries, excavated from
lennium BCE contexts in Eurasia, there is no direct 2005 to 2022,59 that are the focus of our discussion
evidence of it in Africa prior to the Plague of Justinian here.60 One lies to the southeast of the city (the South
(ca. 541 CE).55 Tombs Cemetery) and the other three are located to its
northeast (the North Cliffs, North Desert, and North
amarna, its burials, and bioarchaeology Tombs Cemeteries). We estimate there are around
Amarna is a large and diverse archaeological site, 11,350–12,950 people buried across these four sites.
with a long excavation history.56 The remains of the While all the cemeteries have been disturbed by loot-
ancient city, its buildings made mostly of mudbrick, ing, which complicates excavation and analysis, they
lie on the east bank of the Nile River, framed by the have produced one of the largest and most systemati-
cliffs of the river valley (see fig. 2). The major temples cally recorded assemblages of cemetery data (889 in-
and palaces were located in the Central City. The terments) yet available from ancient Egypt.
main housing areas occupied three zones (the Main While each cemetery is distinctive in its burial pro-
City, North Suburb, and North City), with the Work- file, the majority of interments are pit graves contain-
men’s Village and a second peripheral settlement (the ing a single individual, the bodies surviving mostly
Stone Village) in the desert east of the city. Occasional in skeletal form. No evidence for embalming has yet
burials have been encountered within the settlement been encountered.61 The bodies were usually wrapped
areas, probably dating to the very earliest and latest in textiles and then rolled in a mat of plant material, or
years of the city’s occupation.57 Akhenaten founded a sometimes placed in a wooden, pottery, or mud cof-
fin.62 The graves were occasionally marked by a stele
which, to judge from the best-preserved examples,
52
Panagiotakopulu 2004, 271, table 1. showed the deceased receiving offerings.63 Burial
53
A situation that may change as more systematic study of objects are not common but include pottery vessels
rodent bones is undertaken; see Panagiotakopulu 2004, 273. (sometimes containing food), jewelry, and cosmetic
54
Kozloff 2006, 45; Gestoso Singer 2017, 237 (but cf. Ges- and personal items (e.g., eye-paint sets, a mirror, a
toso Singer 2023, 61); Jiménez Meroño 2024, 116.
55
Green 2018, 14. Publication of results preliminarily shared
walking staff). Other than occasional inscriptions on
at a recent Paleopathology Association conference may belie this
statement, as Maixner and colleagues (2024) reported finding
molecular evidence of Y. pestis in an unprovenanced Egyptian only patchily recorded and studied; see, e.g., Griffith 1924, 302;
mummy with radiocarbon dates associated with the Second Stevens 2018, 113–16; Nicholson 2023.
Intermediate period / early New Kingdom (1687–1449 BCE). 58
Dolling 2012, 384–411; Stevens and Rose 2020.
If Y. pestis was present in pharaonic Egypt, ongoing genomic re- 59
See, e.g. Kemp et al. 2013; Stevens 2018; Stevens et al. 2023.
search suggests it may not have been as deadly as later outbreaks: 60
The authors codirect the cemetery project at Amarna
the 1.ANT strand of the Y. pestis pathogen, which is today the (Dabbs: bioarchaeology; Stevens: archaeology). Analysis of
most widely recognized, globally dispersed, and deadly of the the skeletal and archaeological remains is ongoing in advance
strains, seems to have only emerged just prior to the Second of final publication, but it is not anticipated that the arguments
Plague Pandemic of 1142–1339 CE; Green 2018, 7. For more presented here will change substantially going forward.
on the important genetic differences between extant and an- 61
I.e., the ethmoid is not systematically fractured, no incisions
cient Bronze Age variants of Y. pestis, see Light-Maka et al. 2025. are visible in preserved soft tissue, no natron has been identified,
56
Kemp 2012. and resinlike substances are rare; Dabbs et al. 2022, 3.
57
Numbering perhaps in the tens to low hundreds, they are 62
Stevens 2018, 106.
an important part of the burial landscape of Amarna but are 63
Kemp et al. 2013, fig. 4; Stevens 2018, fig. 7.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 463
coffins, stelae, and small amulets, there is very little disease progression and its documented presence in
written material and limited direct evidence of per- Egypt from the Predynastic period onward suggest it
sonal details such as names, occupations, and so on. may have been endemic.
Bioarchaeologically, each cemetery is distinct in
certain ways, although they share many commonali- South Tombs Cemetery
ties that reflect living conditions at the city, and many This is the largest of the cemeteries, located in a wadi
variations between them can probably be attributed adjacent to the South Tombs (see figs. 2, 3). It was ex-
to economic and social differences between sectors cavated from 2005 to 2013, with 381 graves excavated
of the urban population. Overall, the individuals bur- and full or partial skeletons of a likely 494 individuals
ied in the four cemeteries exhibit skeletal markers recovered. We currently estimate, by projecting the
of substantial stress loads, such as low adult stature; average number of interments in an excavation square
high frequency of linear enamel hypoplasia; extreme across the approximate area of the cemetery, that
trauma, particularly of the vertebral column;64 and around 6,000 individuals were buried here. Most buri-
high frequency and sometimes severity of degenera- als are single pit graves where the deceased is wrapped
tive joint disease.65 in textiles and matting. The excavated graves include
Few diseases are specifically identifiable in the skele- a single mudbrick-lined tomb, and produced only 40
tal record. Leprosy, some treponemal diseases (syphilis wooden coffins and up to 17 grave stelae or miniature
and yaws), brucellosis, and tuberculosis can be iden- pyramid stones, suggesting the rarity of these embel-
tified from the morphology of lesions they cause and lishments. In terms of the burial population, the exca-
distribution of those lesions in the skeletal system.66 vated sample contained six fetuses (1.2%), 183 sets of
Importantly, influenza, cholera, typhus, schistosomia- subadult remains (birth–14.9 years; 37.0%), and 305
sis, guinea worms, and other common environmentally adult individuals (15 years and over; 61.7%). The sex
driven diseases do not leave skeletal lesions, because ratio for adults reveals slightly more females (n = 157;
the disease progression is rapid and typically fatal or 59.0%) than males (n = 109; 41.0%), where sex can be
recovered from before skeletal response is possible, estimated.69 The full demographic profile is presented
or any skeletal response is nonspecific. The identifi- below,70 but the general pattern is one of overrepre-
cation of intestinal parasites may in the future be pos- sentation of older children and young adults, and an
sible from the matrix samples that have been collected underrepresentation of infants and older adults. The
from abdominal areas of some in situ burials. While analysis of the South Tombs Cemetery is the most ad-
there are several nonspecific indicators of heavy stress vanced of the four cemeteries, and it currently offers
loads,67 to date the only identifiable infectious disease the greatest scope for statistical analysis of the burials
processes observed in the skeletal remains is tuberculo- and human remains in terms of disease and its impacts.
sis, present in one individual of the South Tombs Cem-
etery sample,68 and possibly six additional individuals North Tombs Cemetery
from the North Desert Cemetery. While tuberculosis This cemetery is located in a wadi adjacent to the
has the capacity to become an epidemic disease, slow North Tombs. It was excavated in 2015 and 2017, with
141 graves excavated and 252 individuals recovered.
It is the most divergent of the Amarna cemeteries, in
64
Spondylolysis, Schmorl’s nodes, and vertebral compression part due its large proportion of multiple burials (about
fractures are the most commonly observed vertebral fractures. 50% of sampled graves, as discussed further below).
65
Kemp et al. 2013; Dabbs et al. 2015; Dabbs 2019; Stevens
While textile and matting body wrappings were com-
et al. 2020.
66
Lesions that could indicate malaria have also been iden- mon, burial objects were rare, and no wooden coffins
tified; Smith-Guzmán 2015. Malaria identification, however, or stelae were found. The variable number of individu-
relies on features of anemic response. Anemia is a symptom of als in the multiple burials makes it difficult to calculate
more than 200 conditions, and as yet it has not been reliably an overall population, but a preliminary estimate is
proven that any one of the other 200 conditions cannot also
cause the observed lesion sequence.
67
E.g., low adult stature, high linear enamel frequency, high
frequency of anemia;see Dabbs et al. 2015, Dabbs 2019, Stevens 69
For the methods used in estimating age and sex in the
and Dabbs forthcoming. Amar­na samples, see Dabbs 2019, 178–79.
68
Dabbs 2021. 70
See under “Observed Demography.”
464 Gretchen R. Dabbs and Anna Stevens [aja 129

fig. 3. A view across the Amarna excavations at the South Tombs Cemetery in 2010, facing southeast (courtesy Gwil Owen
and the Amarna Project).

4,000–5,000. The distinctive nature of the cemetery the sex ratio would still be 1.6 (more than three females
extends to the skeletal remains, with the excavated for every two males).
sample displaying a very restricted demography: 96.4%
of individuals for whom a precise age estimate was North Cliffs Cemetery
possible died between the ages of 5 and 24.9 years, This burial ground is situated on the low desert
with peak mortality at ages 12 and 17 years.71 No fetal below the tomb of one of the city’s priests, Panehesy.
remains were recovered and only four individuals less One excavation season was conducted here in 2018,
than 5 years old were identified in the sample. Only with 72 likely graves excavated and 57 individuals re-
two individuals older than 35 years were present. The covered. The sampled burials were mostly single inter-
demographic curve is inverted from the expected, with ments, with occasional wooden coffins and stelae, and
a large number of deaths of older subadults and young a range of burial objects similar to the South Tombs
adults and almost no young or old individuals. There is Cemetery. This site was particularly disturbed by loot-
a dramatic difference in the adult (n = 120; 47.6%) sex ers, which again creates difficulties in estimating the
distribution here compared with the other cemeteries. overall cemetery size, but there were probably around
Of adults for whom a sex estimate is possible (n = 96), 900–1,400 individuals in total buried here. There are
females (n = 73; 76.0%) outnumber males (n = 23; 18 subadults (36.7%) in the excavated sample and 39
24.0%) by a factor of 3.2 (more than three females for adults (63.3%). The sex distribution of adult indi-
each male). The reason for this is not yet clear. Given viduals is relatively equal, with 18 (54.5%) identified
the young age of the skeletal sample, it may in part be as female, 15 as male (45.5%), and the remaining six
due to difficulties in identifying young adult males of indeterminate sex. All age groups are represented
from skeletal remains, although even if all the individu- except fetuses. Broadly, the demography, which is dis-
als of uncertain sex were assumed to be male (n = 24), cussed more fully below, is similar to that of the South
Tombs Cemetery.

71
Dabbs 2019, 179; and discussed below.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 465
North Desert Cemetery to demonstrate high frequency of pathogen presence.73
This cemetery is located southwest of the North As outlined above, there is no direct written evidence
Cliffs Cemetery. Excavations were conducted here in for an epidemic at Amarna, and to date it has not been
2018, 2021, and 2022; its burial population probably possible to undertake DNA analysis of skeletal material
numbers around 450–550. The cemetery encompasses from the site. Bioarchaeologically, the identification
a low escarpment of claylike rock and is distinctive in of an epidemic disease is not as simple as identify-
its range of grave architecture. Pit graves occur on the ing skeletal lesions of a particular disease on a large
desert floor and along the edges of the escarpment, proportion of individuals. With few exceptions, the
while at least five much larger and more elaborate indicators of life events reflected on skeletons cannot
shaft-and-chamber tombs are located on and around be temporally correlated with lifespan, thus it is usu-
the surface of the escarpment. The pit-grave burials ally impossible to know when within their lifespan an
are again of the standard type, with the body usually individual was affected by a particular event. Further,
placed singly and supine, in textiles and matting, with even if the condition is identifiable as an active process
only a few examples of burial items and wooden cof- at the time of death, in cases such as Amarna, where
fins. Three of the shaft-and-chamber tombs were fully the chronology of interments is impossible to identify
excavated. Two were double-chambered with brick- beyond the broader context of the period, there is no
lined shafts 3.0 and 3.8 m deep. Both were heavily guarantee that any two individuals who exhibit skel-
robbed, but in combination produced disarticulated etal lesions of a condition experienced that condition
skeletal elements from a minimum of 10 individu- simultaneously.
als (subadults, adult males, and adult females), and Epidemic diseases are most commonly virulent bac-
painted coffin fragments. The third tomb was smaller terial or viral infections that tend to result in death or
but also had two chambers: one contained at least six recovery relatively quickly (days or weeks), and skele-
intact or partly disturbed interments, while the other tal response to disease is often slow acting, nonspecific,
preserved some small fragments of a painted wooden and, occasionally, age dependent. It generally results
coffin. in lesions only after weeks to months of immune re-
The bioarchaeological analysis of this cemetery is sponse. Skeletal assemblages are then comprised of
ongoing. A minimum of 86 individuals was excavated three groups with respect to a particular disease: indi-
from the cemetery overall. Preservation is generally viduals who never contracted it, those who contracted
poorer than found in the other three cemeteries, and it but died (not necessarily as a result of the disease)
a consequence is a higher proportion of individuals before skeletal lesions could form, and those who con-
identified only by the basic demographic profile.72 tracted the disease and survived long enough to form
One fetus was identified (1.2%). Of the remaining in- skeletal lesions. Analytically, the first two categories
dividuals, 35 (40.7%) are subadults and 50 (58.1%) are indistinguishable, while individuals with lesions
are adults. The demography is discussed below, but indicative of disease processes may have been able to
generally this cemetery most closely approximates resist the disease long enough for lesions to form be-
the expected U-shaped demographic curve of prein- cause they were generally healthier.74
dustrial sites, although like other Amarna cemeteries While paleopathological approaches to epidemic are
there is an underrepresentation of infants (0–5 years). limited when used in isolation, research into cemeter-
Of the adults studied to date, when a sex estimate was ies and urban environments associated with known
possible, the sex ratio is even (22 males; 22 females). epidemic outbreaks (particularly those with cata-
strophic mortality) is helping to isolate features that
epidemic and amarna: perspectives from
the cemetery study
Epidemics are extremely difficult to identify at ar- 73
Even the presence of pathogen DNA would not necessi-
chaeological sites in the absence of historical docu- tate the presence of an epidemic disease, as not all individuals
mentation or molecular (i.e., ancient DNA) analyses who carry a pathogen will develop symptoms, and the degree
to which a single individual within a population would be likely
to develop symptoms given pathogen exposure is an example
of hidden heterogeneity of risk. This is part of a larger concept
known as the osteological paradox; see Wood et al. 1992.
72
E.g., adult female, instead of 25–35-year-old female. 74
Again, part of the osteological paradox; Wood et al. 1992.
466 Gretchen R. Dabbs and Anna Stevens [aja 129
may be traceable in wider archaeological and bio­ need not reflect death rates alone but could also indi-
archaeological data. In terms of settlement patterns, cate varying levels of mobility in the face of epidemic
for example, epidemic events have been shown to outbreaks (e.g., 15–25-year-olds moving away from
sometimes result in the rapid abandonment of sites affected areas).83
with previously long and stable occupation histories, Many of these hallmarks are of course not unique
the relocation of populations from cities out to re- to epidemics. Famine, war, genocide, mass human sac-
gional areas, or a dearth of newly established towns.75 rifice, natural disaster, and large-scale migration may
Changes to building programs can also occur, which result in similar outcomes, and specific causes will not
may manifest as aborted or delayed construction proj- necessarily be identifiable from the available data. For
ects or neglected maintenance that leads to premature the purposes of this discussion, it can be noted that
decay of existing structures.76 As regards burial prac- there is no written or artistic evidence in this period
tice, the presence of mass graves is often considered a of Egyptian history for internal warfare, genocide,
hallmark of epidemic cemeteries, although historians natural disaster, or human sacrifice.84 The Amarna
and archaeologists also acknowledge that most buri- skeletal sample does show evidence of nutritional
als during epidemics were carried out in the culturally insufficiency,85 and while we cannot completely rule
normative way, which typically means a single burial out widespread food shortage, the assemblage does
within an established cemetery.77 Mass burials seem not reflect the typical demography of famine-affected
to be limited to the periods when the death toll over- populations.86 Nor is famine directly attested in his-
whelms the typical mortuary systems.78 What seems to torical records related to Akhetaten.87 If, therefore, the
matter most in identifying an epidemic mortality crisis, hallmarks outlined above are present in the Amarna
therefore, is not the number of individuals in a grave data, disease and migration remain as the most imme-
but deviation from prescribed cultural norms of the diate causes for consideration. Migration is certainly
population. Gallagher and Dueppen suggest evidence an important component of the story of Akhetaten,88
of decomposition before burial as an indicator of an with both its foundation and main abandonment
epidemic mortality crisis,79 which is consistent with forming substantial migration events, and a factor we
reports from plagues where delayed burial resulted.80 return to below. Next, we survey the archaeological
Finally, in addition to evidence of rapid population and bioarchaeological evidence from Amarna against
loss from death,81 indiscriminate catastrophic mor- these potential epidemic hallmarks, through four main
tality events in preindustrial populations have been lines of inquiry: ancient burial practices, settlement
proposed to share three demographic characteristics, and construction patterns, whether rapid population
based on research from a known Black Death cemetery
in London: a relatively low proportion of infants (0–5
years old), a low proportion of 15–25-year-olds, and a 83
Margerison and Knüsel 2002, 139.
peak in the 25–35-years-old adult age group.82 These 84
Nor are such factors visible in the bioarchaeological record
at Amarna. Each would likely result in perimortem (around the
time of death) traumatic injury on a potentially large scale, in-
cluding at least some portion that would have impacted hard
75
Kennedy 2006, 89; Chouin 2018, 5–6. tissues and been detectable through bioarchaeological study.
76
Kennedy 2006, 89; Chouin 2018, 5. While traumatic injury affecting the skeletal system is a fre-
77
Castex 2008, 28–29; Chouin 2018, 12. quently observed condition in the Amarna skeletal samples
78
Brzobohatá et al. 2019, 79. During the Great Plague of there is little evidence of perimortem trauma or of interpersonal
London, e.g., undertakers were able to keep up with burials of violence in general; Kemp et al. 2013; Dabbs et al. 2015; Stevens
the dead for all but the deadliest six or seven weeks in summer et al. 2018, 131–38; Dabbs 2019.
1665, during which time bodies could be seen on the streets; 85
Rose 2006; Rose and Zabecki 2010; Kemp et al. 2013;
Sloane 1973, 274. Dabbs et al. 2015; Dabbs 2019.
79
Gallagher and Dueppen 2018, 5. 86
The greatest impact of famine is seen among the very
80
Such as during the Plague of Athens, when Thucydides young, and females have an approximately 40% mortality ad-
described bodies lying around unburied and fed on by animals; vantage over males; Speakman 2013, 832–33.
Littman 2009, 460. Similar conditions were recorded during the 87
Its occupation predates the climatic downturn that may
Justinian Plague and the Black Death; Gallagher and Dueppen have spurred wider regional famines beginning ca. 1200 BCE;
2008, 6; Green 2018, 14. Butzer 2012, 3634–35.
81
Gallagher and Dueppen 2018, 5. 88
Often overlooked as a potential cause of change in archae-
82
Margerison and Knüsel 2002, 141. ology; Anthony 1990.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 467
loss from death can be identified, and wider demo- where this cannot be explained by potential tapho-
graphic patterns. nomic factors. If victims of an epidemic were among
the dead, there is little sign that they were denied this
Modification to Normative Burial Practices standard treatment, although further analysis—to con-
At a broad organizational level, there is little obvious sider the use of reduced or lower-quality wrappings
sign of divergent practice in the treatment of the dead and containers—is hampered by the generally poor
at Amarna, such as the physical isolation of a portion of preservation of these materials.
the dead in one or more epidemic cemeteries; at least, Multiple Burial. The most obvious divergence from
none of the burial grounds is overtly separated or hid- the standard burial practice noted above are the cases
den. In contrast, they seem to fit fairly cohesively into of multiple burials in a single grave. These occur, with
the urban landscape, lying on the fringes of the city varying frequency, at all four cemeteries, across pit
adjacent to the main suburban areas and in the vicin- graves and in the shaft-and-chamber tombs at the
ity of the rock-cut tombs of officials (see fig. 2). The North Desert Cemetery, and in particularly striking
South Tombs Cemetery seems well positioned to have proportions at the North Tombs Cemetery (table 1).
served the Main City, and the northern cemeteries may In total, approximately 304 of the 889 (34.2%) recov-
have served, at least in part, populations in the city’s ered individuals from the cemeteries as a whole were
northern zone. It is not yet clear why there are three buried with at least one other person.
separate burial grounds here, but there is a range of po- Multiple burials were not uncommon in this period
tential explanations related to community occupation, of Egyptian history, associated mostly with tombs
organization, and socioeconomic standing.89 proper rather than pit graves (although the latter have
Broadly, the predominance of pit graves and ap- been less systematically studied). Burial with family
parent absence of embalming could be put forth as members was one likely motivation behind the prac-
evidence for expedient practice resulting from large- tice.93 The shaft-and-chamber tombs at the North
scale disruption.90 Yet pit graves are so widespread at Desert Cemetery may have served as family tombs,
Egyptian cemeteries, and embalming and mummifica- and presumably also as status markers given their
tion practices of the general population are so poorly larger and more elaborate form.94 There is nothing to
studied,91 that it would be premature to attribute the suggest they were anomalous burials associated with
large proportion of pit graves relative to rock-cut a mortality crisis. The pit graves with multiple buri-
tombs, and the observed simplicity of burial treatment, als are somewhat more difficult to interpret. These
to a mortality crisis. Further, in warm to hot environ- graves generally contain just two or three individuals:
ments, expedience in burial practice may be in part a there are no very large-scale burial pits that more cer-
practical consideration, disposing of the dead before tainly indicate mass casualties. The immediate ques-
active decomposition begins. More specifically, at tion is whether the individuals were buried together
Amarna a single interment with the deceased extended simultaneously. It is impossible to be sure, but this
supine and wrapped in textiles and a more rigid con- may often have been the case. While tombs may have
tainer, usually of matting, is clearly the baseline burial been reopened to accommodate new burials with
treatment.92 What stands out at all four cemeteries is relative ease,95 this is more difficult to envisage with
the nearly universal use of both textiles and an outer the pit graves, where the body and burial objects were
covering, like a matting coffin. There are burials where
one or both of these elements is absent, but few cases
93
Baines and Lacovara 2002, 11; Cooney 2011, 18; Olabarria
2020, 87. The practice increased into the first millennium BCE,
89
See Stevens and Dabbs 2017, 148; Stevens 2018. with economic and political instability a growing influence; Tay-
90
In the pessimistic literature noted earlier (supra n. 27), lor 2010; Cooney 2011.
burial systems are presented as being so overwhelmed by trig- 94
Other potential markers of higher status are found in the
gers like famine or disease that people who could have expected surviving burial equipment and objects from these tombs, which
to be embalmed instead received expedient interment, perhaps include fragments of painted coffins and items of gold jewelry;
in desert pit graves; Enmarch 2008, 96, 107, 122. Stevens et al. 2023, 101, fig. 15.
91
Grajetzki 2021, 26, 35–36. 95
Miniaci 2019, 293–95. One of the North Desert Cemetery
92
An arrangement also widely represented elsewhere in tombs has a staircase that may have facilitated this; Stevens et
Egypt; Grajetzki 2021, 29. al. 2022, 14, figs. 6, 7.
468 Gretchen R. Dabbs and Anna Stevens [aja 129
table 1. Numbers of multiple burials and single burials found in pit graves in the four main Amarna cemeteries. The
shaft-and-chamber tombs at the North Desert Cemetery are excluded.
Graves with
preserved human Likely multiple Likely single burials
Cemetery Excavated graves remains burials (freq.) (freq.)
South Tombs Cemetery 381 359 46 (0.128) 313 (0.872)
Wadi Mouth 79 77 3 (0.039) 74 (0.961)
Lower Site 126 118 7 (0.059) 111 (0.941)
Middle Site 32 32 4 (0.125) 28 (0.875)
Upper Site 136 124 32 (0.258) 92 (0.742)
Wadi End 8 8 0 8 (1.000)
North Tombs Cemetery 145 138 69 (0.500) 69 (0.500)
North Cliffs Cemetery 72 46 1 (0.022) 45 (0.978)
North Desert Cemetery 67 57 4 (0.070) 53 (0.930)

c­ overed with sand. In these multiple burials, the burial usually be determined with reasonable confidence.
containers are often closely and neatly associated with Most (87.0%) contained two or three individuals,
one another in a way that also seems consistent with but there are some larger groupings—the maximum
their having been placed in the grave at the same time. encountered to date is seven. The multiple burials
We should not rule out the possibility that pit graves vary in concentration across the North Tombs Cem-
were sometimes reopened, particularly if professional etery, but are widespread here as a whole.97 These
gravediggers were engaged.96 Overall, however, many multiple burials differ somewhat from those at the
people in the multiple-burial pit graves seem likely to other Amar­na cemeteries, not just in their frequency
have been interred at the same time. Because of the dif- and the numbers of individuals they contain but also
ficulty of preserving bodies in the Egyptian environ- in the morphology of the graves and arrangement of
ment, and the absence of embalming, they probably individuals. At the other cemeteries, graves with mul-
died close to one another in time. tiple individuals are usually cut sufficiently wide so
Multiple burials are by far most prevalent at the that the individuals can be placed side-by-side. At the
North Tombs Cemetery, where they comprise half of North Tombs Cemetery, the graves with multiple buri-
the total interments (see table 1). While disturbance als are often cut only slightly wider than single burials,
from looting often makes it difficult to reconstruct and the bodies are tightly packed into the grave (fig. 4)
these burials, the numbers of individuals in each can or placed on top of each other in layers.98 Textile body
wrappings and matting coffins are poorly preserved at
this site overall, but where these survive in the multi-
96
The pit graves are often cut with a level of precision that
ple burials it is clear that, while some individuals were
suggests the work of specialist gravediggers; Stevens 2018, wrapped separately, there are also cases where multiple
117. Occasionally, graves containing multiple burials also have bodies were wrapped together in textiles and/or mat-
slightly irregular edges that might indicate recutting to enlarge ting. In these instances, there can be little doubt that
the grave, although disturbance from looters limits scope for the individuals died close to one another in time. The
observations of this kind. One grave at the North Tombs Cem-
other unusual feature of the graves here, seen across
etery contained five individuals, each separated by a layer of
sand about 5 cm deep, which potentially represent time lapses examples of both multiple and single interments, is
between each interment; Stevens and Dabbs 2017, 138. The that they are often cut noticeably longer than the body
practice of reopening burials generally, in this case to reinter the
deceased, seems to be attested in one unusual multiple burial at
the South Tombs Cemetery, which contained a reconstructed 97
See the preliminary distribution maps in Stevens and Dabbs
skeleton, padded out with mudbrick and matting as thought it 2017, figs. 3, 4, 6.
had been exhumed and reinterred; Stevens 2018, 115, fig. 13. 98
See Stevens 2018, 112, fig. 9.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 469

fig. 4. An undisturbed burial at the North Tombs Cemetery containing three individuals (Inds. 1167, 1168, and 1169), who
were wrapped together in a single burial mat before interment. The matting survived in patches on the sides and base of the
grave and is visible beside the skull and right forearm of the individual to the right (courtesy the Amarna Project).

or coffin itself. This raises the possibility that grave­ in overload and death.99 Where they lived is not clear.
diggers were preparing burial pits in advance of the If they were housed separately, there are no known
bodies being delivered, not knowing the exact size of workers’ settlements large enough to have created
the coffin; if so, there is no way of knowing how many such an extensive burial assemblage,100 although they
pits might have been prepared at one time. perhaps lived in simple encampments that have not
Archaeologically, the multiple burials at the North survived.101 Few parallels to the North Tombs Cem-
Tombs Cemetery provide fairly compelling evidence etery can yet be cited from other published Egyptian
for a mortality crisis affecting at least a portion of the cemeteries in terms of its restricted demography and
Amarna population, although they do not inform di- proportion of multiple burials. If this is not just a mat-
rectly on its cause. Another distinctive aspect of this ter of preservation or sampling, it raises the question
cemetery, noted earlier, is its very restricted age profile, of whether something unusual occurred at Akhetaten
with most of the individuals buried here aged 5–24.9 to create this burial profile—such as pressure to build
years. In interpreting this site, it is necessary to ask Akhenaten’s new city, an event such as a mass disease
whether it served the general population living in the outbreak, or multiple such stressors. We return later
regular housing suburbs, or might have served one or to the interpretation of this cemetery from a demo-
more communities that lived separate from the wider graphic perspective.
population. If the former, the multiple burials and re- At the other Amarna cemeteries, multiple burials
stricted age profile could imply that there was a mortal- are less common (see table 1) and, as noted, usually
ity crisis affecting children and young adults generally, contained bodies laid side by side in separate burial
and the North Tombs Cemetery served as a dedicated, containers,102 although some stacked interments were
or an overflow, burial ground for those impacted. For
now, however, it can be noted that the individuals bur-
ied at this site also show signs overall of extreme work- 99
Stevens and Dabbs 2017; Dabbs 2019.
loads. Because of this, our preliminary interpretation is 100
The average population size necessary to create this sized
that this cemetery served young laborers functioning burial assemblage is 4,800–8,000, depending on the temporal-
under near-maximum threshold stressor levels, and the ity of the population (15–20 years): Acsádi and Nemeskéri
1970, 65–66.
addition of either psychological or physiological stress- 101
Stevens 2018, 120.
ors (one of which could be disease), or both, resulted 102
See, e.g., Dolling 2008, 17, fig. 7.
470 Gretchen R. Dabbs and Anna Stevens [aja 129
found.103 The fairly low numbers of multiple burials and sex, and the combinations of people in one burial.
in the sampled areas of these cemeteries do not im- Some reasonably clear patterns emerge, which hold
mediately suggest a mortality crisis, although one of across the cemeteries as a whole and when the sites
the excavation areas at the South Tombs Cemetery are considered individually.108 The sample of individu-
(the Upper Site; fig. 5) does contain a slight cluster- als buried with others is roughly evenly split between
ing of multiple burials, which has been commented on adults and subadults (table 2). Ignoring the pattern-
previously in the context of disease rates.104 Here, the ing of burials and considering only the number of
proportion of multiple burials is 26%, compared to 6% individuals interred in single- versus multiple-burial
averaged across the other four areas excavated at this graves yields interesting results (table 3). It is clear
cemetery (Wadi Mouth, Lower, Middle, and Wadi End the frequency of multiple burial is greater in the North
sites). The South Tombs Cemetery is located in a fairly Tombs Cemetery, where about 73% of individuals are
narrow wadi, and the Upper Site lies about 500 m in- found in such graves,109 compared to about 23% in the
side the wadi. Smith-Guzmán et al. posit that the earli- South Tombs Cemetery.110 The difference is statisti-
est burials were placed near the mouth of the wadi and cally significant (p <0.001).111 At the North Tombs
the cemetery then gradually spread along its length.105 Cemetery, the frequency of multiple burials is consis-
They observe that the numbers of multiple buri- tently high (60–78%) across demographic groups,112
als increase along the wadi up to the Upper Site and and the differences between demographic groups is
suggest a “gradual increase in the number of people not statistically significant, as a whole, or between any
dying simultaneously during the Amarna period.”106 two individual groups.113 In the South Tombs Cem-
Caution is needed here, however, in terms of framing etery, however, there is a disparity where adults males
the temporal development of the cemetery. While the were found in multiple-burial graves less frequently
cemetery probably progressed generally from the wadi than some other demographic groups (namely, young
mouth inward, its development was likely guided also adult females and subadults). There is no statistically
by family groups,107 with select areas conceivably re- significant difference between the frequency of multi-
visited over time. It could also be that the Upper Site ple burials when all age groups from the South Tombs
was in use when the city’s population had peaked, and Cemetery are considered collectively (p = 0.058).
more people living in the city resulted in more tem- Finer examination, comparing individual age groups
poral compaction between individual deaths. Finally,
the sampling of the site is not extensive enough to be
confident of what is occurring in the ground between 108
The North Cliffs and North Desert Cemeteries are in-
the excavation areas, and by the time the far end of the
cluded here in discussions of the combined sample but are not
cemetery is reached (Wadi End), no multiple burials considered individually due to their small sample sizes.
are represented in the, albeit small, excavated sample. 109
I.e., 167/228 individuals for whom burial context and
It is possible that the Upper Site reflects a period when full demographic data are discernable from the skeletal and ar-
burial systems were strained, resulting in more mul- chaeological remains.
tiple burials, yet other explanations for these burials
110
I.e., 91/399 individuals for whom burial context and full
demographic data are discernable from the skeletal and archaeo-
are possible (e.g., a form of community practice, or a logical remains.
response to restricted space in a family plot). 111
Chi-squared test using SPSS version 29.0.2.0; χ2 = 152.42.
From a demographic perspective, we look at the in- 112
For this analysis, individuals were sorted into culturally
dividuals in the multiple burials in two ways: in terms relevant age groups consistent with previous publications from
of the overall numbers of individuals of different age the Amarna Project. These include infant (0–2.9 years), early
subadult (3–6.9 years), and late subadult (7–14.9 years). Adults
are individuals 15 years or older and here have been subdivided
into young adults (15–24.9 years) and older adults (25+ years)
103
See Stevens 2018, 116, fig. 13. to accommodate comparisons between the South Tombs Cem-
104
Smith-Guzmán et al. 2016, 311–12, fig. 1.4. etery and the North Tombs Cemetery, where a dearth of indi-
105
Smith-Guzmán et al. 2016, 309–12, fig. 1.4. viduals over 25 years is observed; see Dabbs 2019, 179, for more
106
Smith-Guzmán et al. 2016, 311–12. on these age groups.
107
Suggested in general by the intermixing of burials of adults 113
p = 0.950 for comparisons of all groups at once; p >0.404
and children across the site, and more concretely by emerging for individual age groups compared against each other (Kruskal-
evidence from recent kinship analyses; unpublished data. Wallis test; see Kruskal and Wallis 1952).
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 471

fig. 5. Plan of the Upper Site at the South Tombs Cemetery, showing the multiple-burial graves in gray. The location of the Upper Site
is shown in the inset map of the South Tombs Cemetery (base map by H. Fenwick; courtesy the Amarna Project).

table 2. Basic demographic characteristics of individuals in multiple burials at the four main Amarna cemeteries.
Individuals in South Tombs North Tombs North Cliffs North Desert
multiple burials Cemetery Cemetery Cemetery Cemetery
Adult 148 57 85 2 4
Adult female 92 39 52 0 1
Adult indeterminate 23 6 15 0 2
Adult male 33 12 18 2 1
Subadult 156 52 98 1 5
Total 304 109 183 3 9

with one another, however, reveals that the difference than adult males of any age in the South Tombs Cem-
between young adult females and both young adult etery, although not in the North Tombs Cemetery.
males and old adult males is statistically significant On the face of it, the high frequency of multiple
(p = 0.043 and 0.015, respectively), as is the differ- burials at the North Tombs Cemetery and the fact
ence between both young and old adult males when that no one demographic group was distinct from
compared with early subadults (p = 0.034 and 0.010, others could lend support to a pattern of indiscrimi-
respectively).114 Young adult females and early sub- nate burial, one cause of which could be epidemic
adults were found in multiple-burial graves more often disease. However, disease need not be the only cause
of rapid burial in multiple-burial graves, and the ques-
tion of who is buried with whom is important. Here, it
114
Dunn’s post hoc test.
472 Gretchen R. Dabbs and Anna Stevens [aja 129
table 3. Individuals in multiple-burial graves by specific demographic group; South Tombs Cemetery p = 0.058,
North Tombs Cemetery p = 0.950 collectively. Kruskal-Wallis test (Kruskal and Wallis 1952) to compare demograph-
ic groups within individual cemeteries, p >0.05 for individual demographic groups versus others, except as noted.a, b
Individuals identified only as “adult” or “subadult” are excluded, as are individuals of indeterminate sex.
South Tombs Cemetery North Tombs Cemetery
Total no. No. buried in Total no. No. buried in
Demographic group observed multiple (freq.) observed multiple (freq.)
Young adult female 39 13 (0.333a, b) 68 48 (0.706)
Older adult female 86 21 (0.244) 5 3 (0.600)
Young adult male 31 4 (0.129 )
a
23 18 (0.783)
Older adult male 64 8 (0.125 )
b
0 0 (0.000)
Infant 69 18 (0.261) 3 2 (0.667)
Early subadult 41 14 (0.341 )
a, b
4 3 (0.750)
Late subadult 69 13 (0.188) 125 93 (0.744)
Total 399 91 (0.228) 228 167 (0.732)
a Young adult males statistically different from young adult females (p = 0.043) and early subadults (p = 0.034).
b Older adult males statistically different from young adult females (p = 0.015) and early subadults (p = 0.010).

becomes clear that underlying patterns exist.115 When

looking at the basic divisions of multiple burials by
demographic groups (fig. 6), the most frequent com-
bination is that of one or more adult(s) of either sex
with one or more subadults, which comprises about
73% of multiple burials in the South Tombs Cem-
etery and about 47% of multiple burials in the North
Tombs Cemetery (table 4).116 Of these, around three-
quarters are one or more adult females with one or
more subadults (77% South Tombs Cemetery; 70%
North Tombs Cemetery; see tables 4, 5). The least
common grouping is adult males buried together fig. 6. Proportion of multiple-burial graves by type of individu-
als buried in them.
exclusively (about 2% South Tombs Cemetery; 4%
North Tombs Cemetery).117 Some of this patterning
is certainly driven by the underlying population de- lack of difference in frequency of individuals buried
mography, but these comparisons, coupled with the by demographic groups (see table 3), suggest that the
burial of adult females with subadults in particular was
a choice being made based on culturally specific rules.
115
Here, each grave pit with multiple individuals is consid- The broad patterns are consistent across the two larg-
ered as a unit and assigned to a type that includes adults only est cemeteries at Amarna.118
(female only, male only, or mixed sexes), subadults only (same
age group, i.e., early or late only, or different age groups, both
There are few ancient Egyptian written records
early and late), and mixed burials of adult(s) with subadult(s) that inform on the practice of co-interment, although
(with the subadult[s]: adult female[s] only, adult male[s] only, the burial of infants or children with female adults is
or mixed adults).
χ = 7.38; p = 0.025.
116 2
117
At the North Tombs Cemetery, along with adult males
only, subadults of different age groups and adult male(s) with Of note, the remaining five multiple burials from the North
118

subadult(s) are each observed only twice (i.e., 4% of multiple Cliffs and North Desert Cemeteries are all of the form of one or
burials). more adults buried with one or more subadults.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 473
table 4. General demographic composition of multiple burials in the South Tombs and North
Tombs Cemeteries; p = 0.025, chi-squared test. Individual graves excluded if all demographic charac-
ters could not be estimated for all individuals.
No. multiple burials South No. multiple burials North
Tombs Cemetery (freq.) Tombs Cemetery (freq.)
Adults only 6 (0.146) 11 (0.193)
Subadults only 5 (0.122) 19 (0.333)
Adult/subadult mixed 30 (0.732) 27 (0.474)
Total 41 57

table 5. Demographic composition within individual graves with multiple burials in the South
Tombs and North Tombs Cemeteries.
No. multiple burials South No. multiple burials North
Tombs Cemetery (freq.) Tombs Cemetery (freq.)
Adults only
Females only 2 (0.049) 6 (0.105)
Males only 1 (0.024) 2 (0.035)
Mixed sexes 3 (0.073) 3 (0.053)
Subadults only
Same age group 2 (0.049) 17 (0.298)
Different age group 3 (0.073) 2 (0.035)
Mixed adults and subadults
Adult female(s) with subadult(s) 23 (0.561) 19 (0.333)
Adult male(s) with subadult(s) 5 (0.122) 2 (0.035)
Mixed adult(s) with subadult(s) 2 (0.049) 6 (0.105)
Total 41 57

paralleled frequently enough at Egyptian ­cemeteries— age difference, assuming they were interred simulta-
if not yet studied in depth—to support the idea that neously.121 It is conceivable that infants or children
it was an established cultural practice.119 The natural were sometimes buried with older females from or
assumption is that these represent direct mother- outside their family who died around the same time,
offspring dyads, although this is difficult to confirm
without molecular analysis120 or written evidence.
For the Amarna cemeteries, around 23% of the mul-
121
The difference between the upper margin of the female
age estimate and the lower margin of the subadult age estimate
tiple burials contain an adult female and at least one must exceed 15 years, the estimated minimum age of menarche
subadult who could be her offspring based on their at Amarna (see Dabbs 2023), with the exception of two fetuses
found in situ in the pelvic cavity of an adult female. Three of
these burials (12%) are of adult women with fetal remains found
119
E.g., Brunton 1937, 133–35, pls. LXX, LXXVIII; Brunton outside the pelvic cavity, suggesting the mother and fetus died
1948, pls. LIV–LVI; Onderka and Toivari-Viitala 2014, 60; Min- during birth or miscarriage, although neither the relationship
iaci 2019, 298. We rely here in part on early 20th-century age and between the individuals, nor the timing of death is certain.
sex estimations, which will not always be accurate. Another 36% of these possible mother-offspring dyads had at
120
Limitations on funding and the availability of facilities has least one other individual buried in the grave. Nine graves have
meant that molecular analysis has not yet been possible on the a single adult female with a single subadult where both have ages
skeletal sample from Amarna. that may suggest a mother-offspring relationship (about 8%).
474 Gretchen R. Dabbs and Anna Stevens [aja 129
to model a mother-child relationship; care might have work shows that prone burials are present, although
been sought not only in the afterlife but also during again not predominant, at the other three cemeteries
the dangerous transition through the underworld. as well. For ancient Egypt, while the orientation of the
Other scenarios can be imagined, and it would be deceased body could have had symbolic importance,124
premature to rule out the possibility that male adults including the direction it faced, it is not clear if spe-
served also as caregivers in a burial context. Whether cial meaning was attached to prone burials.125 We also
multiple burials that contain only subadults or only need to consider that the prone position could be un-
adults should be seen as indiscriminate is impossible intentional, if perhaps the correct orientation of the
to know; future molecular work might offer, as a start, body was forgotten after it was wrapped in a matting
insight on whether the individuals in these burials were cylinder (allowing that wider crises could contribute
biologically related. to lapses in concentration or care). In one case at the
The importance of the multiple burials for the ques- South Tombs Cemetery, however, an individual buried
tion of disease at Amarna, therefore, remains compli- prone also had their hands behind their back, rather
cated to unpack. They do stand as potential markers than below the abdomen or by the sides as expected
of temporally concentrated deaths for at least some if the burial was simply a misoriented supine burial
among the Amarna community, and particularly at (Ind. 35, a probable female aged 15–20 years; see
the North Tombs Cemetery, but there is little in these fig. 7). This might indicate a purposeful disruption of
demographic patterns that immediately suggests in- regular practice. It is, however, the only clear case of
discriminate groupings. The fact that around half of this among the South Tombs Cemetery prone burials
the multiple burials across the sampled cemeteries (although most have been disturbed and often only
as a whole are of adult females and subadults could the legs remain in situ).
imply the opposite. This combination, in particular, The South Tombs Cemetery has produced occa-
could instead speak to a careful consideration of both sional other interments, among its undisturbed burials,
who should be buried together and perhaps a system where the limbs of the deceased are somewhat irregu-
of care into the afterlife. Such systems of care may, of larly placed compared to the standard layout, where
course, have held up during times of crisis. Insofar as the head is upright, legs are side by side, and arms and
it is possible to say, burial in a multiple-burial grave at hands are either beside the torso or hips, or the hands
Amarna may often have had more cultural significance meet over the abdomen (see, e.g., Ind. 121 in fig. 7).
than value as an expedient method to dispose of the Ind. 191, an adult male, had his head turned to the left,
dead, particularly beyond the North Tombs Cemetery. left leg disarticulated at the knee, right leg and foot
Posture and Presentation of the Dead. In terms of pos- turned toward the right, and lower arms tucked under
ture, bodies found in the Amarna cemeteries do some- the torso (see fig. 7). He was also buried in a mat that
times diverge from the typical supine position. Some was scattered with potsherds and small pieces of mud-
lie on their sides, although these are often in graves brick or plaster, as though the mat was reused to en-
that are quite narrow, requiring the body to be twisted case the body. Ind. 240, an adult female, had her head
a little. There are two unusual crouched burials, but turned to the left, left arm folded across the abdomen
their rarity suggests a more specific cause than wide- with the hand bent at the right hip, and right arm bent
spread crisis.122 A number of prone burials also occur so the hand was tucked against the right shoulder (see
(fig. 7), raising the question of whether this posture fig. 7). These cases illustrate how divergent practice,
had the kind of negative associations sometimes at- potentially stemming from rushed interment, might
tested beyond Egypt.123 At the South Tombs Cemetery, appear among pit graves, but what again stands out is
up to 17 bodies were interred prone, of the 302 burials their rarity: of the 53 certain undisturbed burials at the
where body position could be determined. Preliminary

124
E.g., toward the sun, or offerings; Raven 2005, 41; Spence
122
They may reflect non-Egyptian burial traditions; Dabbs 2010.
et al. 2022, 4. 125
Some funerary texts do include spells to avoid spend-
123
E.g., for medieval Europe, where a connection of prone ing eternity inverted (i.e., head down and feet up), a symbol of
burial with epidemics has also been suggested: Alterauge et powerlessness, defeat, and disorder; Munro 2017, 57, fig. 4.13.
al. 2020. Might prone burial have been intended to mimic this posture?
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 475

fig. 7. A sample of burials from the South Tombs Cemetery showing a typical supine body layout (Ind. 121), a prone burial (Ind.
42), a disturbed prone burial where the hands are placed behind the back (Ind. 35), and two supine burials where the body is also
irregularly laid out (Inds. 191 and 240) (courtesy the Amarna Project).

South Tombs Cemetery, these are the only two that include: flying insects or their pupae present in graves
have noticeably irregular postures.126 (deposited at the time of burial), separation of limb
Delayed Burial. Delayed burial is another potential elements, skeletons with only slight misalignment of
characteristic of epidemics, as outlined above. Again, joints, and the systematic loss of small skeletal ele-
there is little direct evidence for this in the Amarna ments (e.g., distal phalanges of hands and feet, hyoid,
cemeteries. Mammalian or avian scavenging leaves and anterior teeth). None of these indirect character-
distinctive traces and can result in the loss of elements istics have been observed at the Amarna cemeteries,
to scattering or consumption,127 but no such traces of although many would be obscured in all but the un-
scavenging behavior have been identified in any of the robbed graves. One possible exception is the undis-
Amarna cemeteries. Indirect indicators of excarnation turbed burial of Ind. 191 just noted (see fig. 7): his
hands were tucked beneath his torso, but the right
hand was represented only by a small number of bones
126
Both individuals also have distinguishing bioarchaeological (including one carpal, three metacarpals, and a pha-
features that conceivably point to some sort of otherness as an- lanx), none still articulated with the disrupted fore-
other framework for understanding their divergent burial pos- arm. The right radius was inverted such that the head
tures. Ind. 191 exhibits lesions on the left scapula similar to those
previously identified as possible evidence of corporal punish-
of the radius was resting next to the head of the ulna.
ment; Dabbs and Zabecki 2015, 20. Ind. 240 was unusually tall, The left knee was also disarticulated. If not the result
and in addition exhibits both a healed cranial depression fracture of catastrophic trauma, these characteristics raise the
suggestive of interpersonal violence, for which skeletal evidence possibility of a delay in burial during which the hand
is rare at Amarna, and evidence of a salivatory gland tumor that was partly lost, the right forearm disturbed, and the
may have been visible in the face. Could the placement of her
knee became dislodged, but if so, this is again an iso-
right hand reflect a common posture in life—­attempting to
hide her tumor? lated case at present.
127
Haglund et al. 1988, 985; 1989, 587.
476 Gretchen R. Dabbs and Anna Stevens [aja 129
For the most part, therefore, in terms of the basic earlier), while contemporaneous data from other
preparation, presentation, and condition of the body, Egyptian settlements are too limited at present to allow
and especially at the South Tombs Cemetery where for analysis of wider settlement patterns. It is impor-
the data are largely processed, there are hints at lapses tant, though, to note that the abandonment of Akhet-
of care, but the scope of these is limited, and on their aten was not as abrupt or complete as is sometimes
own they are not strong indicators of a mortality crisis. implied. Archaeological, textual, and iconographic evi-
The overall impression is that the typical burial was dence suggest ongoing occupation, including state-led
likely timely and orderly. activities, for several decades after Akhenaten’s death.
This may have included an initial period into the reign
Changes in Settlement Patterns and Urban Building of Horemheb in which the city’s temples were main-
Programs tained, before a phase that continued to the reign of
As a second line of inquiry, we turn to the wider Ramesses II in which stone buildings were systemati-
urban record of the site. As noted earlier, changes in cally dismantled.131 Kemp and Gabolde have produced
settlement and construction patterns—such as the a “speculative diagram” to capture this general timeline
abandonment of stable, long-established settlements, (fig. 8).132 While the population fluctuations it hypoth-
delayed or aborted building programs, and premature esizes cannot be fixed to absolute or relative figures,
decay of structures—have been identified as potential the diagram does capture the general sense that the
indicators of mortality crises, when they co-occur with city saw progressive growth throughout its time as the
other evidence. The abandonment of Akhetaten after seat of the royal family, at least in the years preceding
just 15–20 years as the seat of the royal family is strik- Tutankhamun’s reign. The evidence for this is some-
ing, although on its own it is not a strong indicator of what piecemeal, but it includes inscriptional material
a mortality crisis because of the likelihood there was from the house of an official named Ranefer that sug-
political instability at this time, and the fact that it was gests he was building his home in the years just after
not a long-established city. The primary catalyst for the Akhenaten’s death,133 stratigraphic evidence that the
abandonment of Akhetaten is usually taken by scholars city’s main temple saw major remodeling in or after
to be the rejection of Akhenaten’s religious movement Akhenaten’s Year 12,134 and an inscription dating to
following his death.128 Evidence such as jar labels in- Akhenaten’s Year 16 at a nearby limestone quarry im-
scribed with kings’ year dates suggests this relocation plying that building materials were still being sourced
occurred within the first few years of Tutankhamun’s at this time.135 There is little immediate sense in Amar­
reign.129 It is assumed that many of the population na’s urban footprint of a city where growth had been
also left the city at this time. Their movements are arrested through crisis before Akhenaten’s death.
not clearly documented in surviving sources, nor While Kemp and Gabolde’s diagram does suggest
with temporal precision, but the impression from the the beginning of population decline from the start of
archaeological record is that the city was abandoned Tutankhamun’s reign, this is simply conjectured on
in an orderly manner that gave people time to collect the grounds that Tutankhamun may have immediately
and remove possessions, sometimes leaving behind begun delegitimizing Akhenaten’s reign, and shifts
items that may have been superfluous or difficult to away from Akhetaten could have accompanied this.
transport, but with no sign of the mass abandonment The diagram also proposes that a reasonably substan-
of personal or household items.130 tial population remained at the city; archaeological
The issue of whether disease might also have been evidence suggests they may have lived initially in the
a factor in spurring the city’s abandonment is difficult Main City and over time consolidated to an area now
to approach through urban data. Akhetaten’s short oc- known as the River Temple, at the south end of this
cupation restricts the kind of wider diachronic views of district (see figs. 2, 8).136 Some of the city’s intramural
settlement patterns that can sometimes isolate periods
of potential social crisis (e.g., for Tell el-Dab’a, noted
131
Kemp and Gabolde 2022.
132
Kemp and Gabolde 2022, 36, fig. 10.
133
Kemp and Stevens 2012, 95, 105, 120–27, 476–77.
128
Kemp and Gabolde 2022, 34. 134
Kemp 2013a, 28; 2013b.
129
Fairman 1951. 135
Van der Perre 2014.
130
Kemp 1995a, 161. 136
Kemp 1995b, 446–48; Kemp and Gabolde 2022, 34–36.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 477

fig. 8. Hypothetical reconstruction of the population size of Akhetaten during the main period of occupation through
the early years of Tutankhamun, and beyond. The numbers above the black bars represent years of reign. X is the brief
and much-debated period when Smenkhkare, Neferneferuaten, or both, may have ruled. The authors note that the reign
lengths for each king will be slightly exaggerated, as the death of each king marked the beginning of his successor’s first
year of reign, regardless of where this fell in the calendar year (modified from Kemp and Gabolde 2022, fig. 10).

burials appear to belong, in part, to this ongoing oc- be used in conjunction with the estimated city size
cupation. While these burials represent significant and length of occupation to estimate the total num-
shifts in situating the dead, they are probably best un- ber of interments that should be expected at Amarna,
derstood for now as reflecting the evolving roles and and provide insight into whether the number of dead
size of the city and its population—and changes to falls in the range of attritional usage, or is in excess of
accommodate these—rather than disruptions from this in a way that may suggest a catastrophic mortal-
health-related crises.137 The available evidence gives ity event.139 The first question, however, is whether
the impression, although not one to be pressed too we have enough reliable data to meet the conditions
far, of both pragmatism in dealing with the city and required by this formula.
its legacy and that any sense of misfortune (disease- Life Expectancy at Birth (e00). While ideally we would
related or otherwise) that may have constituted part use the life expectancy calculated from the Amarna
of that legacy was not so powerful as to see official ac- cemeteries for this analysis, the single generation of
tivity entirely discontinued. the city’s occupation and the attendant effects on
burial composition (namely, the underrepresentation
The Number of Dead at Amarna of infants and older individuals; see below) makes this
As a third line of inquiry, we ask whether paleo­ untenable. Instead, we might look to other temporally,
demographic modeling might be applied to Amar­na geographically, and socially similar samples. Nagar esti-
to offer perspective on whether the number of in- mated average life expectancy at birth to be 24 years for
dividuals buried here is consistent with expected Late Bronze Age individuals in the Levant (at Horvat
numbers for similarly sized cities in the premodern Zelef, n = 100).140 This value is generally consistent
era. Acsádi and Nemeskéri published an equation to with preindustrial assemblages, and Horvat Zelef ’s
estimate the size of a population (P) feeding into a temporal and geographic proximity to Akhetaten make
given cemetery based on average life expectancy at it a suitable proxy.
birth (e00), estimated (or actual) number of dead (D),
Duration of Cemetery Use (t). We are on relatively firm
duration of cemetery use (t), and a correction factor
ground in determining the main occupation period
that equals roughly 10% of cemetery-use duration
of Akhetaten at around 15–20 years. The foundation
(k): P = k + D(e00)/t.138 In theory, this formula might
texts on Akhetaten’s boundary stelae are first inscribed
in Akhenaten’s Year 5, while written evidence noted
above suggests the city was abandoned by the royal
137
See discussions in Stevens 2018, 113–16, and Stevens et al.
2024, 34–35, fig. 19, and the original excavation report, Griffith court by around the third year of Tutankhamun’s reign.
1924, 302. They seem largely to be placed in pits, use burial
mats and sometimes coffins, and have occasional pottery and
other burial objects. There are no clear examples yet published
of disorderly treatment that could indicate, e.g., catastrophic 139
In this example, the equation is solved for D and becomes
circumstances in which bodies were simply left in the streets. d = ((P − k)*t)/e00 , where d is now the expected number of dead.
138
Acsádi and Nemeskéri 1970, 65–66. 140
Nagar 2011, 7.
478 Gretchen R. Dabbs and Anna Stevens [aja 129
With Akhenaten’s highest recorded regnal year at 16 or This estimate, however, relates to the population
17,141 and allowing leeway for one or two short reigns peak, and not an average across its occupation (P).
between him and Tutankhamun, a 15–20-year main We know little about how the population grew. Amar­
occupation period can be reasonably estimated (see na’s ground plan suggests, from concentrations of
fig. 8) and is the figure used here. The main problem, buildings and debris, that housing areas were initially
however, with regard to the use of the cemeteries is contained to a band of land parallel to the river and
whether they were only used during this 15–20-year gradually spread eastward,147 but there is little scope to
period, or were also used by people who stayed on at quantify this growth or tie it to temporal frameworks.
the site afterward. While some of this residual popula- To establish a tentative average population, we might
tion was evidently buried among housing areas, there nonetheless utilize Kemp and Gabolde’s “speculative
is no way of excluding the possibility that use of the diagram” (see fig. 8), which offers an educated guess
cemeteries continued, and no way knowing how long as to growth rates. The diagram assumes a substan-
this might have continued. The effective result of a tial founding population (about 46% of the eventual
longer period of cemetery use, even by a substantially maximum), a large population increase around the be-
smaller population, would be an inflated number of ginning of the city’s fourth year of occupation (Akhen­
burials relative to the 15–20-year use period. aten’s Year 9; about 40% of ultimate maximum) and a
Average Population Size (P). It has proven difficult to final, smaller increase around the beginning of its ninth
estimate population figures for Akhetaten. Because no year (Akhenaten’s Year 14; about 15% of total). The
census lists or clear textual accounts survive for the site, diagram also postulates, as noted earlier, a decrease in
nor for ancient Egypt generally, estimates have been population size of about 13% around the time of Tu-
made by calculating the number of houses and multi- tankhamun’s ascension to the throne. Applying these
plying that by proposed household size.142 Difficulties rates to a maximum population size of 20,000, the av-
arise in determining the original area covered by resi- erage population can be estimated at 17,000 and the
dential buildings, dealing with variations in housing difference between a 15-year and a 20-year occupation
density, and estimating how many people would have is negligible (16,800 versus 17,000). The latter is also
lived in houses of different sizes. Kemp arrived at an true for the estimated 50,000 individual maximum
estimate of 20,400–28,790 for the city, the lower limit population, with an average of approximately 42,350
using an average of five household members and the (42,147 versus 42,548). The average population esti-
upper using six.143 Janssen arrived at a higher estimate mates (P) utilized here will be 17,000–42,350.
of 50,000–100,000,144 the difference occurring in part Estimated Number of Dead (D). We are on slightly more
from his use of generally higher estimates of household solid ground in estimating the number of people bur-
occupants. Both authors stress the tenuous nature of ied at Amarna, as a figure that can be compared with
the figures because of not only the issues noted above the expected number of dead (d) over the selected
but also uncertainty regarding the existence of upper 15–20-year occupation period (t). The number of in-
stories in Amarna houses.145 More recently, and to ac- dividuals at the four main cemeteries can be estimated
commodate the difficulties in ascertaining the upper as: 6,000 (South Tombs Cemetery) + 4,000–5,000
population limit in particular, an estimate of 20,000– (North Tombs Cemetery) + 900–1,400 (North Cliffs
50,000 has generally been put forth.146 Cemetery) + 450–550 (North Desert Cemetery),
summing to approximately 11,350–12,950. The num-
ber of people interred in the rock-cut tombs of officials
is unknown but was probably low. A further unknown
is how many people were buried in graves in hous-
141
Van der Perre 2014, 77.
142
Kemp 1981, 93–97; Janssen 1983, 282–87.
ing areas—or what were to become housing areas—­
143
Kemp 1981, 93–97. during the main 15–20-year occupation period. If,
144
Janssen 1983, 282–87. say, 50 additional interments are added to accommo-
145
The latter seemingly accommodated to a degree in their date these, our overall estimate rises to approximately
estimates, although upper stories have since been suggested to
have been more widespread than many thought at the time these
estimates were made; see Spence 2004.
146
Kemp 2012, 272. 147
Kemp and Garfi 1993, 69–70.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 479
11,400–13,000. The two small, and little-studied, Population Length of Expected
cemeteries at the Workmen’s Village and Stone Village, estimate (P) occupation (t) cemetery size (d)
which might have contained 100–300 individuals,148
15 years ca. 10,700
can be excluded on the basis that the village popula-
17,000
tions are not included in the average population esti- 20 years ca. 14,200
mate (P).
Again, however, the situation is complicated by a
number of factors. First, there is the uncertainty over 15 years ca. 26,600
where the population buried at the North Tombs 42,350
Cemetery lived, with the possibility they resided in 20 years ca. 35,400
encampments, perhaps even as temporary work gangs,
rather than in the houses used to approximate the pop- Archaeologically attested cemetery size: 11,400–13,000
ulation size (P). If this population lived separate from
city residents, the additional base population would fig. 9. Estimated cemetery sizes (i.e., number of buried indi-
increase estimates of the expected number of dead for viduals) based on average population estimates for Akhetaten,
using the Acsádi and Nemeskéri (1970) formula for archaeo-
the city (d), as would continued use of the cemeteries
logical cemeteries.
for an extended period outside the 15–20-year period
we seek to isolate here (t). Conversely, it could be
that some people maintained ties to ancestral burial 10,700–35,400. The archaeologically attested burial
grounds and were interred in cemeteries distant from population of 11,400–13,000 falls at the lower end
Akhetaten,149 which would deflate d; although it does of this range and offers no support for identifying a
seem unlikely that the majority of residents could com- mortality crisis. To reiterate the main caveats outlined
mand the resources to bury their dead far away. above, however: if the city’s population was larger than
A further issue is whether any other cemeteries re- currently estimated or the so-called postabandonment
main to be discovered. The four main cemeteries were occupation was longer or more substantial than cur-
identified during a systematic survey of Amarna’s east rently thought, the case for such crises would weaken
bank,150 and it is unlikely that large burial grounds re- further, but if there are substantially more burials for
main to be identified here. However, there are a num- the Amarna population than yet identified, the case for
ber of poorly studied cemeteries about 4 km north of such crisis could strengthen. Of course, an epidemic
Amarna, in the same wadi system that contains the might also be invisible in these figures if the disease
North Tombs Cemetery and in areas where quarrying was morbid, rather than mortal. In the interest of
and industrial activity were undertaken in the Amarna considering a more extreme situation, we might look
period.151 It is conceivable that some served popula- to Paine’s modeling of epidemic populations using
tions associated with Akhetaten, although if so, the a variety of life expectancy and fertility rates to un-
question would again be whether these people lived derstand the long-term consequences of catastrophic
in parts of the city used to generate the population mortality events. He found that simulated epidemic
estimate (P). disease outbreaks with a 30% mortality rate resulted
Bearing these issues in mind, and without overstat- in greater than 10 times the number of deaths as in a
ing the degree to which the Amarna data are complete, stable year.152 Applying this expectation to Amarna
we applied the Acsádi and Nemeskéri formula to these where the previous calculations suggest an average of
data (fig. 9). We used a life expectancy of 24 years (e00), 700–1,800 deaths per year, the mortality rate of a single
occupation period of 15–20 years (t), and average city catastrophic year would result in 7,000–18,000 deaths
population of 17,000–42,350 (P). The formula pro- in addition to the normal mortality of the remaining
duced an expected cemetery population of roughly years of the cemetery’s use, and it is clear that the ceme-
tery data do not align with such a catastrophic scenario.

148
Stevens 2018, 112.
149
Dabbs and Zabecki 2014, 241.
150
Fenwick 2004.
151
Willems et al. 2009. 152
Paine 2000, 187.
480 Gretchen R. Dabbs and Anna Stevens [aja 129
Observed Demography: Who Was Buried in the demographic curve, although all three have lower than
Cemeteries expected representation of infants and older adults.
Finally, there is the question of whether the demo- The overall demography of Amarna is therefore
graphic profile of the Amarna cemeteries, in terms of atypical for a preindustrial urban environment, but
the age and sex of the deceased, offers any insights this need not be due to epidemic. The two key differ-
on a possible epidemic outbreak. The identification ences between it and the expected age distribution are
of an epidemic disease based on cemetery demog- the relative underrepresentation of infants and of older
raphy alone is again difficult.153 Different diseases adults. Significantly, in epidemics with both attritional
create different mortality patterns, as do epidemics and catastrophic mortality patterns, the numbers of
of the same disease in different environments, at dif- deaths among these categories would be expected to
ferent times, and within the context of individual dis- rise.158 It is also not uncommon for archaeological cem-
ease histories. During an epidemic at any one locale, etery excavations to produce a dearth of infants, in par-
age-specific mortality may also have catastrophic or ticular. This is sometimes attributed to their interment
attritional patterns,154 or characteristics of both.155 in separate cemeteries or in isolated parts of cemeteries
For Amar­na, in addition, there are the outstanding that might not have been sampled, to the poor preser-
questions around the city’s life history and underly- vation of their remains, and to excavator error.159 These
ing population ­profile—including the urban growth are unlikely to have had a substantial impact on the
rate, mobility and migration, founding population— Amarna assemblage given the fairly widespread sam-
and the fact of its short occupation period in general. pling strategy, good preservation overall, and careful
These factors have potential to drastically alter the age- excavation methods, although for the North Cliffs and
at-death profiles of the cemeteries from the expected. North Desert Cemeteries, the relatively small sample
The typical demographic curve associated with pre­- sizes may be a factor. Underrepresentation of infants
industrial urban communities tends to have a U shape, in bioarchaeological samples can also be a reflection of
which arises from high rates of mortality in infancy low fertility rates,160 fewer infants being born resulting
(birth–5 years), followed by a rapid decline in mor- in fewer infant deaths. This may be pertinent for Ama-
tality through childhood and young adulthood, and a rna. Low fertility has been proposed for ancient Egypt
subsequent rise with advancing age.156 The overall de- as a whole,161 and recent work on Amarna-specific data
mography of the Amarna cemeteries is the inverse of has suggested the late onset of menarche,162 which can
this U-shaped curve (table 6; fig. 10), with the skeletal also be associated with slightly higher risks of subfe-
sample exhibiting fewer infants (birth–5 years) and cundity and infertility.163
older adults (45+ years) and more children, adoles- The migratory nature of the Amarna population
cents, and young adults (5–24.9 years) than expect- also bears mention for the possibility that the mass
ed.157 The greatest proportion of the total sampled relocations at the beginning and end of its use as a
population (about 63%) sits in the age range 5–24.9 royal city may have truncated the demographic pro-
years. Examination of the data by individual cemeter- file. Under­enumeration of the young and old may in
ies (fig. 11) shows the South Tombs, North Cliffs, and part be caused by their burial elsewhere, either before
North Desert Cemeteries are the closest to a normal families moved to Akhetaten or after they moved away.
This has the potential to cause an overrepresentation
153
Margerison and Knüsel 2002, 142.
154
Attritional epidemics follow the general pattern of death 158
Chamberlain 2006, 7.
for a given community, with higher death rates. This typically 159
For Egypt, see Richards 2005, 66. Domestic interment is
means the young and old are affected the most. In catastrophic sometimes used to explain the low representation of infants in
death assemblages, the risk of death is relatively equal across all Egyptian cemeteries (e.g., Richards 2005, 66), but this practice
ages and reflects the living age structure of the group; Cham- appears uncommon at Amarna; Stevens 2006, 209–10.
berlain 2006, 7. 160
McFadden et al. 2022, 198.
155
Chamberlain 2006, 74. 161
Masali and Chiarelli 1972, 168.
156
Siler 1979. It is comparison against this pattern, not ab- 162
At 15–19 years; Dabbs 2023, 10.
solute frequencies, that is important in developing an overall 163
Guldbrandsen et al. 2014, 2062. Other factors also affect
understanding of site-specific mortality; DeWitte 2010, 3400. fertility rates, including poor overall nutrition, heavy workloads,
157
Based on those skeletons that can be aged more precisely the length of breastfeeding and interbirth interval, and cultural
than simply “adult” or “subadult.” attitudes on the appropriate number and timing of births.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 481
table 6. Demography of the Amarna cemeteries compared with data (Margerison and Knüsel 2002, 138) from the
Royal Mint (London) plague cemetery and the St. Helen-on-the-Walls (York, U.K.) attritional cemetery; Kruskal-
Wallis test (Kruskal and Wallis 1952).
Royal Mint St. Helen-on-
All Amarna South Tombs North Tombs North Cliffs North Desert Cemetery the-Walls
Cemeteries Cemetery Cemetery Cemetery Cemetery (plague) (attritional)
n (freq.) n (freq.) n (freq.) n (freq.) n (freq.) n (freq.) n (freq.)
0–4.9 years 103 (0.141) 80 (0.212) 4 (0.016) 4 (0.082) 15 (0.273) 68 (0.144) 91 (0.143)
5–14.9 years 241 (0.331) 84 (0.223) 128 (0.516) 14 (0.286) 15 (0.273) 110 (0.233) 209 (0.329)
15–24.9 years 220 (0.302) 81 (0.215) 111 (0.448) 19 (0.388) 9 (0.164) 54 (0.114) 54 (0.085)
25–34.9 years 82 (0.112) 69 (0.183) 3 (0.012) 5 (0.102) 5 (0.091) 99 (0.209) 55 (0.086)
35–44.9 years 59 (0.081) 50 (0.133) 2 (0.008) 1 (0.020) 6 (0.109) 71 (0.150) 128 (0.201)
45+ years 24 (0.033) 13 (0.034) 0 (0.000) 6 (0.122) 5 (0.091) 71 (0.150) 99 (0.157)
Total 729 377 248 49 55 473 636
p-value versus <0.001 <0.001 <0.001 0.344 0.003
plague
p-value versus <0.001 0.009 <0.001 0.710 0.018
attritional

of individuals in the middle age groups in a propor-

tional analysis. Many families, for example, might
already have had infants who died and were buried
in hometowns before they relocated to Akhetaten.
While the absence of a few infants from the founding
population may balance out over the long-term use of
a cemetery, the Amarna cemeteries were utilized for
only one or two generations and the effect may still
be visible. At the other end of the spectrum, when the
main occupation period ended, many of the individu-
als who should have grown old at Amarna, and been
buried there, left and were buried elsewhere: they out-
lived the city itself. We know this is true for some of
Akhenaten’s advisors who created tombs at Amarna fig. 10. Skeletal demography of individuals buried in Amarna’s
but ended up not using them. four main cemeteries (n = 729), divided into common paleo­
The striking demography of the North Tombs Cem- demographic age categories, and excluding individuals for
whom precise age estimates are not possible.
etery requires particular comment (see fig. 11). The
large number of young individuals there, especially in
multiple burials that suggest they died at or around the was a dedicated or overflow cemetery for the general
same time, could be a hallmark of epidemic disease. population who were dying from a disease that primar-
The deaths of large numbers of individuals in the later ily affected children through to young adults. From a
childhood through early adult years is unexpected in pathological perspective, however, it would be unusual
a normal attritional cemetery. As noted earlier, there for otherwise uncompromised people in this age range
are two main scenarios to be considered for this cem- to be affected in this way. Instead, the extremely low
etery: that it served people living in regular housing infant representation (1.6%) and complete lack of
suburbs, or it served one or more special-purpose com- older adults (45+ years) at the North Tombs Cem-
munities who lived separately. For the first scenario, etery strongly suggests this cemetery did not serve a
the restricted age profile could indicate that the site typical population. If the demography is coupled with
482 Gretchen R. Dabbs and Anna Stevens [aja 129

fig. 11. Skeletal demography of each of the four main cemeteries at Amarna, divided into common paleodemographic age categories;
for sample sizes, see table 6.

the paleopathological evidence of extreme workloads, being particularly vulnerable to damage from urban
which includes evidence of degenerative joint disease growth.166 At the moment, we do not have enough
in the youngest individuals represented, it is much data to assess whether burial grounds with similar
more likely that the North Tombs Cemetery served characteristics existed at cities elsewhere, and it would
as the burial ground for a demographically restricted be premature to attribute the Amarna North Tombs
labor force.164 Cemetery to a disease outbreak based on its unusual
The likely need for mass labor to create a new city nature. The deeper research value of this cemetery—
further supports the interpretation of this popula- and the Amarna cemeteries overall for the history of
tion as young laborers, employed under conditions (urban) disease—may lie less in the study of isolated
and for tasks that are not yet fully understood, but outbreaks and more in the wider interconnections
who conceivably lived separate from the main popu- between disease, urban processes, and social change.
lation.165 The cemetery is highly unusual for its large Synergism among poor nutrition, trauma, and heavy
number of multiple burials and restricted demogra- workloads has long been recognized to exacerbate
phy, with no close parallels yet known. Nonelite pit- disease processes, 167 and has more recently been
grave cemeteries are, however, both understudied and
underrepresented in Egypt’s archaeological record,
166
Cemeteries for workers’ communities are known, such
as at the New Kingdom tomb-builders’ settlement at Deir
164
Stevens and Dabbs 2017; Dabbs 2019. An epidemic dis- ­el-Medina and proposed pyramid-builders’ burials on the Giza
ease could of course have affected a demographically restricted plateau, and similar demographic and burial features are not
population, but the demography alone cannot elucidate that. evident in these cemeteries. The former is almost certainly for
If nothing else, we would need to know the size of the popula- higher-status workers than those buried at the North Tombs
tion feeding into the North Tombs Cemetery to consider this Cemetery, however, while the Giza burials so far identified can
argument. represent only a small sample of any workforce associated with
165
Stevens and Dabbs 2017; Stevens 2018, 119–20; Dabbs this site; see Der Manuelian 2009, 134–39.
2019. 167
See Scrimshaw et al. 1968.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 483
i­ntegrated into the concept of syndemics,168 where significant difference in mortality pattern is the North
the interconnected nature of biological, cultural, eco- Cliffs Cemetery (p = 0.710; others p ≤0.018).172
logical, and economic factors can converge to increase Through comparison of the Royal Mint demo-
health burdens. The skeletal remains of the people of graphic data with that from St. Helen-on-the-Walls,
Amarna exhibit characteristics and lesions indicative Margerison and Knüsel went further to propose a
of extreme stress over their life course,169 and the syn- suite of three demographic characteristics to identify,
ergistic stress response to these would have drawn on if all are simultaneously present, an indiscriminate
the limited available resources, perhaps mimicking an catastrophic mortality event, such as epidemic, in the
epidemic experience without a single, or possibly any, preindustrial world:173 (1) a relatively low proportion
pathogen being the root cause. of infants (0–5 years); (2) a low proportion of indi-
Finally, some wider provisional perspective on the viduals 15–25174 years old (possibly due to their abil-
Amarna demographic data might be gained through ity to move out of affected areas);175 and (3) a peak in
comparison with other known epidemic cemeteries, the young adult age group of 25–35 years. Of these
although for data we are at present forced well beyond three required demographic characteristics, the overall
the pharaonic Nile Valley. The Royal Mint cemetery Amarna sample reflects only low infant representation
(East Smithfield, London, U.K.) is one site that pro- (see table 6; fig. 10), as do the individual demographic
vides a rare opportunity to examine epidemic demog- profiles of the South Tombs, North Tombs, and North
raphy, in this case related to the Black Plague. It was Cliffs Cemeteries (although not the North Desert
opened specifically in anticipation of large numbers Cemetery; see fig. 11). While there are questions
of deaths due to the plague outbreak and used only for around whether young adults at Amarna would have
a few months in 1349 CE. Most, if not all, individuals enjoyed the same mobility as their Medieval-period
buried here probably died of plague.170 Although the counterparts, and whether the underlying population
potential for underlying population, ecological, and structures are similar enough for direct comparison, as
cultural differences makes direct comparison tenuous, it stands the Amarna data do not fully align with this
it can be noted that when comparing the proportion particular model for a catastrophic mortality event,
of individuals in each age group between the Royal including the North Tombs Cemetery with its very
Mint plague cemetery and Amarna’s collective cem- distinctive demographic profile.176
etery sample, there is a statistically significant differ- To summarize, therefore, while each Amarna cem-
ence in the proportional representation by age group etery has a distinctive demography that does not nec-
(p <0.001). Individually, each Amarna cemetery, save essarily conform to expectations for a preindustrial
the North Cliffs, also exhibits a proportional com- society, the unexpected aspects of the demography
position different from the Royal Mint cemetery are more likely a reflection of the city’s short-lived
(p ≤0.049; see table 6).171 Margerison and Knüsel com-
pared the Royal Mint demographic data with a con-
temporaneous plague-period attritional cemetery St. 172
Again, this difference may reflect the small North Cliffs
Helen-on-the-Walls (York, U.K.), and comparison of
sample and the large disparity in sample sizes.
Amarna with this site also yields statistically significant 173
Margerison and Knüsel 2002, 141.
differences as a collective (p <0.001) The only individ- 174
Skeletal age estimation of older adults can provide broad
ual Amarna cemetery that fails to show a statistically age ranges that are biased toward underaging, but the dual de-
velopmental and degenerative changes ongoing in early adult-
hood allows for more precision in age estimation than for older
adults, without known systematic bias. Margerison and Knüsel
listed their age groups as in our table 6, with overlap at age 25,
and we have followed their convention, though our age groups
168
See Perry and Gowland 2022; Robbins Schug and Hal- do not overlap. Each age group terminates just before the next
crow 2022. begins; e.g., 15–25 years is actually 15–24.9 years.
169
Kemp et al. 2013; Dabbs et al. 2015; Dabbs 2019; Stevens 175
Margerison and Knüsel 2002, 139.
and Dabbs forthcoming. 176
We acknowledge that a discriminant disease event such
170
Margerison and Knüsel 2002, 136. as cholera that differentially targets some age groups may be
171
This may be as much a reflection of the small sample size obfuscated in demographic analyses, but generally speaking,
of the North Cliffs Cemetery (n = 49) and the extreme differ- diseases with differential impact on the population by age have
ences in sample size as an actual similarity between the two. the greatest impacts on the very young and old.
484 Gretchen R. Dabbs and Anna Stevens [aja 129
occupation and the associated migration events, along does not mean bioarchaeology is silent on the ques-
with low fertility rates, heavy workloads, and nutri- tion. The skeletal remains at Amarna lack indicators of
tional deficiency, than the result of epidemic. While delayed burial sometimes noted in epidemic contexts,
we cannot definitively rule out the possibility of a such as carnivore or rodent modification and decom-
catastrophic epidemic disease event at Amarna based position before burial. From a demographic perspec-
on the demographic patterns of the cemeteries, either tive, modeling of expected cemetery size suggests, at
individually or collectively, neither does the cemetery least provisionally, that while the number of individu-
demography provide sufficient support for such an als buried at the city (estimated between 11,400 and
argument on its own. 13,000) may seem high when noted as an absolute
value, it is consistent with what would be expected for
conclusion an ancient city of Amarna’s estimated size and length of
The archaeological and bioarchaeological data avail- occupation. Finally, while the burial demography does
able from the ongoing study of Amarna’s cemeteries, not conform to expectations for preindustrial societ-
in combination with insights based on long-term work ies, these deviations likely reflect factors such as use by
in the site’s urban areas, provide the first opportunity restricted communities (North Tombs Cemetery), low
to empirically assess claims of an epidemic at ancient fertility rates, and the migration events bookending
Akhetaten. Attempts to do so underscore the difficul- the relatively short occupation (relevant to all Amarna
ties in recognizing epidemics in ancient biological and cemeteries). Looking forward, it is conceivable that
archaeological data, but the balance of evidence sug- infectious disease that can cause an epidemic may be
gests nonetheless that the city was not affected by an identified at Amarna through molecular analysis—
epidemic disease with considerable mortality. and the North Tombs Cemetery would be one prior-
Examination of burial treatments emphasizes a gen- ity assemblage for study—but in this case, any results
eral consistency in the use of simple wrappings with should be integrated with the factors presented here
textile and plant-fiber matting before placement in the to assess whether they represent actual epidemic out-
grave. The few anomalies that can be identified serve to breaks, or simply presence of an ancient pathogen.177
emphasize the overall uniformity of treatment, rather In terms of the wider significance of these results,
than divergent practice that could indicate widespread they suggest, first, that the Hittite plague prayers and
disruption to burial systems. While most individuals related texts do not indicate a disease event that was so
were buried in single graves, the presence of multiple- geographically (and perhaps temporally) widespread
burial graves, and particularly the high frequency of as to substantially affect Akhenaten’s city, 400 km from
them at the North Tombs Cemetery and in the Upper the Mediterranean. They are a reminder of the impor-
Site of the South Tombs Cemetery, could suggest the tance of integrating local data and interpretations with
indiscriminate and expedient burial often noted in epi- internationalist perspectives on the Late Bronze Age
demic cemeteries. The stacking of individuals in graves Mediterranean. Epidemic has sometimes been applied
at the North Tombs Cemetery in a manner that might as a fairly blunt interpretive tool to explain aspects
indicate reduced care for the dead and larger numbers of the disrupted political and social history of Egypt
of dead in these graves further suggest a mortality crisis in the late 14th century BCE. The Amarna results dem-
for the community buried here. However, examination onstrate the value of combined bioarchaeological and
of who was buried with whom in the multiple burials archaeological perspectives in revealing something
across the cemeteries also points to care and protection more of the complexity of ancient disease and its im-
as other potential influencing principles, particularly pacts, particularly when set against an urban backdrop.
in graves containing adult females and subadults, the They also show the contributions of these types of
most common form of multiple burial here. Expedi-
ence may be a factor, but it need not always be due
to disease. 177
Great care must also be taken in considering ancient patho-
The nature of most diseases with epidemic potential gens, as evolutionary change may have increased or decreased
and the slow skeletal response to disease in general de- their virulence, transmissibility, mechanisms of disease develop-
prive us of a suite of macroscopically observable skele- ment, duration and type of symptoms, incubation periods, and
tal lesions for identifying their presence. However, that duration of infectivity.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 485
data, including from modest burials that have often Mittleren Reiches und andere Forschungsergebnisse
been overlooked in Egyptological research, to broader aus dem östlichen Nildelta (Tell el-Dab’a 1979–1984).”
historical questions, such as those surrounding the Zeitschrift der philosophisch-historischen Klasse der Öster-
reichischen Akademie der Wissenschaften 121:312–49.
motivations of state leaders. The wider challenge is to ———. 1991. “Egypt and Canaan During the Middle Bronze
continue to situate disease—infectious, catastrophic, Age.” BASOR 281:27–72.
or otherwise—relative to cultural processes rather Bland, D.M., and B.J. Hinnebusch. 2016. “Feeding Behav-
than as isolated phenomena, and Amarna offers a ior Modulates Biofilm-Mediated Transmission of Yer-
prime opportunity to pursue this for the Bronze Age sinia pestis by the Cat Flea, Ctenocephalides felis.” PLOS
Neglected Tropical Diseases 10(2):e0004413. https://doi
Mediterranean. .org/10.1371/journal.pntd.0004413.
Bonita, R., R. Beaglehole, and T. Kjellström. 2006. Basic Epi-
Gretchen R. Dabbs
demiology. Geneva: World Health Organization.
School of Anthropology, Political Science, and
Brunton, G. 1937. Mostagedda and the Tasian Culture. Lon-
Sociology
don: Bernard Quaritch.
Southern Illinois University
———.1948. Matmar. London: Bernard Quaritch.
Carbondale, Illinois
Bryce, T. 2005. Kingdom of the Hittites. Oxford: Oxford Uni-
[email protected]
versity Press.
Brzobohatá, H., J. Forlík, and E. Zazvonilová. 2019.
Anna Stevens
“Bioarchaeology of Past Epidemic- and Famine-Related
Monash University
Mass Burials with Respect to Recent Findings from the
Melbourne, Australia
Czech Republic.” Interdisciplinaria Archaeologica: Natural
University of Cambridge
Sciences in Archaeology 10(1):79–87.
Cambridge, United Kingdom
Butzer, K.W. 2012. “Collapse, Environment, and Society.”
[email protected]
Proceedings of the National Academy of Sciences 109(10):
3632–39.
Byerly, C.R. 2010. “The U.S. Military and the Influenza
Works Cited Pandemic of 1918–1919.” Public Health Reports 125(3):
82–91.
Acsádi, G.Y., and J. Nemeskéri. 1970. History of Human Life Castex, D. 2008. “Identification and Interpretation of Histori-
Span and Mortality. Budapest: Akadémiai Kiadó. cal Cemeteries Linked to Epidemics.” In Paleomicrobiolo-
Aldred, C. 1988. Akhenaten: King of Egypt. London: Thames gy, edited by D. Raoult and M. Drancourt, 23–48. Berlin:
& Hudson. Springer. https://doi.org/10.1007/978-3-540-75855-6_2.
Alfani, G., and T.E. Murphy. 2017. “Plague and Lethal Epi- Chamberlain, A. 2006. Demography in Archaeology. Cam-
demics in the Pre-Industrial World.” Journal of Economic bridge: Cambridge University Press.
History 77(1):314–42. Chouin, G. 2018. “Reflections on Plague in African History
Alterauge, A., T. Meier, B. Jungklaus, M. Milella, and S. (14–19th c.).” Afriques: Débats, méthodes et terrains d’his­
Lösch. 2020. “Between Belief and Fear: Reinterpreting toire 9. https://doi.org/10.4000/afriques.2228.
Prone Burials During the Middle Ages and Early Mod- Cohen, R., and R. Westbrook, eds. 2000. Amarna Diplomacy:
ern Period in German-Speaking Europe.” PLOS One The Beginnings of International Relations. Baltimore: Johns
15(8):e0238439. https://doi.org/10.1371/journal.po Hopkins University Press.
ne.0238439. Cooney, K.M. 2011. “Changing Burial Practices at the End
Anthony, D.W. 1990. “Migration in Archaeology: The Baby of the New Kingdom: Defensive Adaptations in Tomb
and the Bathwater.” American Anthropologist 92(4): Commissions, Coffin Commissions, Coffin Decoration,
895–914. and Mummification.” JARCE 47:3–44.
Baines, J., and P. Lacovara. 2002. “Burial and the Dead in Dabbs, G.R. 2019. “Bioarchaeology of the Non-Elite North
Ancient Egyptian Society: Respect, Formalism, Neglect.” Tombs Cemetery at Amarna: A Preliminary Assessment
Journal of Social Archaeology 2(1):5–36. of the Non-Elite Individuals of the North Tombs Cem-
Bianucci, R., G. Mattutino, R. Lallo, P. Charlier, H. Jouin- etery at Tell el-Amarna, Egypt.” Bioarchaeology Interna-
Spriet, A. Peluso, T. Higham, C. Torre, and E. Rabino tional 3(3):174–86.
Massa. 2008. “Immunological Evidence of Plasmodium ———. 2021. “Tuberculosis at Tell el-Amarna: A Theoreti-
falciparum Infection in an Egyptian Child Mummy from cal Exercise in Economic and Social Effects of Chronic,
the Early Dynastic Period.” JAS 35(7):1880–85. Terminal Disease in Ancient Egypt.” In The Ancient Egyp-
Bianucci, R., A. Araujo, C.M. Pusch, and A.G. Nerlich. 2015. tians and The Natural World: Flora, Fauna, and Science, ed-
“The Identification of Malaria in Paleopathology: An In- ited by S. Ikram, J. Kaiser, and S. Porcier, 83–98. Leiden:
depth Assessment of the Strategies to Detect Malaria in Sidestone Press.
Ancient Remains.” Acta Tropica 152:176–80. ———. 2023. “Menarche at Amarna: Timing and the Further
Bietak, M. 1984. “Eine Palastanlage aus der Zeit des späten Implications.” American Journal of Biological Anthropology
486 Gretchen R. Dabbs and Anna Stevens [aja 129
183(2):e24856. Enmarch, R. 2008. A World Upturned: Commentary on and
Dabbs, G.R., and M. Zabecki. 2014. “Abandoned Memo- Analysis of the Dialogue of Ipuwer and the Lord of All. Ox-
ries: A Cemetery of Forgotten Souls?” In Remembering ford: Oxford University Press.
and Commemorating the Dead: Recent Contributions in Fairman, H.W. 1951. “The Inscriptions.” In City of Akhenat-
Bioarchaeology and Mortuary Analysis from the Ancient en, III: The Central City and the Official Quarters. Vol. 1,
Near East, edited by B.W. Porter and A.T. Boutin, 217–50. by J.D.S. Pendlebury, 143–223. London: Egypt Explora-
Boulder: University of Colorado Press. tion Society.
———. 2015. “Slot-Type Fractures of the Scapula at New Fenwick, H. 2004. “Ancient Roads and GPS Survey: Model-
Kingdom Tell el-Amarna, Egypt.” International Journal of ling the Amarna Plain.” Antiquity 78(302):880–85.
Paleopathology 11:12–22. Gabolde, M. 1998. D’Akhenaton à Toutânkhamon. Lyon:
Dabbs, G.R., J.C. Rose, and M. Zabecki. 2015. “The Bio­ Université Lumière-Lyon 2, Institut d’archéologie et d’his­
archaeology of Akhetaten: Unexpected Results from a toire de l’antiquité.
Capital City.” In Egyptian Bioarchaeology: Humans, Ani- Gallagher, D.E., and S.A. Dueppen. 2018. “Recognizing
mals, and the Environment, edited by S. Ikram, J. Kaiser, Plague Epidemics in the Archaeological Record of West
and R. Walker, 43–52. Leiden: Sidestone Press. Africa.” Afriques: Débats, méthodes et terrains d’histoire 9.
Dabbs, G.R., A. Stevens, and A. Mérat. 2022. “Teasing https://doi.org/10.4000/afriques.2198.
Identity from Burial Treatment Among the Non-Elite at Gautschy, R. 2014. “A Reassessment of the Absolute Chro-
­Amarna.” Horizon: The Amarna Project Newsletter 22:3–5. nology of the Egyptian New Kingdom and Its ‘Brotherly’
David, R. 2017. “Illness from Afar: Epidemics and Their Af- Countries.” Egypt and the Levant 24:141–58.
termath.” In Pharaoh’s Land and Beyond: Ancient Egypt and Gestoso Singer, G. 2017. “Beyond Amarna: The ‘Hand of
Its Neighbours, edited by P.P. Creasman and R.H. Wilkin- Nergal’ and the Plague in the Levant.” Ugarit-Forschungen
son, 271–85. Oxford: Oxford University Press. 48:223–47.
Dean, K.R., F. Krauer, L. Walløe, O.C. Lingjærde, B. Braman- ———. 2023. “Beyond Amarna: Exorcists Without Borders
ti, N.C. Stenseth, and B.V. Schmid. 2018. “Human Ecto- in the Levant.” Diacrítica 37:51–70.
parasites and the Spread of Plague in Europe During the Goedicke, H. 1984. “The Canaanite Illness.” Studien zur Alt­
Second Pandemic.” Proceedings of the National Academy of ägyptischen Kultur 11:91–105.
Science 115(6):1304–9. Grajetzki, W. 2021. The Archaeology of Egyptian Non-Royal
Der Manuelian, P. 2009. “On the Early History of Giza: The Burial Customs in New Kingdom Egypt and Its Empire. Cam-
‘Lost’ Wadi Cemetery (Giza Archives Gleanings, III).” bridge: Cambridge University Press.
JEA 95:105–40. Green, M.H. 2018. “Putting Africa on the Black Death Map:
DeWitte, S.N. 2010. “Age Patterns of Mortality During the Narratives from Genetics and History.” Afriques: Débats,
Black Death in London, A.D. 1349–1350.” JAS 37(12): méthodes et terrains d’histoire 9. https://doi.org/10.4000
3394–400. /afriques.2125.
———. 2014. “The Anthropology of Plague: Insights from Griffith, F.L. 1924. “Excavations at El-‘Amarnah, 1923–24.”
Bioarchaeological Analyses of Epidemic Cemeteries.” Pan- JEA 10:299–305.
demic Disease in the Medieval World: Rethinking the Black Guldbrandsen, K., L.B. Håkonsen, A. Ernst, G. Toft, J.
Death 1(1):97–124. Lyngsø, J. Olsen, and C.H. Ramlau-Hansen. 2014. “Age
Dodson, A. 2018. Amarna Sunset: Nefertiti, Tutankhamun, Ay, of Menarche and Time to Pregnancy.” Human Reproduc-
Horemheb, and the Egyptian Counter-reformation. Rev. ed. tion 29(9):2058–64.
Cairo: American University in Cairo Press. Habicht, M.E., P.E. Eppenberger, and F. Rühli. 2021. “A
———. 2020. Nefertiti, Queen and Pharaoh of Egypt: Her Life Critical Assessment of Proposed Outbreaks of Plague
and Afterlife. Cairo: American University in Cairo Press. and Other Epidemic Diseases in Ancient Egypt.” Interna-
Dolling, W. 2008. “South Tombs Cemetery.” In B. Kemp, tional Journal of Infectious Diseases 103(February):217–19.
“Tell el-Amarna, 2007–8.” JEA 94:13–31. Haglund, W.D., D.T. Reay, and D.R. Swindler. 1988. “Tooth
———. 2012. “Excavations at the Quarries and Cemetery Artifacts and Survival of Bones in Animal-Scavenged Hu-
on the Eastern Spur: Trenches 7 and 8.” In Akhenaten’s man Skeletons.” Journal of Forensic Sciences 33(4):985–97.
Workers: The Amarna Stone Village Survey, 2005–9. Vol. ———.1989. “Canid Scavenging/Disarticulation Sequence
1, The Survey, Excavations and Architecture, by A. Stevens, of Human Remains in the Pacific Northwest.” Journal of
361–411. London: Egypt Exploration Society and Amar­ Forensic Sciences 34(3):587–606.
na Trust. Hawass, Z., Y.Z. Gad, S. Ismail, et al. 2010. “Ancestry and
Eisen, R.J., and K.L. Gage. 2012. “Transmission of Flea- Pathology in King Tutankhamun’s Family.” Journal of the
Borne Zoonotic Agents.” American Journal of Entomol- American Medical Association 303(7):638–47.
ogy 57:61–82. Helck, W. 1962. Die Beziehungen Ägyptens zu Vorderasien im
Eisen R.J., J.N. Borchert, J.L. Holmes, et al. 2008. “Early- 3. und 2. Jahrtausend v. Chr. Wiesbaden: Harrassowitz.
Phase Transmission of Yersinia pestis by Cat Fleas (Cten- Janssen, J.J. 1983. “El-Amarna as a Residential City.” BibO
cephalides felis) and their Potential Role as Vectors in a 40:273–88.
Plague-Endemic Region of Uganda.” American Journal of Jiménez Meroño, B. 2024. “Social Exclusion in Ancient
Tropical Medicine Hygiene 78(6):949–56. Egypt: A Sociological Approach.” CAJ 34(1):111–26.
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 487
Johnson, N.P.A.S., and J. Mueller. 2002. “Updating the Ac- Margerison, B.J., and C.J. Knüsel. 2002. “Paleodemographic
counts: Global Mortality of the 1918–1920 ‘Spanish’ In- Comparison of a Catastrophic and an Attritional Death
fluenza Pandemic.” Bulletin of Historical Medicine 76(1): Assemblage.” American Journal of Physical Anthropology
105–15. 119:134–43.
Jones, D.S. 2006. “The Persistence of American Indian Health Martin, G.T. 1989. The Royal Tomb at el-’Amarna II: The Re-
Disparities.” American Journal of Public Health 96(12): liefs, Inscriptions, and Architecture. London: Egypt Explo-
2122–34. ration Society.
Kemp, B.J. 1981. “The Character of the South Suburb at Tell Masali, M., and B. Chiarelli. 1972. “Demographic Data on the
el-‘Amarna.” MDIK 113:81–87. Remains of Ancient Egyptians.” Journal of Human Evolu-
———. 1995a. “Site Formation Processes and the Recon- tion 1:161–69.
struction of House P46.33.” In Amarna Reports VI, ed- McFadden, C., B. Muir, and M. Oxenham. 2022. “Determinants
ited by B.J. Kemp, 146–68. London: Egypt Exploration of Infant Mortality and Representation in Bioarchaeological
Society. Samples: A Review.” American Journal of Biological An-
———.1995b. “Outlying Temples at Amarna.” In Amarna thropology 177(2):196–206. http://doi.org/10.1002
Reports VI, edited by B.J. Kemp, 411–62. London: Egypt /ajpa.24406.
Exploration Society. Miarinjara, A., B.M. Bland, J.R. Belthoff, and B.J. Hinnebusch.
———. 2012. The City of Akhenaten and Nefertiti: Amarna 2021. “Poor Vector Competence of the Human Flea, Pulex
and Its People. London: Thames & Hudson. irritans, to Transmit Yersinia pestis.” Parasite Vectors 14(1),
———. 2013a. “Tell El-Amarna, 2012–13.” JEA 99:1–34. no. 317. https://doi.org/10.1186/s13071-021-04805-3.
———. 2013b. “The House of the Aten: The Work of the Miniaci, G. 2019. “Multiple Burials in Ancient Societies:
Spring 2013 Season.” Horizon: The Amarna Project News- Theory and Methods from Egyptian Archaeology.” CAJ
letter 13:8–9. 29(2):287–307.
Kemp, B., and M. Gabolde. 2022. “Rediscovering an Old Mitchell, P.D. 2024. “Parasites in Ancient Egypt and Nubia:
Discovery: F.G. Newton’s Cartouche Pot.” Horizon: The Malaria, Schistosomiasis and the Pharaohs.” Advances in
Amarna Project Newsletter 22:27–37. Parasitology 123:23–49.
Kemp, B.J., and S. Garfi. 1993. A Survey of the Ancient City of Moran, W. 1992. The Amarna Letters. Baltimore: Johns Hop-
el-‘Amarna. London: Egypt Exploration Society. kins University Press.
Kemp, B., and A. Stevens. 2012. Busy Lives at Amarna: Exca- Morris, E. 2023. Famine and Feast in Ancient Egypt. Cam-
vations in the Main City (Grid 12 and the House of Ranefer, bridge: Cambridge University Press.
N49.18). Vol. 1, The Excavations, Architecture and Envi- Mourad, A. 2021. “Strategies of Survival? Change, Continu-
ronmental Remains. London: Egypt Exploration Society ity and the Adaptive Cycle Across the Middle to Early Late
and Amarna Trust. Bronze Age at Tell el-Dab’a, Egypt.” JAnthArch 64:1–25.
Kemp, B.J., A. Stevens, G.R. Dabbs, M. Zabecki, and J.C. Rose. Munro, I. 2017. “The Significance of the Book of the Dead
2013. “Life, Death, and Beyond in Akhenaten’s Egypt: Ex- Vignettes.” In Book of the Dead: Becoming God in Ancient
cavating the South Tombs Cemetery at Amarna.” Antiq- Egypt, edited by F. Scalf, 49–63. Chicago: University of
uity 87(335):64–78. Chicago Press.
Kennedy, H.N. 2006. “Justinian Plague in Syria and the Ar- Murnane, W.J., and C.C. Van Siclen, III. 1993. The Boundary
chaeological Evidence.” In Plague and the End of Antiquity: Stelae of Akhenaten. London: Kegan Paul International.
The Pandemic of AD 541–750, edited by L.K. Little, 87–96. Nagar, Y. 2011. “Human Osteological Database at the Israel
Cambridge: Cambridge University Press. Antiquities Authority: Overview and Some Examples of
Kozloff, A.P. 2006. “Bubonic Plague During the Reign of Use.” Bioarchaeology of the Near East 5:1–18.
Amenhotep III.” KMT: A Modern Journal of Ancient Egypt Nerlich, A.G., B. Schraut, S. Dittrich, T. Jelinek, and A.R.
17(3):36–46, 83–84. Zink. 2008. “Plasmodium falciparum in Ancient Egypt.”
Kuhrt, A. 1995. The Ancient Near East: c. 3000–330 BC. Vol. Emerging Infectious Diseases 14(8):1317–19.
1. Oxfordshire: Routledge. Nicholson, P. 2023. “The Burials at Amarna Site O45.1.”
Laboury, D. 2010. Akhénaton. Paris: Pygmalion. JEA 109:305–6.
Light-Maka, I., T.R. Hermes, R.A. Bianco, et al. 2025. “Bronze Norrie, P. 2016. A History of Disease in Ancient Time: More
Age Yersinia pestis genome from sheep sheds light on hosts Lethal than War. London: Palgrave Macmillan.
and evolution of a prehistoric plague lineage.” Cell 188:1– Nunn, J.F. 1996. Ancient Egyptian Medicine. London: Brit-
15. https://doi.org/10.1016/j.cell.2025.07.029. ish Museum Press.
Littman, R.J. 2009. “The Plague of Athens: Epidemiology Olabarria, L. 2020. Kinship and Family in Ancient Egypt: Ar-
and Paleopathology.” Mount Sinai Journal of Medicine chaeology and Anthropology in Dialogue. Cambridge: Cam-
76:456–67. bridge University Press.
Maixner, F., E. Gaul, A. Mussauer, et al. 2024. “The 3,300- Onderka, P., and J. Toivari-Viitala. 2014. “The Eastern Cem-
year-old Yersinia pestis Genome of an Ancient Egyptian etery Reconsidered.” In The Deir El-Medina and Jaroslav
Mummified Individual.” Paper presented at the European Černý Collections, edited by P. Onderka, H. Navrátilová, J.
Meeting of the Paleopathology Association, Leiden, 21– Toivari-Viitala, A. Hallmann, A. Hilbig, K-H. von Stülpna-
24 August. gel, and H. Svobodová, 51–66. Prague: National Museum.
488 Gretchen R. Dabbs and Anna Stevens [aja 129
Paine, R.R. 2000. “If a Population Crashes in Prehistory Disease.” Behavioral Ecology and Sociobiology 75, no. 122.
and There Is No Paleodemographer to Hear It, Does It https://doi.org/10.1007/s00265-021-03055-8.
Make a Sound?” American Journal of Physical Anthropol- Singer, I. 2002. Hittite Prayers. Atlanta: Society of Biblical
ogy 112(2):181–90. Literature.
Panagiotakopulu, E. 2001. “Fleas from Pharaonic Egypt.” Sloane, A.W. 1973. “Medical and Social Aspects of the Great
Antiquity 75(289):499–500. Plague of London in 1665.” South African Medical Journal
———. 2004. “Pharaonic Egypt and the Origins of Plague.” 47(7):270–76.
Journal of Biogeography 31:269–75. Smith-Guzmán, N.E. 2015. “The Skeletal Manifestation of
Panagiotakopulu, E., and P.C. Buckland. 2010. “The Insect Malaria: An Epidemiological Approach Using Document-
Remains from Ranefer’s House and Grid 12.” In Busy Lives ed Skeletal Collections.” American Journal of Physical An-
at Amarna: Excavations in the Main City (Grid 12 and the thropology 158(4):624–35.
House of Ranefer, N49.18). Vol. 1, The Excavations, Archi- Smith-Guzmán, N.E., J.C. Rose, and K. Kuckens. 2016. “Be-
tecture and Environmental Remains, edited by B. Kemp and yond the Differential Diagnosis: New Approaches to the
A. Stevens. 453–71. London: Egypt Exploration Society Bioarchaeology of the Hittite Plague.” In New Directions
and Amarna Trust. in Biocultural Anthropology, edited by M.K. Zuckerman
Perry, M.A., and R.L. Gowland. 2022. “Compounding Vul- and D.L Martin, 295–316. New York: John Wiley & Sons.
nerabilities: Syndemics and the Social Determinants of Speakman, J.R. 2013. “Sex- and Age-Related Mortality Pro-
Disease in the Past.” International Journal of Paleopathol- files During Famine: Testing the ‘Body Fat’ Hypothesis.”
ogy 39:35–49. Journal of Biosocial Science 45(6):823–40.
Platt, C. 1996. King Death: The Black Death and Its After- Spence, K. 2004. “The Three-Dimensional Form of the Amar­-
math in Late-Medieval England. Toronto: University of na House.” JEA 90:123–52.
Toronto Press. ———. 2010. “Establishing Direction in Early Egyptian Buri-
Raven, M. 2005. “Egyptian Concepts on the Orientation of als and Monumental Architecture: Measurement and the
the Human Body.” JEA 91:37–53. Spatial Link with the ‘Other.’” In The Archaeology of Mea-
Redford, D.B. 1984. Akhenaten: The Heretic King. Princeton: surement: Comprehending Heaven, Earth and Time in An-
Princeton University Press. cient Societies, edited by I. Morley and C. Renfrew, 170–80.
Richards, J. 2005. Society and Death in Ancient Egypt: Mortu- Cambridge: Cambridge University Press.
ary Landscapes of the Middle Kingdom. Cambridge: Cam- Stevens, A. 2006. Private Religion at Amarna: The Material
bridge University Press. Evidence. Oxford: Archaeopress.
Robbins Schug, G., and S.E. Halcrow. 2022. “Building a ———. 2018. “Death and the City: The Cemeteries of Ama-
Bioarchaeology of Pandemic, Epidemic, and Syndem- rna in Their Urban Context.” CAJ 28(1):103–26.
ic Diseases: Lessons for Understanding COVID-19.” Stevens, A., and G.R. Dabbs. 2017. “The North Tombs Cem-
Bioarchaeology International 6:1–22. etery Excavations and Skeletal Analysis.” In B. Kemp, “Tell
Rose, J.C. 2006. “Paleopathology of the Commoners at Tell El-Amarna, Spring 2017.” JEA 103:137–51.
Amarna, Egypt, Akhenaten’s Capital City.” Memorias do ———. Forthcoming. “Excavating the Amarna Cemeter-
Instituto Oswaldo Cruz 101, suppl. 2:73–76. ies: Toward the Lived City.” In Places. Pots. People, edited
Rose, J.C., and M. Zabecki. 2008. “The Quality of Life.” Ho- by A.L. Gascoigne and G. Pyke. Oxford: Archaeopress.
rizon: The Amarna Project Newsletter 4:5. Stevens, A., and P.J. Rose 2020. “Death and burial at the Amar-
———. 2010. “The Commoners of Tell el-Amarna.” In Be- na Workmen’s Village: A Community Cemetery in Con-
yond the Horizon: Studies in Egyptian Art, Archaeology and text.” In Dust, Demons and Pots: Studies in Honour of Colin
History in Honour of Barry J. Kemp, edited by S. Ikram and A. Hope, edited by A. Warfe, J. Gill, C. Hamilton, A. Pett-
A. Dodson, 409–22. Cairo: Publications of the Supreme man, and D. Stewart, 681–704. Leuven: Peeters.
Council of Antiquities. Stevens, A., G.R. Dabbs, C. Rogge, et al. 2018. “Tell El-Amarna,
Sabbahy, L. 2020. “Did Akhenaten’s Founding of Akhetaten Autumn 2017 and Spring 2018.” JEA 104(2):121–44.
Cause a Malaria Epidemic?” JARCE 56:175–79. https:// Stevens, A., G.R. Dabbs, J. Bos, A. Mérat, A. Garnett, and G.
doi.org/10.5913/jarce.56.2020.a011. Tully. 2020. “Tell El-Amarna, Autumn 2018 to Autumn
Scott, S., and C. Duncan. 1998. Human Demography and Dis- 2019.” JEA 106(1/2):3–15.
ease. Cambridge: Cambridge University Press. Stevens, A., G.R. Dabbs, J.E.M.F. Bos, and A. Mérat. 2022.
Scrimshaw, N.S., C.E. Taylor, and J.E. Gordon. 1968. Inter- “Tell el-Amarna, Autumn 2020 to Autumn 2021.” JEA
actions of Nutrition and Infection. Geneva: World Health 108:9–22.
Organization. Stevens, A., G.R. Dabbs, F. Balestra, G. Tully, S. Schiødt,
Shaw, I. 2000. “Chronology.” In The Oxford History of An- P. Rose, A.J. Clapham, P. Docherty, A. Garnett, and M.
cient Egypt, edited by I. Shaw, 480–89. Oxford: Oxford Bertram. 2023. “Tell el-Amarna, 2022.” JEA 109:89–116.
University Press. Stevens, A., F. Balestra, D. Driaux, et al. 2024. “Tell el-Amarna,
Siler, W. 1979. “A Competing-Risk Model for Animal Mor- Autumn 2023 to Summer 2024.” JEA 110:13–45.
tality.” Ecology 60(4):750–57. Strouhal, E., B. Vachala, and H. Vymazalová. 2014. The Med-
Silk, M.J., and N.H. Fefferman. 2021. “The Role of Soc­ial icine of the Ancient Egyptians. Vol. 1, Surgery, Gynecology,
Structure and Dynamics in the Maintenance of Endemic Obstetrics, and Pediatrics. Cairo: American University in
2025] Bioarchaeology of the Late Bronze Age Mediterranean Epidemic 489
Cairo Press. of Egyptology, edited by W. Grajetzki and W. Wendrich. Los
Taylor, J.H. 2010. “Changes in the Afterlife.” In Egyptian Ar- Angeles: University of California Los Angeles. https://
chaeology, edited by W. Wendrich, 220–40. Chichester: escholarship.org/uc/item/77s6r0zr.
Wiley-Blackwell. Wood, J.W., G.R. Milner, H.C. Harpending, and K.M. Weiss.
Tyldesley, J.A. 1998. Nefertiti: Egypt’s Sun Queen. New York: 1992. “The Osteological Paradox: Problems of Infer-
Viking. ring Prehistoric Health from Skeletal Samples.” CurrAnth
Van der Perre, A. 2014. “The Year 16 Graffito of Akhenaten in 33(4):343–70.
Dayr Abū Ḥ innis: A Contribution to the Study of the Later Zabecki, M. 2008. “Human Bones from the South Tombs
Years of Nefertiti.” Journal of Egyptian History 7:67–108. Cemetery.” In B. Kemp, “Tell El-Amarna, 2007–8.” JEA
Willems, H., S. Vereecken, L. Kuijper, et al. 2009. “An Indus- 94:61–67.
trial Site at Al-Shaykh Sa’id/Wadi Zabayda.” Egypt and the Zabecki, M., and J. Rose. 2010. “Bioarchaeological Findings
Levant 19:293–331. from the Amarna South Tombs Cemetery.” Horizon: The
Williamson, J. 2015. “Amarna Period.” In UCLA Encyclopedia Amarna Project Newsletter 8:5–9.