












































Bioarchaeology International 

DOI: 10.5744/bi.2024.0027

Copyright © 2025 University of Florida Press
—-1
—0
—+1

Detailed Recovery Methods Show the Complexity 
of Ancient Mortuary Practices in Later Stone Age 
Hunter- Gatherers of Southern- Central Africa
Jessica I. Cerezo- Román,a,*  Elizabeth A. Sawchuk,b,c,d  Alex Bertacchi,e,f 
Annalys Hanson,g Megan Walsh,a E. Grace Veatch,h  Potiphar Kaliba,i  
and Jessica C. Thompsone,j,k  
aDepartment of Anthropology, University of Oklahoma, Norman, OK, USA
bCleveland Museum of Natural History, Cleveland, OH, USA
cDepartment of Anthropology, University of Alberta, Edmonton, AB, Canada
dDepartment of Anthropology, Stony Brook University, Stony Brook, NY, USA
eDepartment of Anthropology, Yale University, New Haven, CT, USA
fDepartment of Earth and Planetary Sciences, Yale University, New Haven, CT, USA
gSol Solutions, LLC, Paradise Valley, AZ, USA
hHuman Origins Program, Department of Anthropology, National Museum of Natural History, Smithsonian 
Institution, Washington, D.C., USA

iResearch Division, Department of Museums and Monuments, Lilongwe, Malawi
jYale Peabody Museum of Natural History, New Haven, CT, USA
kInstitute of Human Origins, Arizona State University, Tempe, AZ, USA

* Correspondence to: Jessica I. Cerezo- Román, University of Oklahoma– Anthropology, 455 W. Lindsey
St. Dale Hall Tower, Room 509, Norman, OK 73019- 0390, USA

E- mail: jessica . cerezoroman@ou . edu

ABSTRACT This article examines mortuary practices by terminal Pleistocene and Holocene hunter- gatherers from Malawi 
and eastern Zambia in southern- central Africa, with a focus on the evidence for secondary burial and postmor-
tem body manipulation over the past ~16,000 years. Published regional archaeological literature documents 
widespread but variable reports of incomplete or isolated remains from these contexts. While pre- burial body 
exposure, post- burial manipulation, and/or secondary burial with selective removal or interment of certain 
body parts are possible explanations, inconsistent recovery and reporting of archaeological contexts and bone 
modifications makes it challenging to fully exclude taphonomic processes. The social significance of such prac-
tices in regional hunter- gatherer lifeways is also undertheorized because most remains were reported in appen-
dices as element lists that focus on population characteristics. Here, we report human remains recovered 
between 2016 and 2019 through detailed archaeological excavations from five rock shelters in Malawi, which 
provide a way to investigate burial practices with high- resolution data sets. We recovered remains from 19 indi-
viduals, 16 of whom are represented only by isolated elements. Individuals found in primary burials have evi-
dence for pre- burial loss or post- burial removal of elements that are not readily explained by non- anthropogenic 
processes. We interpret some of these patterns as likely evidence of mortuary treatments involving posthumous 
exposure, manipulation, and/or curation of body parts.

Keywords: mortuary practices; hunter-gatherers; Africa

En este artículo se examinan las prácticas mortuorias de los cazadores- recolectores de Malawi y el este de Zam-
bia en el centro- sur de África durante el Pleistoceno terminal y el Holoceno. Particularmente nos concentramos 

Received 3 October 2024 
Revised 21 December 2024 
Accepted 31 January 2025

Published Online 13 June 2025

BI_0-0_08_Cerezo-Román_3pp.indd   1BI_0-0_08_Cerezo-Román_3pp.indd   1 27-05-2025   15:37:3627-05-2025   15:37:36

https://orcid.org/0000-0002-7752-8079
https://orcid.org/0000-0003-4398-2174
https://orcid.org/0000-0002-6952-7666
https://orcid.org/0000-0003-1627-4949


Mortuary Practices in Hunter-Gatherers of Southern-Central Africa2

-1—
0—

+1—

Introduction

Mortuary archaeology allows us to look at how peo-
ple in the past handled and disposed of the bodies of 
their dead in a ritualized fashion (Nilsson Stutz 2016). 
The practices related to the treatment and burial of 
the deceased can be viewed as deeply entangled, with 
a wide range of emotional responses to death, includ-
ing love, anxiety, fear, grief, loss, and standardized 
behaviors directed by cultural norms and traditions. 
The nature and materiality of the corpse may also 
play a significant role in thinking about the responses 
of the living. The cadaver, ambiguous by its very  
nature, located between subject and object, is often  
described as an object with different degrees of per-
sonhood depending on the culture, time period, and 
the identity of the dead itself (Cerezo- Román 2015; 
Kristeva 1980; McClelland and Cerezo- Román 2016; 
Nilsson Stutz 2003, 2016).

What people do with the human remains as part 
of mortuary rituals varies depending on the culture 
and time period. While some people decide to cre-
mate their loved ones, other groups bury their dead, 
then reopen the graves for a variety of reasons, and 
perform secondary mortuary treatments. The use of 
relics, token burials, and curation of human remains 
as part of secondary mortuary rituals in past societ-
ies has been well documented in many cultures 
around the world and throughout time (e.g., Chap-
man et  al. 2017; Chávez 2018; Edlich- Muth 2020; 
Kjellström 2020; Loston 2007; Stratouli et  al. 2010). 
While finding isolated bones in secondary deposits 
of human remains in archaeological sites is not un-
usual, such finds are not typically at the center of the 
research. Furthermore, archaeological identification 

of secondary mortuary practices is challenging be-
cause not all the behaviors associated with their cre-
ation and performance leave archaeological traces, 
osteobiographies are not easily reconstructed from 
isolated elements, and generally, the archaeological 
visibility of these practices is low. In some cases, the 
traces that remain may resemble unintentional or 
non- anthropogenic disturbance.

While understandable from the perspective of pri-
oritizing data- rich features, the bias in mortuary  
archaeology toward primary burials inhibits under-
standing of the full repertoire of past behaviors. Sec-
ondary deposits of human remains in the form of 
isolated findings are often not analyzed, quickly ig-
nored, and discarded as disturbances without further 
exploration. In many cases, isolated findings of hu-
man remains and human remains in secondary de-
posits are a product of bioturbation and taphonomic 
processes, but in other instances, they could be the 
product of human behavior, secondary mortuary 
burial deposit, and part of the archaeological record 
that should not be ignored.

In this article, we explore how mortuary rituals 
and the social significance of diverse practices were 
incorporated into the lifeways of Later Stone Age 
(LSA) foragers of Malawi and eastern Zambia. This 
region and time period have long been described as 
containing human remains in both primary inhuma-
tions and secondary deposits from archaeological 
rock shelters but with little bio archae ol o gi cal atten-
tion dedicated to why this may be. This is a missed 
opportunity to understand the underlying diversity 
and complexity of the forager social and ritual world 
across Africa and beyond. Unlike with earlier work, 
most of the remains described in this study were 

en la evidencia de entierro secundario y manipulación de cuerpos post- mortem durante los últimos ~16,000 
años. La literatura arqueológica regional publicada documenta muchos entierros incompletos o aislados de di-
versas formas en estos contextos. La exposición del cuerpo antes del entierro, manipulación del cuerpo después 
del entierro, entierro secundario con remoción selectiva de restos óseos y/o entierro de ciertos partes del cuerpo 
son explicaciones posibles. Sin embargo, la recuperación y las modificaciones de los huesos en los informes de 
hallazgos arqueológicos son inconsistentes y eso dificulta la exclusión total de los procesos tafonómicos de los 
contextos de restos humanos. La importancia social de tales prácticas en los modos de vida regionales de los 
cazadores- recolectores también es poco teorizado, ya que la mayoría de los restos y hallazgos se describen en 
apéndices que se centran en las características de la población y solo hay listas de los elementos óseos. Aquí in-
formamos sobre los restos humanos recuperados entre 2016 y 2019 a través de excavaciones arqueológicas detal-
ladas de cinco refugios rocosos en Malawi, que proporcionan una manera de investigar las prácticas funerarias 
usando un conjunto de datos de alta resolución. Recuperamos restos de diecinueve individuos, dieciséis de los 
cuales están representados sólo por elementos aislados. Individuos encontrados en entierros primarios tienen 
evidencia de remoción de elementos óseos antes de ser enterrados o después del entierro que no se explican 
fácilmente a través de procesos no antropogénicos. Interpretamos que algunos de estos patrones son posibles 
evidencias de tratamientos mortuorios que involucran exposición póstuma, manipulación y/o curación de par-
tes del cuerpo.

Keywords: prá cticas mortuorias; cazadores- recolectores; África

BI_0-0_08_Cerezo-Román_3pp.indd   2BI_0-0_08_Cerezo-Román_3pp.indd   2 27-05-2025   15:37:3627-05-2025   15:37:36



Cerezo- Román et al. 3

—-1
—0
—+1

recovered using high- resolution excavation tech-
niques and documentation. This enables the first as-
sessment of the diversity of mortuary practice in the 
region, including the potential role of secondary 
mortuary treatment.

Between the 1950s and 1970s, LSA human remains 
were recovered from nine rock shelter sites in eastern 
Zambia and Malawi (Fig. 1) and published as site re-
ports or dissertations (Clark 1956, 1973; Mgomezulu 
1978; Phillipson 1976; Sandelowsky 1972; also see 
Supplementary Materials). These include general de-
scriptions of recovery methods, preliminary counts 
of elements, and a small number of conventional ra-
diocarbon ages. As was commonplace at the time, 
bio archae ol o gi cal research was limited to basic oste-
ological reporting and largely confined to appendices 
(Buikstra 1991).

Excavations at five rock shelters in the Kasitu Val-
ley of the Mzimba District of northern Malawi be-
tween 2016 and 2019 by the Malawi Ancient Lifeways 
and Peoples Project (MALAPP) have been finer in 
detail and offer new information that we articulate 
with older records to reveal considerable variation in 
regional LSA mortuary practice and a far more ex-
tensive bio archae ol o gi cal record than was previously 
known. Our two research questions are as follows: (1) 
What are the different ways of treating the bodies at 
each site? (2) What do these mortuary rituals suggest 
about social memory and the use of space across the 
landscape, as well as the role of human remains in 
these processes?

We first examine the immediate questions of 
where remains were deposited (which sites contain 

them), how they manifest in the archaeological re-
cord (from isolated elements in secondary deposits to 
primary inhumations), and the quality of the evi-
dence for deliberate association with other objects 
within or as part of the grave feature. We also consol-
idate information about the minimum number of in-
dividuals, the evidence for variation in how they were 
treated in the perimortem period, and how these 
variables may relate to other factors such as the onto-
genetic ages and biological sexes of the deceased. 
Drawing from published literature and analyses of 
newly recovered remains, we use spatial analysis to-
gether with taphonomic indicators, such as element 
part representation, fragmentation, and modifica-
tion, to explore the evidence for how human remains 
arrived at these sites and what happened to them 
during and after deposition. By comparing the qual-
ity of evidence from recently excavated sites and  
“legacy” collections (i.e., those excavated decades ago 
prior to contemporary documentation methods and 
often with incomplete contextual information; King 
2016), we are able to assess the relationship between 
recovery practices, data quality, and ability of the 
bioarchaeologists to “see” diverse ancient mortuary 
practices.

Background

The bio archae ol o gi cal record prior  
to recent study

Recent analyses of ancient DNA (aDNA) from 11 an-
cient individuals from Malawi and eastern Zambia 

Figure 1. Locations of sites mentioned in the text. (A) Boxed area on the small map of Africa. (B) Detail of the Kasitu Valley. African countries and 
water bodies shapefiles from Natural Earth. Ecoregions from The Nature Conservancy. Topography from SRTM, a product of NASA. Circles with 
crosses or dots of different colors represent locations of rock shelters with remains of human foragers.

BI_0-0_08_Cerezo-Román_3pp.indd   3BI_0-0_08_Cerezo-Román_3pp.indd   3 27-05-2025   15:37:3827-05-2025   15:37:38

https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3628


Mortuary Practices in Hunter-Gatherers of Southern-Central Africa4

-1—
0—

+1—

show that terminal Pleistocene and Holocene forgers 
in the region were genetically distinct from Bantu- 
speaking people who occupy the region today (Lip-
son et al. 2022; Skoglund et al. 2017). The ancestry of 
these ancient foragers forms a three- way cline be-
tween both ancient and living foragers in eastern, 
central, and southern Africa. This cline was estab-
lished after ~50,000 years ago, but by ~20,000 years 
ago, populations in the region had become more re-
gionally differentiated. This emphasizes the signifi-
cance of Malawi and eastern Zambia as ancient 
crossroads for population interactions and their sub-
sequent role as loci for population differentiation. Be-
cause mortuary practice is heavily embedded in 
cultural norms and traditions, understanding varia-
tion in the treatment of the dead may offer unique 
insight into the emergence of different cultural iden-
tities in tandem with biological divergence. However, 
although Malawi and eastern Zambia have a substan-
tial bio archae ol o gi cal record of LSA foragers, there 
has been little investigation into spatiotemporal  
variation in mortuary treatment (Supplementary 
Materials).

In this article, we define primary inhumation as a 
feature where the individual was deposited as a com-
plete body and the body decomposed in situ (Duday 
2009; Knüsel and Schotsmans 2022). Secondary  
inhumation deposits refer to those with human re-
mains that are isolated, incomplete, and/or unarticu-
lated, indicating relocation to a secondary location 
after initial decomposition or disarticulation else-
where. Surveyed reports include variable descriptions 
of both types, including at least 45 individuals from 
nine sites (17 children, 24 adults, and four with un-
reported age- at- death). Eighteen individuals from 
seven of these sites have enough contextual evidence 
to suggest skeletal elements were part of burial fea-
tures comprising multiple elements from the same 
individual, although the degree of completeness and 
commingling is highly variable. For the other indi-
viduals, either context was not described or described 
as isolated remains. We also note that associated 
worked bone objects and parts of faunal remains for 
five individuals from four sites were reported as po-
tential burial objects, which may offer additional in-
sight into variation in mortuary practice (Table 1).

We discuss all radiocarbon dates in this article in 
calibrated years B.P., calibrated in Oxcal 4.4 using the 
SH20 calibration curve (Hogg et  al. 2020). Date 
ranges for human remains recovered prior to 2016 
were based on conventional radiocarbon dating of 
associated charcoal and only reported for Chencher-
ere II (Clark 1972), Fingira (Sandelowsky 1972), Mtuzi 
(Mgomezulu 1978), and Kalemba, Makwe, and 
Thandwe (all in Phillipson 1976). These placed all 

remains in the middle to late Holocene, except for one 
individual from the terminal Pleistocene at Kalemba. 
Although recent aDNA work (Lipson et al. 2022; Sko-
glund et al. 2017) has largely corroborated this general 
age division through direct ages on individuals from 
Fingira (~6,000– 2,500 years cal. B.P.), Chencherere II 
(~5,000  years cal. B.P.), and Kalemba (~5,000  years 
cal. B.P.), only three individuals reported in earlier lit-
erature from the region have been directly dated: SK- 
5, an adult female from Kalemba (5,285– 4,975 years 
cal. B.P.); Hora 1, an adult male from the Hora 1 site 
(HOR- 1 site, 9,090– 8,770 years cal. B.P.); and Hora 2, 
an adult female from the HOR- 1 site (8,175– 7,944 years 
cal. B.P.)(Clark 1956). All other direct dates on human 
remains are from individuals who were more recently 
recovered or cannot be identified as the same individ-
uals from earlier reports. In total, 10 ancient foragers 
from Malawi (four from Fingira, four from HOR- 1, 
and two from Chencherere II) and one from Kalemba 
in Zambia have published aDNA results (Lipson et al. 
2022; Skoglund et al. 2017).

Patterns in mortuary practice based  
on early research

Initial excavation reports emphasized complete buri-
als and/or clusters of human remains. However, 
many more individuals are represented by isolated 
remains. At Chencherere II, careful study of the fau-
nal remains by Crader (1984a) revealed 21 additional 
fragmentary human elements from at least seven 
more individuals than reported by Clark (1973), rais-
ing the minimum number of individuals (MNI) from 
one to eight. Of the entire sample of 47 LSA individu-
als from the region published to date (the 45 reported 
prior to MALAPP, plus the two infants recovered by 
MALAPP in 2019), it is notable that only six (four 
from HOR- 1, one from Fingira, and one from 
Chencherere II) were represented by any remains in 
clear articulation in a primary inhumation burial. 
All other individuals were represented by commin-
gled, isolated, and/or fragmentary elements, with the 
MNI calculated in the original reports on the basis of 
repeated elements, spatial location within the site, 
and/or ontogenetic ages (Table 2).

This total record comprising mainly fragmentary 
and/or isolated remains (Table 2) could be a conse-
quence of preservation or recovery bias, including 
post- depositional taphonomic disturbance unrelated 
to mortuary behavior. However, early excavation re-
ports do describe original observations that some 
sets of remains appeared intentionally clustered or 
patterned by element in ways suggesting some degree 
of postmortem body exposure, secondary burial, 
and/or manipulation of primary burials (Table 1). For 

BI_0-0_08_Cerezo-Román_3pp.indd   4BI_0-0_08_Cerezo-Román_3pp.indd   4 27-05-2025   15:37:3827-05-2025   15:37:38

https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3628
https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3628


Cerezo- Román et al. 5

—-1
—0
—+1

Ta
b

le
 1.

 A
ge

 a
nd

 se
x 

di
st

ri
bu

tio
n,

 b
ur

ia
l t

yp
e 

an
d 

su
pp

or
tin

g 
ev

id
en

ce
, a

ss
oc

ia
te

d 
ob

je
ct

s, 
an

d 
cu

rr
en

t r
ep

os
ito

ri
es

 o
f a

nc
ie

nt
 h

un
te

r-
 ga

th
er

er
 in

di
vi

du
al

s b
as

ed
 o

n 
hu

m
an

 re
m

ai
ns

 fr
om

 M
al

aw
i a

nd
 

ea
st

er
n 

Za
m

bi
a 

ba
se

d 
on

 re
m

ai
ns

 re
co

ve
re

d 
an

d 
re

po
rt

ed
 p

ri
or

 to
 2

01
6.

 S
ite

s l
is

te
d 

no
rt

h 
(to

p)
 to

 so
ut

h 
(b

ot
to

m
) f

or
 M

al
aw

i a
nd

 Z
am

bi
a,

 re
sp

ec
tiv

el
y. 

N
R 

= 
no

t r
ep

or
te

d.

Si
te

C
ou

nt
ry

Ye
ar

 o
f 

Re
co

ve
ry

In
fa

nt
  

< 
1 y

ea
r

Yo
un

g 
C

hi
ld

  
1–

 5  
Ye

ar
s

O
ld

er
  

C
hi

ld
 

6–
 19

  
Ye

ar
s

Yo
un

g 
A

du
lt 

17
– 3

0 
 

Ye
ar

s

A
du

lt 
30

+ 
Ye

ar
s

In
de

t.
N

R
Se

x*
Bu

ri
al

 
Ty

pe
C

on
te

xt
ua

l E
vi

de
nc

e 
fo

r 
Se

co
nd

ar
y/

M
an

ip
ul

at
ed

Bu
ri

al
 O

bj
ec

ts
Re

po
sit

or
y

Re
fe

re
nc

e

Fi
ng

ir
a 

 
Ro

ck
 

Sh
el

te
r

M
al

aw
i

19
66

0
5

3
0

2
6

0
In

de
t. 

(1
4)

 
M

 (2
)

P 
(1)

 
F/

U
 

(1
5)

N
/A

N
on

e
N

at
ur

al
 

H
is

to
ry

 
M

us
eu

m
, 

Lo
nd

on
, 

U
ni

te
d 

K
in

gd
om

Sa
nd

el
ow

sk
y 

19
72

H
or

a 
1

M
al

aw
i

19
50

0
0

0
1

1
0

0
M

 (1
) 

F 
(1)

P 
(1)

 
PS

 (1
)

“. 
. . 

 w
ith

 th
e e

xc
ep

tio
n 

of
 h

ee
l 

bo
ne

s a
nd

 th
re

e o
th

er
 b

on
es

 
of

 th
e a

nk
le

 o
f t

he
 ri

gh
t f

oo
t, 

al
l t

he
 o

th
er

 b
on

es
 o

f t
he

 fe
et

 
w

er
e m

iss
in

g.
 Th

e l
eft

 fi
bu

la
 

w
as

 m
iss

in
g 

an
d 

on
ly

 th
e 

br
ok

en
 sh

aft
 o

f t
he

 ri
gh

t o
ne

 
w

as
 p

re
se

nt
. A

 n
um

be
r o

f t
he

 
bo

ne
s o

f t
he

 h
an

ds
 w

er
e a

lso
 

m
iss

in
g.

 Th
e h

ee
lb

on
e o

f t
he

 
ri

gh
t f

oo
t w

as
 fo

un
d 

w
he

re
 

th
e k

ne
e c

ap
 sh

ou
ld

 h
av

e 
be

en
—

 th
us

 su
gg

es
tin

g 
th

at
 it

 
ha

d 
be

en
 p

la
ce

d 
th

er
e 

in
te

nt
io

na
lly

. I
t w

ou
ld

 
ap

pe
ar

, t
he

re
fo

re
, t

ha
t t

he
 

bo
dy

 h
ad

 b
ee

n 
su

bj
ec

te
d 

to
 

so
m

e e
xp

os
ur

e b
ef

or
e b

ur
ia

l 
to

ok
 p

la
ce

 . .
 .”

 (1
07

)

“. 
. .

  u
pp

er
 ja

w
 

(p
al

at
e)

 o
f a

 sm
al

l 
an

te
lo

pe
, 

pr
ob

ab
ly

 a
 b

us
h 

bu
ck

, w
hi

ch
 h

ad
 

be
en

 p
la

ce
d 

un
de

r 
th

e 
ri

gh
t s

ho
ul

de
r 

bl
ad

e.
 . 

. .
  T

w
o 

of
 

th
e 

lu
m

ba
r 

ve
rt

eb
ra

e 
of

 
bu

ri
al

 N
o.

 1 
re

st
ed

 
on

 a
 p

er
cu

ss
io

n 
fla

ke
d 

st
on

e 
ax

e,
 

bu
t w

he
th

er
 th

is
 

ca
n 

be
 ta

ke
n 

to
 b

e 
a 

tr
ue

 a
ss

oc
ia

tio
n 

or
 n

ot
 is

 
un

ce
rt

ai
n.

” (
10

7)

U
ni

ve
rs

ity
 

of
 C

ap
e 

To
w

n,
 

So
ut

h 
A

fr
ic

a;
 

N
at

ur
al

 
H

is
to

ry
 

M
us

eu
m

, 
Lo

nd
on

, 
U

ni
te

d 
K

in
gd

om

C
la

rk
 19

56

C
BT

S 
(D

Z4
0)

M
al

aw
i

19
76

1
0

0
0

0
0

0
In

de
t. 

(1)
P 

(1)
N

/A
N

on
e

N
at

io
na

l 
Re

po
sit

or
y, 

N
gu

lu
di

, 
M

al
aw

i

M
go

m
ez

ul
u 

19
78

M
tu

zi
 

(D
Z1

26
)

M
al

aw
i

19
76

0
0

0
4

1
0

1
M

 (6
)

S 
(3

)
“S

ku
lls

 1 
an

d 
3b

 w
er

e 
co

m
pl

et
el

y 
se

pa
ra

te
d 

fr
om

 
th

e r
es

t o
f t

he
ir

 p
os

t- c
ra

ni
al

 
[s

ic
] a

nd
 w

er
e f

ou
nd

 fa
ci

ng
 

do
w

n 
. .

 . 
 Sk

ul
ls 

2 
an

d 
3a

, 
al

th
ou

gh
 e

ac
h 

w
ith

 a
 sp

in
al

 
co

lu
m

n 
le

ad
in

g 
to

 it
, a

lso
 

fa
ce

d 
do

w
n 

w
hi

le
 th

ei
r 

id
en

tifi
ab

le
 p

os
t- c

ra
ni

al
 

fa
ce

d 
up

. Th
e r

ib
- c

ag
e o

f 
‘B

ur
ia

l 2
’ w

as
 sq

ua
sh

ed
 o

ve
r 

th
e p

el
vi

s. 
‘B

ur
ia

l 1
’ h

ad
 a

ll 
th

e l
on

g 
lim

b 
bo

ne
s p

ut
 

to
ge

th
er

 li
ke

 p
ie

ce
s o

f w
oo

d 
in

 a
 p

ile
, 3

0 
cm

. f
ro

m
 th

e  
[s]

ku
ll.

” (
10

3)

“. 
. .

  b
on

e 
po

in
t, 

pr
ob

ab
ly

 u
se

d 
as

 
an

 a
rr

ow
 o

r l
in

k 
sh

aft
 . 

. .”
 (2

34
)

N
at

io
na

l 
Re

po
sit

or
y, 

N
gu

lu
di

, 
M

al
aw

i

M
go

m
ez

ul
u 

19
78 (c

on
tin

ue
d)

BI_0-0_08_Cerezo-Román_3pp.indd   5BI_0-0_08_Cerezo-Román_3pp.indd   5 27-05-2025   15:37:3827-05-2025   15:37:38



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa6

-1—
0—

+1—

Si
te

C
ou

nt
ry

Ye
ar

 o
f 

Re
co

ve
ry

In
fa

nt
  

< 
1 y

ea
r

Yo
un

g 
C

hi
ld

  
1–

 5  
Ye

ar
s

O
ld

er
  

C
hi

ld
 

6–
 19

  
Ye

ar
s

Yo
un

g 
A

du
lt 

17
– 3

0 
 

Ye
ar

s

A
du

lt 
30

+ 
Ye

ar
s

In
de

t.
N

R
Se

x*
Bu

ri
al

 
Ty

pe
C

on
te

xt
ua

l E
vi

de
nc

e 
fo

r 
Se

co
nd

ar
y/

M
an

ip
ul

at
ed

Bu
ri

al
 O

bj
ec

ts
Re

po
sit

or
y

Re
fe

re
nc

e

C
he

nc
he

re
re

 
II

M
al

aw
i

19
72

0
1

4
1

0
2

0
In

de
t. 

(7
) 

M
 (1

)
P 

(1)
 

F/
U

 (7
)

N
/A

“. 
. .

  sm
al

l, 
fin

el
y 

po
in

te
d 

bo
ne

 a
w

l 
ne

ar
 th

e 
ri

gh
t 

ar
m

 . 
. .”

 (C
la

rk
:9

)

N
at

io
na

l 
H

is
to

ry
 

M
us

eu
m

, 
Lo

nd
on

, 
U

ni
te

d 
K

in
gd

om

C
la

rk
 19

72
; 

C
ra

de
r 1

98
4

N
ac

hi
ku

fu
 

C
av

es
Za

m
bi

a
19

48
0

0
0

0
0

0
1

In
de

t. 
(1)

U
 (1

)
N

/A
N

on
e

U
nk

no
w

n
C

la
rk

 19
50

Th
an

dw
e

Za
m

bi
a

19
70

0
1

0
0

2
0

0
In

de
t. 

(1)
 

M
 (1

) 
F 

(1)

P 
(1)

 
S 

(1)
 

F/
U

 (1
)

“I
t a

pp
ea

rs
 th

at
 th

e s
ke

le
to

n 
w

as
 p

ar
tly

 d
is

ar
tic

ul
at

ed
 o

r 
di

sm
em

be
re

d 
be

fo
re

 b
ur

ia
l. 

O
ne

 fe
m

ur
 w

as
 b

ro
ke

n 
an

d 
th

e t
w

o 
ha

lv
es

 w
id

el
y 

se
pa

ra
te

d.
 Th

e f
ac

ia
l r

eg
io

n 
of

 th
e s

ku
ll,

 to
ge

th
er

 w
ith

 th
e 

m
ax

ill
a,

 w
as

 m
is

si
ng

, b
ut

 th
e 

m
an

di
bl

e w
as

 p
re

se
nt

. Th
e 

at
la

s a
nd

 a
xi

s v
er

te
br

ae
 w

er
e 

fo
un

d 
ar

tic
ul

at
ed

 w
ith

 th
e 

sk
ul

l b
ut

 se
pa

ra
te

d 
fr

om
 th

e 
re

st
 o

f t
he

 v
er

te
br

al
 co

lu
m

n.
” 

(5
0)

“O
ve

r t
he

 fe
et

 o
f 

th
is 

[e
ld

er
ly

 
in

di
vi

du
al

] 
sk

el
et

on
 w

er
e 

pl
ac

ed
 tw

o 
fr

ag
m

en
ts

 o
f 

w
ar

th
og

 sk
ul

ls,
 

ea
ch

 re
ta

in
in

g 
a 

pa
ir

 o
f t

us
ks

. O
ve

r 
th

e t
ib

ia
e w

as
 se

t 
an

ot
he

r s
uc

h 
fr

ag
m

en
t w

ith
 a

 
si

ng
le

 m
as

siv
e 

tu
sk

.” 
(5

0)

Li
vi

ng
st

on
e 

M
us

eu
m

, 
Li

vi
ng

st
on

e,
 

Za
m

bi
a

Ph
ill

ip
so

n 
19

76

K
al

em
ba

Za
m

bi
a

19
71

1
0

1
2

0
1

0
In

de
t. 

(3
) 

F 
(2

)
P 

(1)
 

S 
(3

) 
F/

U
 (1

)

SK
- 2

 (a
du

lt 
F)

: o
nl

y 
a 

sk
ul

l 
an

d 
ce

rv
ic

al
 v

er
te

br
ae

 
pr

es
en

t; 
SK

- 4
 (c

hi
ld

 
7–

 8 
ye

ar
s)

: “
Th

e s
ku

ll 
ha

d 
ev

id
en

tly
 b

ee
n 

br
ok

en
 b

ef
or

e 
bu

ri
al

 a
nd

 th
e p

ie
ce

s w
er

e 
se

pa
ra

te
d.

”; 
SK

- 5
 (a

du
lt 

F)
: 

“Th
e p

it 
w

as
 li

ne
d 

w
ith

 th
re

e 
pi

ec
es

 o
f a

 h
um

an
 p

el
vi

s a
nd

 
on

e s
ku

ll 
fr

ag
m

en
t. 

In
 th

e 
ar

ea
 th

us
 d

em
ar

ca
te

d 
w

er
e 

pl
ac

ed
, o

ne
 a

t e
ac

h 
en

d,
 th

e 
tw

o 
ha

lv
es

 o
f a

 m
an

di
bl

e.
 

Be
tw

ee
n 

th
es

e w
as

 a
 p

ile
 o

f 
ve

rt
eb

ra
 a

nd
 ri

b 
fr

ag
m

en
ts

, a
 

fe
w

 b
its

 o
f s

ha
tte

re
d 

lo
ng

- b
on

e,
 a

nd
 tw

o 
pi

ec
es

 o
f 

th
e t

em
po

ra
l.”

 (1
25

)

N
on

e
Li

vi
ng

st
on

e 
M

us
eu

m
, 

Li
vi

ng
st

on
e,

 
Za

m
bi

a

Ph
ill

ip
so

n 
19

76

M
ak

w
e

Za
m

bi
a

19
66

0
0

0
0

0
0

3
In

de
t. 

(3
)

F/
U

 (3
)

N
/A

N
on

e
Li

vi
ng

st
on

e 
M

us
eu

m
, 

Li
vi

ng
st

on
e,

 
Za

m
bi

a

Ph
ill

ip
so

n 
19

76

*S
ex

 e
st

im
at

es
 a

re
 re

po
rt

ed
 o

nl
y 

fo
r a

du
lt 

in
di

vi
du

al
s w

ith
 su

ffi
ci

en
tly

 c
om

pl
et

e 
m

or
ph

ol
og

y;
 la

te
r c

or
ro

bo
ra

te
d 

w
ith

 a
D

N
A

 o
n 

th
e 

tw
o 

ad
ul

ts
 fr

om
 H

or
a 

1 (
Sk

og
lu

nd
 e

t a
l. 

20
17

) a
nd

 o
ne

 a
du

lt 
fr

om
 K

al
em

ba
 

(L
ip

so
n 

et
 a

l. 
20

22
)

**
Bu

ri
al

 ty
pe

 is
 a

s p
er

 o
ri

gi
na

l e
xc

av
at

or
 in

te
rp

re
ta

tio
n.

 P
 =

 p
ri

m
ar

y;
 P

/S
 =

 p
ri

m
ar

y 
w

ith
 p

os
td

ep
os

iti
on

al
 m

an
ip

ul
at

io
n;

 S
 =

 se
co

nd
ar

y;
 F

/U
 =

 fr
ag

m
en

ta
ry

 b
ut

 u
ns

pe
ci

fie
d;

 U
 =

 u
ns

pe
ci

fie
d

Ta
b

le
 1.

 (C
on

tin
ue

d)

BI_0-0_08_Cerezo-Román_3pp.indd   6BI_0-0_08_Cerezo-Román_3pp.indd   6 27-05-2025   15:37:3827-05-2025   15:37:38



Cerezo- Román et al. 7

—-1
—0
—+1

Table 2. Minimum number of elements from 43 (minimum) individuals from seven sites excavated prior to 2016. Individual element data were not 
reported from the Changoni Bible Training School or for the Nachikufu Caves, so they are not included here, but individuals from these sites do 
appear as part of a total MNI of 45 in Table 1.

Eastern Zambia Central Malawi Northern Malawi

Body Region Kalemba Makwe Thandwe Mtuzi Chencherere II Hora 1 Fingira Rock Shelter TOTAL

HEAD
Cranium 29 0 13 14 1 2 30 89
Maxilla 0 0 3 0 0 0 0 3
Mandible 2 0 2 4 0 1 4 13
Hyoid 0 0 0 0 0 0 3 3
Tooth 32 4 25 84 39 48 74 306
SHOULDER
Scapula 0 0 3 1 2 4 9 19
Clavicle 0 0 2 2 3 3 3 13
NECK/CHEST/BACK
Sternum 0 0 1 0 1 1 2 5
Manubrium 0 0 0 0 0 0 1 1
Vertebra 8 0 4 2 27 39 97 177
Rib 2 0 3 2 27 25 3 62
PELVIS
Pelvis 1 0 5 2 7 3 8 26
Sacrum 0 0 1 0 0 0 0 1
ARM
Humerus 0 0 4 2 2 4 6 18
Radius 0 1 5 2 2 3 3 16
Ulna 1 0 4 1 5 4 2 17
WRIST
Carpal 0 0 3 9 16 8 15 51
HAND
Metacarpal 0 0 7 11 10 4 18 50
Manual element 0 0 0 0 0 1 0 1
LEG
Femur 1 0 5 3 3 4 5 21
Patella 0 0 2 1 0 0 4 7
Tibia 1 0 4 3 2 4 5 19
Fibula 1 0 3 2 3 3 2 14
ANKLE
Tarsal 0 0 6 8 10 11 40 75
FOOT
Metatarsal 0 0 2 1 10 2 28 43
Metatarsal or Philangeal 0 0 0 0 0 0 28 28
Pedal Element 0 0 0 0 0 3 0 3
WRIST/ANKLE UNSPECIFIED
Carpal/Tarsal 0 0 0 6 0 0 0 6
HAND/FOOT UNSPECIFIED
Phalanx 1 0 11 15 57 5 92 181
TOTAL 79 5 118 175 227 182 482 1,268

example, at HOR- 1, in northern Malawi, Clark 
(1956:107) describes the conditions of recovery of the 
adult female skeleton: “with the exception of the heel 
bones and three other bones of the ankle of the right 
foot, all the other bones of the feet were missing. The 
left fibula was missing and only the broken shaft of 
the right one was present. A number of the bones of 
the hands were also missing. The heelbone of the 
right foot was found where the knee cap should have 
been- thus suggesting that it had been placed there in-
tentionally” (Fig. 2).

At Mtuzi, in the central part of Malawi, six indi-
viduals are represented by fragmentary crania with 

a small quantity of fragmented postcranial elements 
but considered by Mgomezulu (1978) to have been 
clustered within three graves. Mgomezulu (1978:103) 
writes that “‘Burial 1’ had all the long limb bones put 
together like pieces of wood in a pile, 30 cm. from 
the [sic] Skull.” At Kalemba in eastern Zambia, Phil-
lipson (1976:125) indicates there are “four separate 
human burials” and specifies three as secondary 
burials (Table  1). At Thandwe, Phillipson (1976:50) 
describes evidence based on the arrangement and 
fragmentation of elements that “the skeleton was 
partly disarticulated or dismembered before burial” 
(Table 1).

BI_0-0_08_Cerezo-Román_3pp.indd   7BI_0-0_08_Cerezo-Román_3pp.indd   7 27-05-2025   15:37:3827-05-2025   15:37:38



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa8

-1—
0—

+1—

The use of stones as part of mortuary practice adds 
additional information about regional variation and 
may also have facilitated identifying the locations of 
existing burials as foragers reoccupied the same sites. 
At Thandwe, an elderly female was interred with a 
large triangular rock that Phillipson (1976:50) inter-
prets as a marker stone. Overlying this burial were the 
commingled remains of an elderly male and a child 
aged two to three years at death, both of which were 
also subsequently covered by a stone slab. Clark 
(1973:9) also describes a “pile of stones resting in a 
shallow depression dug into the rotten granite” floor 
of Chencherere II, into which the body of an adult 
male was interred. Sandelowsky (1972:120– 122) reports 
that for the adult male burial from Fingira, “Stones lay 
above the fractures of the right shoulder, arm and leg 
as well as above two breaks in the bones of the left leg.”

Combining previous and new research

Taken together, these observations open the possi-
bility of broader variation in mortuary treatment 
than previously considered, including primary and 
secondary inhumations. Furthermore, osteological 
analysis (Table  1) revealed individuals in inferred 

secondary inhumations span all age and both sex cat-
egories. They also show that for some individuals, 
there was community investment in these interments, 
including the use of stones that could have deterred 
scavengers but also may have served as markers. The 
well- described and disproportionate representation 
of cranial elements inside commingled burial features 
at Kalemba and Mtuzi (Table  2) is consistent with 
 original excavator interpretations that there was some 
degree of body manipulation (Mgomezulu 1978; Phil-
lipson 1976). Recovery of remains from different lay-
ers (e.g., Kalemba) or with thousands of years of 
separation between directly dated individuals (e.g., 
Fingira) also shows that the same places were used for 
mortuary practice over periods of time that far exceed 
individual living memory. At this time, the associated 
archaeological record is too coarsely resolved to know 
if this represents local continuity in cultural practice 
or simply convergence on the solution of rock shelters 
as receptacles for the dead.

Understanding ancient mortuary practices in LSA 
foraging communities requires shifting attention 
away from clear primary inhumations to a more bal-
anced approach where secondary burial deposits of 
human remains, as well as isolated bones, are also an-
alyzed and considered important aspects of the bio-
archae ol o gi cal record. This is important for our area 
of study and any site that presents secondary deposits 
of human remains. At any rock shelter site with com-
plex formation history, fragmentation and commin-
gling of remains may be the result of unrelated human 
activities at the site such as pit- digging, trampling, or 
campfires. Non- anthropogenic taphonomic processes 
may have also played a role (e.g., sediment compac-
tion, bone dissolution, and/or bioturbation) (Ha-
glund and Sorg 2002; Pokines et al. 2022). It is also 
possible that secondary burial deposits and body ma-
nipulation may have been an important aspect of 
postmortem treatment in Late Pleistocene and Holo-
cene foragers of southern- central Africa prior to 
~2,000  years ago that has largely gone undescribed. 
With the exception of Crader’s (1984a) work at 
Chencherere II and possibly the lists of elements from 
Finigra in the appendix of Sandelowsky’s (1972) dis-
sertation, most collections have not undergone care-
ful study of faunal assemblages to identify isolated 
human elements. Human remains may therefore be 
present, in isolated and/or highly fragmented form, at 
sites that do not report them— or they may be more 
numerous than realized at sites where clear primary 
inhumations have also been recovered. If so, then this 
would change the archaeological conceptualization of 
the mortuary landscape as it was experienced by LSA 
foragers by increasing the number of localities an-
cient foragers used as repositories for their dead and 
adding nuance about specific kinds of treatments. 

Figure 2. Hora 2 adult female in situ prior to removal, reproduced 
from Clark (1956:Figure VI) and enhanced for clarity using Photoshop 
version 24.1.1. (A) Missing left fibula. (B) Right calcaneus displaced to 
the position of the left knee. Photo credit: Society of Malawi.

BI_0-0_08_Cerezo-Román_3pp.indd   8BI_0-0_08_Cerezo-Román_3pp.indd   8 27-05-2025   15:37:3827-05-2025   15:37:38



Cerezo- Román et al. 9

—-1
—0
—+1

Although foragers may well have used other localities 
outside of shelters as mortuary places, these places 
were not preserved or have not yet been located. There-
fore, here we focus on rock shelter deposits by provid-
ing new data about element representation, MNI, 
spatial relationships, and taphonomic variables from 
five recently excavated rock shelter sites in the Kasitu 
Valley of the Mzimba District of Malawi.

Study Area, Sites, and Methods

The Kasitu Valley of Malawi separates the eastern 
highlands of the Viphya Plateau from plains that 
grade west into the Luangwa Valley of Zambia (Fig. 1). 
It is characterized by Zambezian miombo open 
woodland interspersed with edaphic grasslands 
known as dambos (DeBusk 1998; Wright et al. 2024). 
Prominent local landmarks are the Kasitu River, a pe-
rennial water source at the foothills of the highlands, 
and Mount Hora, a granite- gneiss inselberg that rises 
110 m from the plain to 1,716 m above sea level (AMSL).

Two of the excavated rock shelters (HOR- 1 and 
HOR- 5) are located near the base of the inselberg, at 
1,470 m and 1,503 m AMSL, respectively. HOR- 1 was 
first excavated by Clark in 1950 (Clark 1956) and 
again by MALAPP in 2016, 2017, 2018, and 2019. 
HOR- 5 was excavated by Sandelowsky in 1966 (Sand-
elowsky 1972) and again by MALAPP in 2018. A third 
site, Mazinga 1, or MAZ- 1 (1,401 m AMSL), was exca-
vated by MALAPP in 2017, 2018, 2019, 2022, and 2023 
and is located at the foothills of the highlands, where 
the Luwelezi stream flows into the Kasitu River. The 
Kadawonda 1 and 2 (KAD- 1 and KAD- 2) sites, exca-
vated by MALAPP in 2017, are part of a complex of 
rock shelters formed in a small outcrop in the upper 
Luwelezi catchment, at 1,709 m AMSL. Full descrip-
tions of cultural sequences, stratigraphy, and chronol-
ogy are presently underway, so for the purposes of 
this analysis, we used broad age categories such as 
“Holocene” and “Pleistocene” and refer to published 
ages for the sites for more specific information. As 
analysis is ongoing, only remains recovered through 
the 2019 excavation seasons are reported here. Exca-
vation strategies, site recording, and processing of re-
covered materials were the same across all sites and 
included recovery and analysis of all finds larger than 
3 mm and, in the case of HOR- 1, larger than 1 mm 
(Supplementary Methods).

The data generated in our examination of the hu-
man remains were analyzed using SPSS 24 and Mic-
rosoft Office Excel 2016. The variables recorded 
during the study of the deposits consisted of three 
main types: biological data, posthumous treatment of 
the body, and archaeological context. J.C- R. recorded 
biological data that includes estimations of sex, 

age- at- death, pathological conditions, and trauma. 
J.C.T. recorded further taphonomic attributes such as 
degree of matrix coverage, fossilization, and micro-
scopic bone surface modifications. The variables doc-
umented for the posthumous treatment of the body 
and archaeological context include (a) body manipu-
lation and treatment; (b) spatial analysis, deposit 
type, and context; and (c) associated objects.

The human skeletal remains consisted of primary 
inhumations, bones found in a burned charcoal and 
ash feature, and isolated remains. The degree of frag-
mentation and fossilization of the remains limited 
analytical observations, particularly clinical diagno-
ses of pathological conditions and the frequency and 
etiology of traumas and cut marks. The protocols for 
osteological data collection were based primarily on 
those of Buikstra and Ubelaker (1994) and subse-
quent revisions (Arizona State Museum 2018; Cun-
ningham et al. 2016). Skeletal data collection consisted 
of documenting metric and morphological observa-
tions of remains. For each individual and isolated re-
mains, a detailed skeletal inventory was generated, 
which included recording the presence of elements 
and their conditions. These analyses allow for inter-
pretations of body completeness at the moment of 
burial and the number of individuals represented in 
each deposit (refer to supplement for details on the 
methods that were used to estimate the age- at- death, 
sex, trauma, and pathologies). After bio archae ol o gi-
cal identification, the spatial positions of all elements 
were plotted in ArcGIS versions 8+ in plan and three- 
dimensional view. Methods for obtaining MNI esti-
mates based on spatial and ontogenetic data are 
provided in the supplement.

We analyzed the posthumous treatment of the bod-
ies using primary and secondary data. The primary 
data were generated by analyzing the practices per-
formed directly or indirectly on the body as evidenced 
in the human skeletal remains (e.g., burning, cut 
marks, grave items). Secondary data were collected 
from basic contextual information (e.g., spatial associ-
ations, orientations and positions of the remains, 
chronologies, field observations about sedimentary 
context) and publications. Body manipulations were 
analyzed to reconstruct the ritual practices done to the 
deceased individuals. The intentional deposits that 
contain human bone were classified as either burials 
or non- burials, as well as whether they were in burned 
features or in secondary or mixed contexts (Knüsel 
2014; Knüsel and Schotsmans 2022; Roksandic 2001).

Results

Human remains recovered by MALAPP across  
the five sites represent a minimum number of 19 

BI_0-0_08_Cerezo-Román_3pp.indd   9BI_0-0_08_Cerezo-Román_3pp.indd   9 27-05-2025   15:37:3827-05-2025   15:37:38

https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3628


Mortuary Practices in Hunter-Gatherers of Southern-Central Africa10

-1—
0—

+1—

individuals, which together with the two adults re-
ported by Clark (1956) total 21 individuals from the 
Kasitu Valley sites (Table  3; Supplementary Data 
Set). Here, we detail new results from HOR- 1, HOR- 5, 
and MAZ- 1 and in the discussion combine them 
with results reported prior to initiation of MALAPP. 
Table 3 also includes isolated elements from two ad-
ditional sites for which there is currently insufficient 
context to reconstruct mortuary practices in spatial 
detail because there is only one from each site: the 
crown of an upper right maxillary molar from a 
child aged ≥2.5 to ≤4.5  years from KAD- 1 and a 
manual distal phalanx from a child ≤5  years from 
KAD- 2 (Fig. 1).

HOR- 1 is to date the only reported site in the  
Kasitu Valley containing primary inhumations  
(Table  3). The first two individuals are the adults 

recovered by Clark in 1950, which he designated as 
“Hora Burial 1” (directly dated to ~9,200  years ago) 
and “Hora Burial 2” (directly dated to ~8,100  years 
ago); both were found to preserve ancient DNA (Lip-
son et al. 2022; Skoglund et al. 2017a). Bio archae ol o gi-
cal and contextual details of Hora 1 and Hora 2 from 
published literature are provided in the supplement. 
MALAPP excavations between 2016 and 2019 resulted 
in the recovery of three additional inhumations. One 
was represented by a concentration of burned adult 
human bone, here designated “Hora 3,” that may rep-
resent a primary deposit within an ash feature. The 
feature, but not the remains within them, dates to 
~9,500 years ago, as reported by Lipson et al. (2022), 
and will be described in detail together in a separate 
publication. The other two inhumations were pri-
mary burials of infants (“Kahora 1” and “Kahora 2”), 

Table 3. Minimum number of individuals identified from five sites in the Kasitu Valley.

Site Individual Age at Death Sex
Sex 
Determination

aDNA 
Individual Context Chronological Age Reference

HOR- 1 Hora 1 30– 50 yrs M Morph, aDNA I2966 Burial (primary) 9, 124– 8, 972 cal 
BP (direct enamel)

Clark 1956; Lipson 
et al. 2022; 
Skoglund et al. 2017

HOR- 1 Hora 2 20– 30 yrs F Morth, aDNA I2967 Burial (primary with 
missing major 
elements)

7, 960– 8, 170 Cal 
BP (direct collagen)

Clark 1956; 
Skoglund et al. 
2017

HOR- 1 Hora 3 18– 50 yrs F N/A N/A Burial (primary in a 
burned feature)

~9500 Cal BP 
(indirect, 
charcoal)

Lipson et al. 2022

HOR- 1 Hora 4 ≥14 yrs Indet N/A N/A Isolated Element Pleistocene This paper
HOR- 1 Hora 5 ≥14 yrs Indet N/A N/A Isolated Element Pleistocene This paper
HOR- 1 Horal 6 ≥14 yrs Indet N/A N/A Isolated Element 

(potential disturbed 
burial)

Pleistocene 
(intrusive into)

This paper

HOR- 1 Hora 7 ≥14 yrs Indet N/A N/A Isolated Element Holocene This paper
HOR- 1 Kahora 1 0– 3 mths M aDNA I19528 Burial (primary  

with missing major 
elements)

~14,000 Cal BP 
(indirect, charcoal)

Lipson et al. 2022

HOR- 1 Kahora 2 7.5±3 mths M aDNA I19529 Burial (primary  
with missing  
major elements)

~16,000 Cal BP 
(indirect, charcoal 
and snail shell)

Lipson et al. 2022

HOR- 1 Kahora 3 2– 7.5 yrs Indet N/A N/A Isolated Element Pleistocene This paper
HOR- 1 Kahora 4 0– 6 mths Indet N/A N/A Isolated Element Pleistocene This paper
HOR- 1 Kahora 5 2– 7.5 yrs Indet N/A N/A Isolated Element Pleistocene This paper
HOR- 1 Kahora 6 Infant/Child Indet N/A N/A Isolated Element Holocene This paper
HOR- 1 Kahora 7 0– 6 mths Indet N/A N/A Isolated Element Holocene This paper
HOR- 5 HOR- 5 

Individual 1
≥4.5 and 
≤7.5±3 mths

Indet N/A N/A Isolated Element Holocene This paper

HOR- 5 HOR- 5 
Individual 2

0– 6 mths Indet N/A N/A Isolated Element Holocene This paper

HOR- 5 HOR- 5 
Individual 3

≥12 yrs Indet N/A N/A Isolated Element Holocene This paper

KAD- 1 KAD- 1 
Individual 1

≥2.5 yrs and 
≤4,5 yrs

Indet N/A N/A Isolated Element Holocene This paper

KAD- 2 KAD- 2 
Individual 1

≤5 yrs Indet N/A N/A Isolated Element Holocene This paper

MAZ- 1 MAZ- 1 
Individual 1

>14 yrs Indet N/A N/A Isolated Element Holocene This paper

MAZ- 1 MAZ- 1 
Individual 2

0– 6 mths Indet N/A N/A Element Cluster 
(potential disturbed 
or secondary burial)

Pleistocene 
(intrusive into?)

This paper

BI_0-0_08_Cerezo-Román_3pp.indd   10BI_0-0_08_Cerezo-Román_3pp.indd   10 27-05-2025   15:37:3827-05-2025   15:37:38

https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3630
https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3630


Cerezo- Román et al. 11

—-1
—0
—+1

previously only reported in terms of their chronology 
and aDNA (Lipson et al. 2022). Here, we describe the 
bio archae ol o gi cal details of Kahora 1 and Kahora 2 for 
the first time (Table 3, Fig. 3, Supplementary Data Set).

Kahora 1 infant burial

This individual was recovered near the start of the 
2019 MALAPP excavation season and was first re-
ported by Lipson et al. (2022). The only remains from 
this individual exported from Malawi were the right 
petrous for aDNA and 14C analysis (no collagen was 
recovered). We estimated the age- at- death for this in-
dividual to be birth ± three months (omnibus skeletal 
estimate) based on the degree of epiphyseal fusion 
and dental development (a crown had formed on the 
mandibular left second deciduous incisor but the 
root had not developed) (AlQahtani et al. 2010; Buik-
stra and Ubelaker 1994). Pathologies or trauma were 
not possible to document. Genetic sex was deter-
mined as male (Lipson et al. 2022).

The individual is nearly complete except for the 
missing lower right tibia and fibula, as well as most of 
the manual and pedal elements. The spatial distribu-
tion of elements indicates a tightly flexed position 
with the head to the south and potentially slightly 
raised above the body based on the distribution of 
crushed cranial fragments on top of the left shoulder 
and thorax region (Fig. 4). Most elements were in an-
atomical position, but the right arm had experienced 
greater postmortem movement at the joints and was 
located to the west of the thorax in association with a 
series of right ribs in anatomical position. Our inter-
pretation is that the individual was initially placed on 
the left side and decomposed in situ. Then, through 
the process of decomposition, the right side of the 
body collapsed to bring the right arm elements and 
right ribcage over to the left side of the body as the 
soft tissue and ligaments decomposed and the skele-
tal elements were displaced (Knüsel and Schotsmans 
2022). This is further supported by the fact that the 

Figure 3. Plan view of the HOR- 1 site showing locations of the Hora 1 
and Hora 2 skeletons recovered by Clark, reconstructed location of his 
excavation grid relative to the MALAPP grid, positions of the Kahora 1 
and Kahora 2 infant burials, and positions of all other human remains 
(white dots). Deeper excavations are darker in shading. The cluster in the 
deepest part is the concentration of burned bone designated as Hora 3.

Figure 4. Plan view of the Kahora 1 and 2 primary inhumations (A), noting that Kahora 2 is nearly 70 cm lower in the sediments than Kahora 1. (B) 
A close- up of the distribution of body regions for Kahora 1 and (C) the same for Kahora 2. Gray outlines in (C) are rocks found in the region ~10 cm 
below the remains.

BI_0-0_08_Cerezo-Román_3pp.indd   11BI_0-0_08_Cerezo-Román_3pp.indd   11 27-05-2025   15:37:4127-05-2025   15:37:41

https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3630


Mortuary Practices in Hunter-Gatherers of Southern-Central Africa12

-1—
0—

+1—

right humerus was slightly elevated above the other 
remains and was therefore the first major element to 
be discovered after the hemimandibles.

The individual was discovered in situ and located 
adjacent to the eroding edge of the previous year’s 
excavation in 2018. We considered if it might be pos-
sible that the right tibia, fibula, and most pedal ele-
ments could have been lost to disturbance or erosion 
between excavation seasons or during section clean-
ings. However, all section cleanings and disturbed/
collapsed profile materials were also sieved and 
completely sorted for both the 3- mm and 1- mm 
sieve and did not result in the recovery of these ele-
ments. This attention to detailed excavation, siev-
ing, and sorting resulted in the recovery of some ray 
elements, which were difficult to identify more spe-
cifically because those in the sieve were out of ana-
tomical context, and the morphology was less 
clearly defined because of the very young age- at- 
death of the individual. Those we did find were well 
preserved and unfragmented, raising the possibility 
of other explanations for missing elements beyond 
taphonomy.

We did not identify any objects that appeared to 
be intentionally placed within the burial or as a part 
of the feature. No burial pit outline or coloration 
was apparent, although the individual must repre-
sent a primary burial to be so complete and well ar-
ticulated. While there was insufficient collagen to 
provide a direct 14C date, antiquity was estimated to 
be ~14,000 years cal. B.P. based on associated radio-
carbon dating of charcoal and giant land snail shell 
from within the burial fill (Lipson et al. 2022: Table 
S4). Although the burial may therefore be younger 
than ~14,000  years old, it cannot be younger than 
~9,500  years old, as it lay entirely below an intact 
cemented ash feature (Lipson et  al. 2022:Supple-
mentary Data).

Kahora 2 infant burial

This individual was fairly complete in terms of skel-
etal representation but fragmentary. Age- at- death 
was estimated to be 7.5 ± 3 months (omnibus skele-
tal age estimate) based on bone growth, develop-
ment, and dental eruption patterns. The maxillary 
right and left first incisors and maxillary second 
right incisors have developed crowns, but the roots 
have not started to develop. The maxillary canines 
and molars also are present, but it was challenging 
to evaluate their exact degree of development be-
cause of poor preservation (AlQahtani et  al. 2010; 
Buikstra and Ubelaker 1994). While teeth were ob-
served in the mandible, the degree of development 

cannot be identified with precision because of ob-
scuring bone. Genetic sex was determined to be 
male (Lipson et al. 2022). No trauma or pathological 
changes were observed based on the initial field re-
view. The spatial distribution of elements indicates a 
primary inhumation on the right side with the head 
to the south, in a flexed and potentially slightly up-
right position, which then experienced postmortem 
disturbance (Fig.  4). Paired elements (radius and 
ulna and tibia and fibula) were tightly associated, 
and it is possible that the knees were touching the 
elbows and the individual was in a curled position. 
The forearms were directly under the collapsed vault 
of the cranium, with the cervical vertebrae and bro-
ken clavicle fragments extending to the west. How-
ever, most of the ribs and vertebrae were missing, 
with only some fragments of the left scapula and 
mostly left ribs recovered. The left lower limb was 
bent at an angle, and the right femur was not in ana-
tomical position with the right tibia and fibula, while 
both innominates and the right scapula were miss-
ing entirely.

Although the elements were fragmentary, the spa-
tial distribution suggests postmortem disturbances 
resulting in loss of elements from the back and pelvic 
region of the skeleton. Three possibilities could be ex-
plored that can contribute to the lack of thoracic and 
pelvic elements from an infant of this age- at- death in 
all the surrounding sediments: (1) poor preservation 
due to enhanced microbial decomposition within the 
abdomen and/or soil pH; (2) absence through natural 
disturbances such as moving water, stratigraphic 
sloping, insect activity, or carnivore activity; or (3) 
intentional or unintentional disturbance and re-
moval by humans. The first option is unlikely because 
bones of similar size and density (ribs from the left 
side, vertebral arches, and centra from other parts of 
the vertebral column) from this individual were pre-
served and recovered in situ. Therefore, preservation 
was unlikely to be affected by soil pH, microbial de-
composition, or exposure to water, including through 
wet- sieving, especially for larger elements like the 
ilia. Thoracic elements that are present were recov-
ered 10 to 30 cm away from the parts of the skeleton 
that were in general articulation, and no other re-
mains from the wide excavated area around and be-
low these could be attributed to this individual. The 
alternative hypotheses are that the removal of ele-
ments more likely occurred instead through carni-
vore scavenging or other natural factors that 
mechanically displaced parts of the skeleton or post-
mortem manipulation by humans. This burial de-
posit was dated by association to a minimum of 
~16,000  years cal. B.P. using giant land snail shell 

BI_0-0_08_Cerezo-Román_3pp.indd   12BI_0-0_08_Cerezo-Román_3pp.indd   12 27-05-2025   15:37:4127-05-2025   15:37:41



Cerezo- Román et al. 13

—-1
—0
—+1

fragments from the feature fill (Lipson et  al. 2022: 
Table S4). As with Kahora 1, the minimum strati-
graphic age is ~9,500 years cal. B.P.

HOR- 1 isolated remains

We identified 55 additional fragments of human bone 
and 12 fragments of potential human bone from 
HOR- 1 in addition to elements assigned to inhuma-
tions (Table 4). Most of these were isolated elements, 
with only minimal clustering apparent (Fig.  5). 
Among the isolated fragments, the most common 
represented elements were the cranial vault, fingers, 
and unspecific long bone fragments. Conservatively, 
the MNI is nine individuals with developmental ages 
of ≥14 years (4), 2– 7.5 years (2), 0 to 6 months (2), and 
≤14 years (1) (Table 1, Table 4). These were calculated 
on the basis of repeating elements, ontogeny, their 
occurrence in either Area I or Area II of the excava-
tion (Fig. 3, Fig. 5), spatial distance from articulated 
individuals of the same developmental age, and/or 
broad chronological associations (e.g., Holocene vs. 
Pleistocene).

Intriguingly, one of the isolated unidentified frag-
ments bears a cut mark (Fig. 7A), although the frag-
ment is not morphologically diagnostic beyond being 
part of a long bone. While most faunal specimens in 
the assemblage are from ungulates, which have dis-
tinctive cortical and trabecular bone qualities, body 
size and morphology make it likely, but not certain, 
that this cut- marked fragment was human.

HOR- 5 bio archae ol o gi cal summary

We identified 12 isolated human elements from the 
site of Hora 5 (HOR- 5). This site was first excavated 
by Sandelowsky in 1967 and did not report any hu-
man remains from an estimated sediment volume of 
~1.4 m3 (Sandelowsky 1972:239). Sandelowsky (1972) 
described all bones from the site as “only minute 
pieces and splinters,” none of which could be identi-
fied. In 2018, MALAPP excavated ~1.0 m3 of sediment 
and recovered numerous faunal remains that are cur-
rently under analysis. The human remains consist of 
two fragments of a deciduous incisor, an indetermi-
nate molar crown, a rib fragment, and eight phalan-
ges (Table 5).

Together, these newly reported remains indicate 
the presence of at least one adult and two non- adults. 
Individual 1 is represented by a well- preserved incisor 
(root ¼ developed) and is estimated to have been an 
individual older than ≥4.5 and ≤7.5 ± 3 months at 
death (AlQahtani et  al. 2010). The molar crown, lo-
cated just over a meter to the north and closer to the 
surface, is also tentatively assigned to Individual 1 

because it has the same ontogenetic age and strati-
graphic provenience in the upper layer dated to the 
Late Holocene based on charcoal fragments (2,946– 
2,777 cal. B.P., 2,800 ± 20, UGAMS- 37844; 2,840– 
2,720 cal. B.P., 2,660±20, UGAMS- 37846; Fig. 6B). The 
incisor was located ~70 cm away horizontally and ~20 
cm above a cluster of infant manual and pedal pha-
langes representing Individual 2, which, together with 
the rib fragment, have an ontogenetic age of zero to 
six months at death and are in a layer with an associ-
ated charcoal age of 4,950 ± 20 (5,718– 5,588 cal. B.P., 
UGAMS- 37845), in the Middle Holocene. The overlap 
in the ranges of age- at- death and the possibility of 
bioturbation between layers makes it possible that all 
the infant remains belong to a single individual, but 
the spatial and chronological differences suggest that 
at least two are present. The adult remains (Individual 
3) consist of two intermediate manual phalanges and 
a proximal manual phalanx recovered from the upper 
layer (Fig. 6). A site plan and profile showing locations 
of the human remains in situ at HOR- 5 is shown in 
Figure 6. All elements except one part of the incisor 
were recovered from the sieve, so all spatial locations 
can only be ascertained to within a 0.5- m × 0.5- m × 
0.05- m volume of sediment.

A notable aspect of the HOR- 5 assemblage is the 
presence of anthropogenic modifications to all the 
non- dental remains. Both the infant and the adult re-
mains were carbonized, and all three adult phalanges 
also exhibit marks consistent with stone tool modifi-
cation (Fig. 7). J.C.T. examined all modifications un-
der a 10– 40× binocular zoom microscope and applied 
criteria from published literature and experience 
with experimental taphonomic assemblages to assess 
if marks might be recent (post- excavation) modifica-
tions, natural bone features such as vascular grooves, 
if they may have occurred post- depositionally within 
the sediments, or if they were more likely by agent(s) 
that modify bone in the perimortem period 
(Fernández- Jalvo and Andrews 2016; Supplementary 
Materials). Although all three elements have marks 
with some attributes suggestive of cut mark mor-
phology (e.g., V- shaped, subparallel grooves with mi-
crostriations), their spatial associations with fracture 
propagation, together with discrete patches of mi-
crostriations, make them more consistent with per-
cussion marks from a directed impact to the bone 
(Blumenschine et al. 1996).

MAZ- 1 bio archae ol o gi cal summary

Mazinga 1 (MAZ- 1) was first excavated in 2017 by 
MALAPP. Thirty- three isolated human elements and 
four potential human elements were recovered from 
4 m3 of sediment excavated between 2017 and 2019 

BI_0-0_08_Cerezo-Román_3pp.indd   13BI_0-0_08_Cerezo-Román_3pp.indd   13 27-05-2025   15:37:4127-05-2025   15:37:41

https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3628
https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3628


Mortuary Practices in Hunter-Gatherers of Southern-Central Africa14

-1—
0—

+1—

Ta
b

le
 4

. 
Su

m
m

ar
y 

of
 is

ol
at

ed
 e

le
m

en
ts

 fr
om

 H
O

R-
 1 (

th
e 

sit
e)

, o
rg

an
iz

ed
 b

y 
in

di
vi

du
al

 (H
or

a 
4,

 K
ah

or
a 

3,
 e

tc
.).

IN
D

IV
ID

U
A

L

R
EG

IO
N

—
  B

on
e

H
or

a 
4

H
or

a 
5

H
or

a 
6

H
or

a 
7

K
ah

or
a 

3
K

ah
or

a 
4

K
ah

or
a 

5
K

ah
or

a 
6

K
ah

or
a 

7
U

na
ss

ig
ne

d
TO

TA
L

H
EA

D
C

ra
ni

al
0

1
0

0
0

0
0

0
0

0
1

C
ra

ni
al

 —
  F

ro
nt

al
0

0
0

0
5

0
0

0
0

0
5

C
ra

ni
al

 —
  T

em
po

ra
l —

  M
as

to
id

0
1

0
0

0
0

0
0

0
0

1
C

ra
ni

al
 —

  P
ar

ie
ta

l
0

0
0

0
1

0
0

0
1

0
2

H
em

im
an

di
bl

e
0

0
0

0
1

0
0

0
1

0
2

To
ot

h 
—

  In
ci

so
r —

  M
ax

ill
ar

y
0

0
0

0
0

0
1

0
0

0
1

To
ot

h 
—

  C
an

in
e 

—
  M

ax
ill

ar
y

0
1

0
0

1
0

0
0

0
0

2
To

ot
h 

—
  In

ci
so

r —
  M

an
di

bu
la

r
0

0
0

0
0

0
0

0
1

0
1

To
ot

h 
—

  M
ol

ar
 —

  M
an

di
bu

la
r —

  D
ec

id
uo

us
0

0
0

0
1

0
0

0
0

0
1

To
ot

h 
—

  M
ol

ar
0

1
0

0
0

0
0

0
0

0
1

N
EC

K
Ve

rt
eb

ra
 —

  C
er

vi
ca

l
0

0
0

0
0

0
0

0
1

0
1

C
H

ES
T/

U
PP

ER
 B

A
C

K
R

ib
0

1
0

0
1

0
0

0
0

0
2

Ve
rt

eb
ra

 —
  Th

or
ac

ic
0

0
0

0
0

0
0

0
2

0
2

A
R

M
H

um
er

us
0

1
0

0
1

0
0

0
0

0
2

Ra
di

us
0

0
0

0
0

0
0

0
2

0
2

H
A

N
D

M
et

ac
ar

pa
l

0
0

0
0

1
0

0
0

0
0

1
Ph

al
an

x 
—

  M
an

ua
l —

  P
ro

xi
m

al
0

0
0

0
1

0
0

0
0

4
5

Ph
al

an
x 

—
  M

an
ua

l —
  In

te
rm

ed
ia

te
0

0
0

0
2

0
0

0
1

1
4

Ph
al

an
x 

—
  M

an
ua

l D
is

ta
l

0
1

0
0

0
0

0
0

0
0

1
LE

G
S

Fe
m

ur
0

0
0

0
0

0
0

0
1

0
1

Ti
bi

a
1

0
0

0
0

0
0

0
0

0
1

FO
O

T
M

et
at

ar
sa

l
0

0
1

0
0

1
0

0
1

0
3

Ph
al

an
x 

—
  P

ed
al

 —
  P

ro
xi

m
al

0
0

0
0

0
0

0
0

1
0

1
Ph

al
an

x 
—

  P
ed

al
 —

  D
is

ta
l

0
1

0
0

0
0

0
0

0
1

2
R

EG
IO

N
 N

O
N

SP
EC

IF
IC

Ve
rt

eb
ra

0
0

0
0

0
0

0
0

1
0

1
Lo

ng
- B

on
e 

Fr
ag

m
en

t
2

1
0

0
0

0
0

0
0

2
5

Ph
al

an
x

0
0

0
0

0
0

0
0

0
1

1
Ph

al
an

x 
—

  In
de

te
rm

in
at

e 
—

  In
te

rm
ed

ia
te

0
0

0
0

0
0

0
0

0
1

1
Ph

al
an

x 
—

  In
de

te
rm

in
at

e 
—

  D
is

ta
l

0
0

0
0

0
0

0
0

1
0

1
TO

TA
L

3
8

1
1

14
2

1
1

12
12

55

BI_0-0_08_Cerezo-Román_3pp.indd   14BI_0-0_08_Cerezo-Román_3pp.indd   14 27-05-2025   15:37:4127-05-2025   15:37:41



Cerezo- Román et al. 15

—-1
—0
—+1

Figure 5. Plan (top) and profile (bottom) views of the locations of adult (A = Area I; B = Area II) and non- adult (C = Area I; D = Area II) remains at 
HOR- 1. Profiles are viewed to the west; scales same as plan view. P = Pleistocene; H = Holocene. Dark/larger and light/small dots are fragments 
definitely/probably from that individual (e.g., large, dark aqua dots in panel A are adult remains assigned to Individual Hora 5 and small, light aqua 
dots in panel A are probable in assignment to that same individual). Light gray stippled background indicates all finds plotted in the excavation as 
non- human remains (e.g., lithics, shells, faunal remains, bone not identifiable to taxon). The tops of the profiles of Areas I and II are of similar 
absolute elevation. Refer to Figure 3 for their true relative horizontal relationship.

Table 5. Summary of isolated elements from HOR- 5 organized by individual. Only fragments with a definite (rather than probable) assignment as 
human are included.

INDIVIDUAL

REGION—  BONE HOR- 5 Indiv. 1 HOR- 5 Indiv. 2 HOR- 5 Indiv. 3 TOTAL

HEAD
Tooth —  Incisor —  Maxillary —  Deciduous 2 0 0 2
Tooth —  Molar 1 0 0 1
CHEST/UPPER BACK
Rib 0 1 0 1
HAND
Phalanx —  Manual —  Proximal 0 2 1 3
Phalanx —  Manual —  Intermediate 0 0 2 2
Phalanx —  Manual —  Distal 0 1 0 1
FOOT
Phalanx —  Pedal —  Distal 0 2 0 2
TOTAL 3 6 3 12

(Table 6); sorting of materials excavated in 2022 and 
2023 is still underway. A site plan and profile show-
ing locations of the human remains we report here in 
situ is shown in Figure 8.

Remains from at least one individual older than 
14  years (Individual 1) and one infant between zero 
and six months (Individual 2) are present (Table  5). 

Individual 1 is represented by a charred and frag-
mented distal humerus with a curved and oblique- 
angled break on the shaft that is typical of bones 
broken while they were in a fresh state (Villa and 
Mahieu 1991). Only 20% of the bone surface was free 
from adhering matrix and therefore available for ex-
amination of modifications, but none were identified. 

BI_0-0_08_Cerezo-Román_3pp.indd   15BI_0-0_08_Cerezo-Román_3pp.indd   15 27-05-2025   15:37:4227-05-2025   15:37:42



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa16

-1—
0—

+1—

Attempts to obtain a direct 14C age were unsuccessful 
given insufficient collagen. A land snail shell found in 
situ within 10 cm of the humerus and at the same 
depth returned an age of 9,010 ± 30 (10,234– 9,916 cal. 
B.P.; UGAMS- 39314). If the humerus represents a dis-
turbed burial, then additional dates are required to 
more precisely ascertain the chronological age of 
deposition. All other adult remains from the Holocene 
layers of this excavation are designated provisionally 
as “Individual 1— Possible” because of their adult sta-
tus, but there is no clear spatial clustering or clear pat-
tern by anatomical region (Fig. 8). In the Pleistocene 
layers, there is a cluster of vertebral fragments together 
with a single manual intermediate phalanx from an 
infant aged zero to six months at death. Unlike the 
faunal remains from this depth, the human remains 
are not fossilized and their surfaces are not heavily en-
crusted. This taphonomic difference in the condition 
of bone surfaces between human and faunal remains 
suggests that the human remains are part of an intru-
sive Holocene burial. Another manual intermediate 

phalanx found approximately 1 m away and at a similar 
depth may be from the same individual. At MAZ- 1, 
Holocene human elements tend to be more common 
toward the eastern margin of the excavation, which also 
is where the humerus was found. This may be further 
evidence of a disturbed burial outside the excavation 
margin. An alternative explanation is that these ele-
ments were part of a secondary burial, which may also 
explain the fragmentation and burning treatments.

Discussion

By combining burial information from older publica-
tions with more recent excavations using fine- detailed 
recovery methods, we show that mortuary practices 
among LSA communities in Malawi and eastern Zam-
bia may be more complex than previously reported 
and that secondary burial and postmortem manipula-
tion should not be ignored as a possible explanation 
for fragmentary and isolated remains. This seems 

Figure 6. Plan (A) and (B) profile views of MALAPP excavations and recovered remains by individual, relative to a digitized plan of the HOR- 5 site 
from Sandelowsky (1972:224). Individuals are different colors. Dark colors represent elements used to generate the MNI and/or other elements 
known from articulation and/or refitting to be from the same individual. Light colors represent elements potentially from the same individual on 
the basis of ontogenetic age and stratigraphic location only (e.g., red dots identify the presence of the HOR- 5 Individual 1 and pink dots are only 
potentially from the HOR- 5 Individual 1). The profile view in (B) uses small black dots to represent plotted finds that are not human remains. These 
define the shape of the excavation and the overall density of finds.

BI_0-0_08_Cerezo-Román_3pp.indd   16BI_0-0_08_Cerezo-Román_3pp.indd   16 27-05-2025   15:37:4327-05-2025   15:37:43



Cerezo- Román et al. 17

—-1
—0
—+1

Figure 7. Stone tool– marked elements from HOR- 1 and HOR- 5. (a) 60086 Long bone fragment from HOR- 1 Pleistocene layers. (b) 54639  
Intermediate manual phalanx from HOR- 5. (c) 53693 Proximal manual phalanx from HOR- 5. (d) 54638 Intermediate manual phalanx from HOR- 5. 
Macro images are taken at 20 and closeup images are at 50 unless otherwise specified.

Table 6. Summary of isolated elements from MAZ- 1 organized by individual. Only fragments with a definite (rather than probable) assignment as 
human are included.

INDIVIDUAL

REGION —  BONE MAZ- 1 Indiv. 1 MAZ- 1 Indiv. 2 Unassigned TOTAL

HEAD
Cranial —  Occipital 1 0 0 1
Tooth —  Molar —  Maxillary 1 0 0 1
Tooth —  Premolar/Molar 1 0 0 1
NECK
Vertebra —  Cervical 0 2 0 2
CHEST/UPPER BACK
Rib 1 0 0 2
Vertebra —  Thoracic 1 1 0 1
LOWER BACK
Vertebra —  Lumbar 0 8 0 8
ARM
Humerus 1 0 0 1
Ulna 1 0 0 1
HAND
Phalanx —  Manual —  Proximal 0 0 1 1
Phalanx —  Manual —  Intermediate 1 2 0 3
Phalanx —  Manual —  Distal 3 0 0 3
FOOT
Metatarsal 2 0 0 2
Phalanx —  Pedal —  Proximal 1 0 0 1
REGION NONSPECIFIC
Vertebra 0 3 0 3
Long- Bone Fragment 2 0 0 2
TOTAL 16 16 1 33

BI_0-0_08_Cerezo-Román_3pp.indd   17BI_0-0_08_Cerezo-Román_3pp.indd   17 27-05-2025   15:37:4527-05-2025   15:37:45



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa18

-1—
0—

+1—

especially plausible (but not excluding other lines of 
evidence) if the skeletal elements present displacement 
of possible anthropogenic origin (e.g., bone piles, cal-
caneus placed on the area of the patella) and cut marks 
and breaks suggest some sort of body manipulation. 
Here, we center on the decedents, the deposits, how 
the individuals were treated, and the social signifi-
cance of the place. Based on some ethnographic  
records (Woodburn 1982) and previously known 
practices of primary inhumation during the LSA in 
south- central Africa, burials were a regular part of the 

bio archae ol o gi cal record of the broader region. Even 
more broadly, archaeologists working at many sites in 
South Africa have found primary inhumations in rock 
shelters (Hall 2000; Parkington et  al. 1980; Pearce 
2008). Here, we add to the discourse by presenting new 
bio archae ol o gi cal evidence from the Kasitu Valley 
documenting the occurrence of (1) possible intentional 
removal/displacement of remains via exposure prior to 
burial and/or reopening of primary burials; (2) over-
representation of certain skeletal elements among 
the isolated remains, suggesting possible intentional 

Figure 8. Plan (A) and (B) profile views of MALAPP excavations and recovered remains by individual. Dark colors represent elements used to 
generate the MNI and/or other elements known from articulation and/or refitting to be from the same individual. Light colors represent elements 
potentially from the same individual on the basis of ontogenetic age and stratigraphic location only (e.g., red dots identify the presence of the 
MAZ- 1 Individual 1 and pink dots are only potentially from the HOR- 5 Individual 1). The profile view in (B) uses small black dots to represent 
plotted finds that are not human remains. These define the shape of the excavation and the overall density of finds.

BI_0-0_08_Cerezo-Román_3pp.indd   18BI_0-0_08_Cerezo-Román_3pp.indd   18 27-05-2025   15:37:4627-05-2025   15:37:46



Cerezo- Román et al. 19

—-1
—0
—+1

curation and inclusion in the sites; and (3) clear modi-
fication of some remains in the form of cut marks.

Removal/displacement of remains

At HOR- 1, the primary adult burials (Hora 1 and Hora 
2) and the infant burials (Kahora 1 and Kahora 2) were 
placed in the grave as complete bodies and decom-
posed in situ based on the osteological and archaeo-
logical evidence. However, at least three of the four 
primary inhumations also show evidence of body ma-
nipulation. For example, the missing lower right tibia 
and fibula of Kahora 1 do not appear to be the result of 
taphonomy or post- depositional processes because a 
thorough excavation of the surrounding area com-
bined with highly precise recovery methods failed to 
recover any evidence of these remains while other ele-
ments were relatively well preserved, even for infant 
remains. In the Kahora 2 infant burial, partially artic-
ulated limbs were not in anatomical position, there 
was displacement of shoulder and vertebral elements 
many tens of centimeters away from their original 
points of articulation with the other remains, and 
most of the lumbar and pelvic elements were missing. 
As with Kahora 1, the surrounding excavation encom-
passed an area large enough and deep enough (at least 
50 cm horizontally in all directions and 15 cm verti-
cally below the remains) to indicate that the missing 
elements were not simply displaced slightly away from 
the main mortuary feature.

In the case of the Hora 2 adult female burial,  
Clark (1956:107) makes a strong case for postmortem 
manipulation by documenting the numerous miss-
ing hand bones, the absence of the left fibula (which 
should have been protected as the individual was ly-
ing flexed on their left side), and the near- complete 
absence of the feet with the exception of the calca-
neus and three other ankle bones of the right foot. 
Furthermore, the right calcaneus was moved to the 
left patella position, as documented in the in situ 
photograph published by Clark and reproduced in 
Figure  2. While Clark (1956:107) interpreted this as 
evidence that the body had been subjected to “some 
exposure before burial took place,” the corollary is 
that the individual’s bones were manipulated after 
the body decomposed by reopening the grave.

The adult male Hora 1 was also missing hand bones, 
although there is no known surviving photograph of 
the individual in situ to assess if element loss was po-
tentially attributable to non- mortuary practices such 
as fragmentation in the ground, coarse recovery meth-
ods, transport, or curation. While small hand bones 
and foot bones are more easily displaced by rodents 
and/or affected by formation processes (Haglund and 
Sorg 2002; Pokines et al. 2022), this would not be the 

case with entire elements and major long bones such as 
Hora 2’s missing feet and fibula. If a carnivore had ac-
cessed this element and removed it while the body was 
in situ, there would not be undisturbed long bones in 
articulation on top. Considering formation processes, 
taphonomy, and placement of all primary burials at 
HOR- 1, a plausible explanation is that body manipula-
tion regularly occurred at the site through reposition-
ing and removal of skeletal elements.

At minimum, we must consider whether the miss-
ing elements (particularly of the long bones) from the 
Hora 2 and Kahora 1 primary inhumations result from 
intentional anthropogenic disturbance and removal. 
Although the sample size is small, there is consistency 
in the missing elements in that they are from the lower 
limb bones, hands, and feet. This suggests a common 
set of processes were in operation, with four possible 
hypotheses. First, elements may have never been re-
covered, or they may not have been recognized by the 
analysts. We reject these hypotheses for Hora 2 be-
cause unlike with Hora 1, details of missing elements 
were reported by the excavator and documented with 
a photograph of the remains in situ (Clark 1956; Fig. 2). 
We also reject the possibility of lack of recovery for 
Kahora 1, because all sediments from MALAPP exca-
vations at HOR- 1 were water sieved through nested 3- 
mm and 1- mm sieves and sorted, without exception. 
Similar preservation of infant remains, including 
bones of similar size, thickness, and density, found in 
situ and in the sieve suggests no destruction occurred 
through wet- sieving or that elements were not pre-
served due to the pH of the soil. Tiny well- preserved 
elements such as distal phalanges were identified 
during sorting, which makes it unlikely that large ele-
ments such as a tibia and fibula would be missed.

Taphonomic factors such as scavenging, sediment 
compaction, and/or selective dissolution are not satis-
factory explanations for some of the missing elements. 
Hora 2’s missing left fibula would have been lying un-
derneath the other lower limb elements based on body 
position, and moving water or rodent burrows would 
not selectively move only a few bones while keeping 
the remainder in articulation or relocating only the 
calcaneus to the patellar position. For Kahora 1, the 
missing elements are on the right side and would have 
been lying on top, so a scavenger may have been more 
able to access them from above the burial pit. How-
ever, if this happened, then it did not disturb the ana-
tomical position of the other elements, including the 
femur. Instead, both individuals appear to have de-
composed in situ and then been slightly disturbed or 
partially disarticulated, with elements removed and 
others repositioned at some point after burial.

For other sites, non- anthropogenic explanations are 
more plausible. For example, the infant remains from 

BI_0-0_08_Cerezo-Román_3pp.indd   19BI_0-0_08_Cerezo-Román_3pp.indd   19 27-05-2025   15:37:4627-05-2025   15:37:46



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa20

-1—
0—

+1—

MAZ- 1 are more suggestive of either a previous inhu-
mation that was removed or heavily disturbed while in 
a partially decomposed state or a secondary deposi-
tion of elements. Vertebral elements from the cervical, 
thoracic, and lumbar regions are all present, but these 
fragments only represent a small part of the total ver-
tebrae in a complete individual. Other elements from 
the skeleton are either missing or spatially displaced 
by more than 1 m in any horizontal direction. This is 
because only one other element (a phalanx) was found 
in the rest of the excavated area. Unlike the adult re-
mains from the Holocene layers, the infant remains 
are clustered and represent the tiniest ossified ele-
ments from an individual this young. Because verte-
brae are among the elements least resistant to 
density- mediated attrition (Lam et  al. 2003), we do 
not consider sediment compaction or dissolution of 
the more robust limb bones to be a likely explanation 
for this pattern. Access to a partially or completely de-
composed body, which was then removed from the 
site by either carnivores or people, could account for 
the disappearance of these elements.

Deviations from expected representation  
of certain elements

Another line of evidence that may suggest some  
skeletal elements were deliberately removed is the 
overrepresentation of teeth and phalanges relative to 
other skeletal parts among isolated elements from  
all of the sites (Tables  4, 5, 6, and Supplementary 
Data). While this could be explained by their highly 
identifiable morphology leading to collection bias, 
this is unlikely based on site recovery methods that 
employed consistent wet- sieving and considered less 
diagnostic elements such as fragmented long bone 
shafts. Although we also searched for these during 
sorting, we found only a very small number that 
could be tentatively identified as human. Moreover, 
our excavations at HOR- 1, HOR- 5, and MAZ- 1 were 
extended laterally enough to be confident that most 
remains not assigned to a primary inhumation truly 
were isolated and not part of a disturbed feature with 
elements displaced to fall outside the excavation area.

We must then consider how the isolated remains 
entered the archaeological records at this site (i.e., 
through intentional collection and mortuary deposi-
tion or through more quotidian means). Across all 
remains recently excavated from all five sites (i.e., not 
including the Hora 1 and Hora 2 adult burials, be-
cause we cannot quantitatively assess element repre-
sentation for them based on published reports or 
discount the possibility of element loss after recov-
ery), we found a total of 20 teeth and 58 phalanges. 
Three of the phalanges present anthropogenic 

modifications (see below). While deciduous teeth are 
naturally shed and teeth in general are commonly 
lost through disease and trauma, this is less so the 
case with fingers and toes. Forensic literature and ex-
perimental studies also show that fingers and toes are 
lost late in the sequence of decomposition and/or car-
nivore scavenging (Darwent and Lyman 2002; Pick-
ering 2001; Pokines and Kerbis Peterhans 2007; 
Pokines et al. 2022). The fact that so many more pha-
langes were found relative to teeth, which tend to be 
the most identifiable and well- preserved human re-
mains in the archaeological record, raises the possi-
bility that phalanges, particularly the ones with 
anthropogenic modifications, were intentionally 
brought to and interred at the site as part of mortuary 
practices.

Anthropogenic modifications to elements

Detailed microscopic study of the bone surfaces of 
the primary inhumations at HOR- 1 is still underway; 
these represent 885 of the 1,006 elements listed in the 
Supplementary Data Set. The balance of the sample 
(N = 121) is from the isolated remains, which we have 
examined for bone surface modifications. Of these, 
only 67 (55%) have at least half the surface unen-
crusted by sandy matrix and visible for observation. 
Three of these are phalanges from HOR- 5 that retain 
stone tool marks from defleshing, disarticulation, 
percussion, or a combination of these based on the 
direction of the marks and their characteristics 
(Fig. 7). These are not flesh- bearing bones that would 
be useful for nutritive cannibalism, which indicates 
to us that the purpose was likely related to mortuary 
practice or other symbolic behaviors. We suggest that 
these actions, combined with the other lines of evi-
dence presented here, are most consistent with be-
haviors associated with curating and subsequently 
transporting the remains for the purpose of second-
ary mortuary deposits.

Contextualizing mortuary practices  
in the regional LSA record

Consistent with our reporting here, archaeological 
human remains from southern- central Africa have 
been historically reported as complete but disturbed 
primary burials or unassociated (isolated) fragments 
(Clark 1973; Fagan and Van Noten, 1971; Phillipson 
1976; Sandelowsky and Robinson 1968). At Fingira, 
the northernmost site in our literature review, Sande-
lowsky (1972; Sandelowsky and Robinson 1968) found 
only one individual with any evidence of articula-
tion, while remains of at least 15 others were scattered 
near the shelter wall and represented by a few elements. 

BI_0-0_08_Cerezo-Román_3pp.indd   20BI_0-0_08_Cerezo-Román_3pp.indd   20 27-05-2025   15:37:4627-05-2025   15:37:46

https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3630
https://journals.upress.ufl.edu/bioarchaeology/article/view/3077/3630


Cerezo- Román et al. 21

—-1
—0
—+1

Sandelowsky (1972:121) describes the state of the par-
tially articulated skeleton as follows: “Many of the 
bones had been broken, in some cases shattered. 
Stones lay above the fractures of the right shoulder, 
arm, and leg, and two breaks in the bones of the left 
leg. All these fractures must have occurred while the 
flesh still held the bones together, and the body must 
have been covered by earth before the decaying bones 
could scatter.”

The image published in the dissertation does not 
show any hand or foot elements (Sandelowsky 
1972:121, Plate 8), and a memoir later published by 
Sandelowsky (2004:16) states that “there was no trace 
of hand or foot bones.” The missing elements cannot 
be explained through rough recovery methods be-
cause the memoir further describes “a great deal of 
blowing and very light brushing” during excavation 
of the remains (Sandelowsky 2004:14). Targeted disso-
lution or fragmentation of only hands and feet is fur-
ther unlikely because the bones “were so hard that we 
thought they could well be semi- fossilised” (Sande-
lowsky 2004:16), and Brothwell and Molleson’s report 
in the dissertation (Sandelowsky 1972:405) confirms 
that “cleaning off hard calcareous matrix” was neces-
sary prior to analysis. Post- depositional disturbance 
therefore appears to be the most likely explanation.

Original excavations at Fingira measured ~10 m2 
and reached a maximum depth of 1.5 m, which is not 
a large space to contain the remains of at least 16 indi-
viduals. Therefore, the fragmented and isolated state 
of other remains may be attributable to anthropo-
genic disturbance from later burials. This appears 
less likely for the Fingira 1 individual, who was miss-
ing both hand and foot elements while retaining 
most other bones in articulation (Sandelowsky 1972, 
2004). Rock collapse and sediment movement could 
explain displacement, but not complete removal, of 
only these body parts. The two remaining possibili-
ties are partial access to the cadaver by scavenging 
animals or intentional mortuary behavior.

Brothwell and Molleson (in Sandelowsky 1972:429) 
offer evidence for the scavenging possibility by de-
scribing the left tibia as “gnawed by animals” and 
only describing the presence of a few distal limb ele-
ments (left radius, proximal right ulna, distal left 
ulna, and complete right tibia). However, the report 
does not provide a complete element inventory for 
this individual, and further inconsistencies make it 
difficult to evaluate if this is the only explanation. For 
example, the image in Sandelowsky (1972:121, Plate 8) 
shows the right tibia and fibula in articulation, but no 
fibulae are mentioned in the report. The report attri-
butes a measurement on a left calcaneus to this indi-
vidual, although the original excavator states that 
feet were not present. Transfer of the remains to 

London occurred in three batches, providing ample 
opportunity for element loss or mix- up.

Approximately 300 km to the south in central Ma-
lawi and eastern Zambia, there is more evidence for 
intentional commingling of individuals in the same 
primary or secondary inhumation (Mtuzi, Thandwe) 
and instances of interments consisting primarily of 
cranial elements (Mtuzi, Kalemba). Only isolated ele-
ments were recovered at Makwe: one incisor, three 
molars, and one radius fragment (Phillipson 1976). 
Thandwe contained a primary burial of a female and 
a second individual that Phillipson (1976) suggests 
was partially disarticulated or dismembered because 
the pelvis and lumbar vertebrae were not recovered 
(Phillipson 1976:50). The remains were associated 
with stone “slabs,” with a second burial of a male in-
dividual on top of the slab covering the remains of 
the female. Next to the male burial were the crushed 
skull bones and very fragmentary postcranial re-
mains from a child (Phillipson 1976).

At Kalemba rock shelter, Phillipson (1976:168– 169) 
reports a total of five individuals mainly represented 
by parts of the skull and cervical vertebrae. Only in 
one case were significant postcranial remains pre-
served, but they were fragmentary and incomplete 
(Table 2). Phillipson (1976:168– 169) also suggests that 
in each case, there were signs that the skull had been 
broken before burial and the pieces separated. The 
only deposit of postcranial remains consisted of frag-
mented, burned bones, potentially evidence for a sec-
ondary cremation deposit. Phillipson (1976:168– 169) 
suggests that the dismemberment and burial of heads 
was a ritual process; he also proposes that the con-
texts may be indicators of cannibalism.

At Mtuzi, in the central region of Malawi ~175 km 
east of Kalemba, Thandwe, and Makwe, Mgomezulu 
(1978) reports that at least six sets of adult male re-
mains were recovered in three separate burials. Mgo-
mezulu (1978) suggests that the “bodies were left in 
the shelter. Later the remains of the bodies in various 
stages of preservation were found by relatives, and it 
is possible that this could explain why several body 
parts are missing . . .  the pile of long limb bones was 
made from bare bones which, when buried, received 
little or no disturbance from the revisits of animals. 
It is easier to make such a pile from bare and partly 
disconnected bones than from connected bone sur-
rounded by the flesh” (Mgomezulu 1978:106– 107).

Similar deposits of human remains with missing 
elements and fragmented burials were found in the 
Thandwe rock shelter (Mgomezulu 1978; Phillipson 
1976). Mgomezulu (1978) mentions that individuals 
may have died close to rock shelters under various 
circumstances, such as warfare between groups or at-
tacks by wild animals. Wild animals such as hyenas 

BI_0-0_08_Cerezo-Román_3pp.indd   21BI_0-0_08_Cerezo-Román_3pp.indd   21 27-05-2025   15:37:4627-05-2025   15:37:46



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa22

-1—
0—

+1—

and vultures may prey on the bodies after death, but 
no mention of animal puncture marks or animal 
damage is described in the publications. Animal 
puncture marks and damage are common in individ-
uals who die in open areas when their bodies are not 
recovered immediately (Beck et al. 2015; Roberts and 
Ingham 2008; Suckling et al. 2016), and future work 
that involves microscopic examination of the sur-
faces of the Mtuzi bones may provide more informa-
tion. Archaeological research on early Holocene 
hunter- gatherer sites in Somalia also found similar 
patterns of missing hands and feet and isolated clus-
ters of teeth (Brant 1988), suggesting a potential ar-
chaeological case of intentional removal of these 
elements in a different African forager context.

Death and meaning in the LSA of Malawi  
and Zambia

We can see that mortuary practices in the LSA of 
southern- central Africa had some commonalities, 
such as placing the dead in rock shelters. Individuals 
undertaking residential or other activities in those 
spaces may have encountered remains and removed 
bones and repositioned them, potentially even thou-
sands of years after initial inhumation. Archaeologi-
cal studies have suggested that the interments of  
a sedentary group tend to symbolize corporate mem-
bership and rights to land and inheritance (McAnany 
1995). Parker Pearson (2008:141) suggests that “fixing 
the dead in the land is a social and political act which 
ensures access and rights over natural resources.” 
These ideas highlight a cross- cultural pattern previ-
ously proposed by Arthur Saxe (1970) and reformu-
lated by Lynne Goldstein (1976), in which formal 
disposal areas such as cemeteries are used by corpo-
rate groups to claim ancestral ties to the ancestors 
and control access to crucial but restricted resources 
and territories. While this cross- cultural pattern 
usually applies to agricultural groups, Charles and 
Buikstra (1983) applied these ideas of territoriality to 
hunter- gatherer groups in an interesting way. They 
proposed and tested that formal corpse- disposal ar-
eas can reflect a corporate unit reflecting corporate 
behavior, and this corporate behavior can be tied to 
territorial behavior. We suggest that these ideas can 
be used for LSA groups in Africa, which are generally 
inferred to have been residentially mobile hunters 
and gatherers (Kusimba 2005; Lipson et  al. 2022; 
Pfeiffer and Harrington 2018).

It is highly plausible that forager groups considered 
the shelters to be important landmarks and part of 
the territory they used. The small amount of archaeo-
logical research in the region overall makes it diffi-
cult to tie these patterns to other aspects of forager 

lifeways, although new zooarchaeological data from 
HOR- 1 suggest small hunting ranges (Bertacchi et al. 
2025). Ancient DNA from individuals at Fingira and 
HOR- 1 provide some effective population size esti-
mates between ~300 and 1,000 individuals (Lipson 
et  al. 2022: Extended Data Figure  9). Together with 
data from other ancient foragers in Malawi, Zambia, 
Tanzania, and Kenya, these ancient genetic data sets 
show that in the terminal Pleistocene to Holocene, 
groups of foragers across southern- central Africa ex-
changed genes mainly within geographic scales of 
100 to 300 km, consistent with the ethnolinguistic 
territory sizes of modern non- equestrian warm- 
climate hunter- gatherers (Marlowe 2005). The impli-
cation is that at the time HOR- 1 was occupied, 
genetically and potentially also culturally differenti-
ated groups of hunter- gatherers lived in the region. 
Because treatment of the dead is deeply culturally 
embedded, variation in mortuary practice may offer 
unique insight into a period of time when ethnolin-
guistic identities among central African foragers 
were becoming more strongly expressed.

While there might be other areas in the landscape 
to deposit the dead, we center the discussion on the 
rock shelters as these types of places are where  
human remains have been more often found and  
excavated. They occur because of rocky outcrops  
on inselbergs, which are largely unvegetated stone 
escarpments that rise abruptly from the surround-
ing landscape and can be seen in distinctive profile 
from many tens of kilometers away (Fig. 9). In addi-
tion to their practical uses (e.g., shelter, vantage 
points), multiple lines of evidence point to their so-
cial significance. Their deposits often contain hun-
dreds of personal ornaments, including debris from 
ornament manufacture on site (Miller et  al. 2021). 
Malawi’s extensive rock art record always occurs in 
shelters (Malijani 2019; Smith 1995). Food- producing 
societies that now live in Malawi also associate ear-
lier populations of foragers with mountains and 
caves in their oral histories (Rangely 1952, 1963; Zu-
bieta 2016). Through careful excavation and docu-
mentation, we are now able to add the mortuary 
record to an emerging understanding of these places 
as part of the socio- symbolic landscape, highlight-
ing their importance through time and among di-
verse groups.

There is considerable time depth (sometimes on the 
order of thousands of years), over which these places 
were repeatedly used for burial. Direct dating of re-
mains we recovered at Fingira showed that the site 
was used for mortuary purposes minimally at ~6,100 
and ~2,500 cal. B.P. (Skoglund et al. 2017). These re-
mains lacked the hard matrix on bones recovered by 
earlier excavations, suggesting some may be even 

BI_0-0_08_Cerezo-Román_3pp.indd   22BI_0-0_08_Cerezo-Román_3pp.indd   22 27-05-2025   15:37:4627-05-2025   15:37:46



Cerezo- Román et al. 23

—-1
—0
—+1

older (Brothwell and Molleson, in Sandelowsky 
1972:405). In addition to stratigraphic placement at 
multiple sites that suggests long- term use, remains 
from HOR- 1 have been directly dated to ~8,200 and 
~9,000 cal. B.P. and through close association with 
dated materials to ~14,000 and ~16,000 (Lipson et al. 
2022; Skoglund et  al. 2017). Our finding that many 
more individuals are represented by isolated remains 
than by primary inhumations reveals these sites as 

greater loci of mortuary behavior than previously 
recognized. People from across the entire commu-
nity structure appear to have been included, as shown 
by the variable developmental age and biological sex 
distributions, and the apparent paucity of specific 
grave objects. At the same time, there is subregional 
variability that may yet be revealed through detailed 
excavation, fine sieving, and careful sorting of all os-
teological materials.

Figure 9. Location of Fingira Rock (top) and Hora Mountain (bottom) relative to landscape features. Insets show their unique local prominence as 
steeply rising inselbergs (~90 m and ~250 m, respectively). There are no similarly large landmarks in the local viewsheds. Image credits: Google 
Earth, Jacob Davis, and Chelsea Smith.

BI_0-0_08_Cerezo-Román_3pp.indd   23BI_0-0_08_Cerezo-Román_3pp.indd   23 27-05-2025   15:37:4727-05-2025   15:37:47



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa24

-1—
0—

+1—

We are not necessarily suggesting that observed 
patterns correlate to territoriality or control of these 
areas on the landscape. However, it is possible that 
in using these locations as homes, shelters, and 
places to bury their dead, LSA foragers created 
memories and established connections to the sites in 
various ways that could be related to family histories 
and ancestral places. The space was used and revis-
ited, creating individual and collective memories 
through time. Creating “persistent places” over time 
through burial practices has been suggested for mo-
bile early food producers elsewhere in Africa, nota-
bly across the Sahara and eastern Africa (di Lernia 
2013; di Lernia and Tafuri 2013; Hildebrand et  al. 
2018; Sawchuk et al. 2018). Among megalithic “pillar 
sites” associated with early herders in Kenya’s Tur-
kana Basin, secondary burial and the inclusion of 
isolated remains (especially fingers and teeth) in an-
thropogenic mortuary cavities is commonplace 
(Sawchuk et al. 2018; Sawchuk et al. 2019; Sawchuk 
et al. 2022). There is no reason to assume, a priori, 
that such complex mortuary behaviors would not 
have deeper roots among LSA foragers and predate 
food production.

Interpreting this evidence, we infer that variation 
in LSA mortuary rituals in south- central Africa fo-
cused on the body itself. While people inhabiting 
these rock shelters may have regularly encountered 
remains and even repositioned them, this did not re-
sult in indiscriminate scattering of elements. Further-
more, it is possible that additional isolated remains 
were brought to these places, potentially as part of 
secondary burial deposits. We are not claiming or 
even suggesting that this is the case for every isolated 
human element found in these sites; post- depositional 
disturbances and taphonomic issues were surely im-
portant factors. Rather, we are opening the possibility 
that evidence of postmortem manipulation in the form 
of intentional removal of skeletal elements, the selec-
tion of particular skeletal elements for burials, and 
even the cutting and burning of the remains as part of 
funeral ceremonies must also be considered when re-
constructing site histories. Our findings are consistent 
with reports of previously excavated sites in the study 
region, which also show highly variable evidence of 
body manipulation and/or secondary treatment.

It is possible that variation in body treatment could 
be used to display and distinguish the identity of each 
deceased individual, as well as the mourners. Particu-
lar body treatments could show the social relations of 
these individuals or be associated with belonging, 
whereby returning with a piece of human remains the 
mourners claim a connection to the specific commu-
nity and their ancestors (Cerezo- Román 2014; Chap-
man and Gaydarska 2007; Geller 2012; Liston 2007; 

McAnany 1995; Rakita 2009). Chapman and Gay-
darska (2007) suggest that through the fragmentation 
process and secondary treatment, a connection be-
tween the objects, the living community, and the an-
cestors can be created and maintained. Secondary 
treatment in the form of exhumation, grave reuse and 
reopening, skeletal relics of the decedent, or skull re-
moval had been documented in different groups 
through time in Africa related to ancestral veneration 
(e.g., Chilver 1965; David 1992; Fagan 1969; Insoll 2016; 
Madden 1940; Muller 1976; Siiriäinen 1977; Volavka 
1998). Variation among mortuary behaviors suggests 
a lack of a unifying way of treating the dead across the 
spans of space and time represented in our study. This 
implies freedom by the mourner to follow their family 
or group preferences but within a more widespread 
tradition of posthumous body manipulation. Al-
though time depth likely accounts for some of this 
variation, examination of the published record to-
gether with our new data from the Kasitu Valley offers 
hints of geographic patterning. Complete primary in-
humations with missing and likely removal hand and 
foot skeletal elements are more common in the north-
ern part of the study, while the southern part features 
more secondary and/or commingled burials that  
include crania. This may indicate some degree of  
regional cultural continuity in mortuary practice, 
which is consistent with genetic evidence for LSA 
groups choosing partners within their immediate 
geographic region during the terminal Pleistocene 
and Holocene (Lipson et al. 2022).

The use of sites over thousands of years as living 
areas (territories/investment in “places”) and the 
deposition of “tokens” or parts of people from else-
where seems relevant. There are many reasons why 
groups practice secondary burials of a few or one 
skeletal elements related to different ideas of person-
hood, ancestral veneration, symbolic representation 
of the decedent, acts of purification, and acts of dese-
cration, among others (Cerezo- Román 2015; Cerezo- 
Román et al. 2017; Graham 2009; Graham et al. 2018; 
Liston 2007). We suggest that perhaps in the instances 
that we found, “token” or secondary burials could 
have had symbolic power as a source of remembrance 
of the deceased and the connections and networks 
that previously existed between the deceased and 
mourners. They may also connect with specific places 
on the landscape that are both prominent visual 
landmarks and specific points of social memory.

Conclusions

Here, we show with careful recovery and complete 
sorting of osteological remains from new excavations 

BI_0-0_08_Cerezo-Román_3pp.indd   24BI_0-0_08_Cerezo-Román_3pp.indd   24 27-05-2025   15:37:4727-05-2025   15:37:47



Cerezo- Román et al. 25

—-1
—0
—+1

in the Mzimba District of Malawi that there are an-
cient human remains in every excavated rock shelter, 
reiterating the important role of these places as re-
positories for the dead among ancient African forag-
ers. We also show that there is biased skeletal part 
representation that is, in some cases, best explained 
through anthropogenic removal and deposition of 
remains, and for at least some of these, there is evi-
dence of defleshing and/or fragmentation using stone 
tools. This supports a conclusion that intentional 
postmortem body manipulation was practiced in the 
area and sets the scene for broader interpretations of 
mortuary ritual and the embodiment of memory.

We caution, however, that it is necessary to go 
back and reanalyze other human remains from sites 
in the region that reported intentional and exten-
sive postmortem treatment and evaluate the pres-
ence or absence of non- anthropogenic damage 
alongside new analyses of anthropogenic modifica-
tions. There is good archaeological evidence of non- 
anthropogenic factors that contribute to the finding 
of human remains in rock shelters, such as redepos-
ited human bones, disturbance, and reasons not as-
sociated with secondary burials in rock shelters at 
several LSA sites in South Africa (Hall 2000; Park-
ington et al. 1980; Pearce 2008). Although the South 
African record is relatively geographically distant 
from Malawi, it offers the closest available data set 
for comparison. This approach may also be a fruit-
ful avenue of research for ancient forager remains 
from northern Africa, which also show substantial 
evidence for intentional disturbance of burials and 
movement of bodies or body parts into specific shel-
tered locations with long records of human occupa-
tion (Haverkort and Lubell 1999). Some patterns, 
however, may only be possible to understand with 
new excavations that maintain careful spatial con-
trol and documentation.

Our work shows that an excavation, recovery, and 
analytical focus on more obvious mortuary features, 
such as burials, may have rendered invisible other 
important and meaningful cultural practices. Evi-
dence of selective removal and transport of elements 
requires careful consideration and analysis of the 
data and can reveal a tradition of embodied memory- 
making that carries the relationship between com-
munity members and the bodies of their dead beyond 
the brief moment of interment. It draws out the dura-
tion of the relationship between the body of the de-
ceased and the people tending to their remains, 
effectively extending the mortuary ritual over time. It 
also extends the ritual spatially, across the broader 
cultural landscape (Watts et al. 2020). This interpre-
tive approach offers a richer view of the complexity of 
ancient forager social and symbolic worlds than 

simply analyzing more complete burial features as 
the physical traces of singular, unassociated events 
and isolated remains as random collections of ele-
ments. Our findings illustrate a need for more in- 
depth discussions about mortuary complexity about 
ancient foragers in south- central Africa and beyond.

Acknowledgments

We thank the Malawi Department of Museums and 
Monuments for their long- term collaboration and 
permission to do this work. We are especially grateful 
to Inkosi ya Makhosi M’mbelwa V, Inkosi Chindi, 
Inkosi Kampingo Sibande, Inkosana Thomas Nkosi, 
the Mzimba Heritage Association, and the communi-
ties around the sites for welcoming the research for so 
many years. We also thank our representatives from 
the Department of Museums and Monuments, the 
community field and field lab crew, and student par-
ticipants for their careful eyes and many hours spent 
processing finds. Particular thanks are owed to Fred-
rick Mapemba and Raphael Mwafulirwa for their  
operation of the field lab. Menno Welling and Jacob 
Davis were of invaluable logistical and intellectual 
support. Inset images in Figure  9 are by Jacob  
Davis and Chelsea Smith. Fieldwork was funded by 
Wenner- Gren Foundation Grant 9432, National Geo-
graphic Society Award NGS- 53412R- 18, Emory Uni-
versity, Yale University, and Hyde Family Foundations. 
Financial support was provided by the University of 
Oklahoma Libraries’ Open Access Fund and the De-
partment of Anthropology. The Society of Malawi 
gave permission to reproduce the original image of 
the Hora 2 burial from Clark (1956). Three anony-
mous reviewers and an associate editor provided con-
structive feedback on an initial submission.

Ethics Statement

The two HOR- 1 non- adult burials (Kahora 1 and Ka-
hora 2) were recovered in 2019 and are curated by the 
Malawi Department of Museums and Monuments 
(formerly Department of Antiquities). Additional re-
mains from HOR- 1, HOR- 5, and MAZ- 1 were recov-
ered in 2017, 2018, and 2019. Permission for the 
research was provided under permits A/III/3.3/70,  
A/III/3.3/71, AD/23/56, and NCST/RTT/2/6. Export 
was provided under A/1/1/1/3.6/50, A/1/1/1/3.6/44,  
A/II/1.5/33, and MHQ/CUL/1/04/2. Spatial coordi-
nates in the Supplementary Data Set have been mod-
ified from the original UTM to obscure exact site 
locations. They are correct relative to one another 
within each site but do not tie to real- world locations.

BI_0-0_08_Cerezo-Román_3pp.indd   25BI_0-0_08_Cerezo-Román_3pp.indd   25 27-05-2025   15:37:4727-05-2025   15:37:47



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa26

-1—
0—

+1—

References

AlQahtani, S. J., M. P. Hector, and H. M. Liversidge. 2010. Brief 
communication: The London atlas of human tooth develop-
ment and eruption. American Journal of Physical Anthropol-
ogy 142:481– 490.

Arizona State Museum, University of Arizona. 2018. ASM Oste-
ology Recording Packet. Arizona State Museum, University of 
Arizona, Tucson. https:// statemuseum . arizona . edu / file / 712 
[accessed 1 January 2018].

Beck, Jess, Ian Ostericher, Gregory Sollish, and Jason De León. 
2015. Animal scavenging and scattering and the implications 
for documenting the deaths of undocumented border crossers 
in the Sonoran Desert. Journal of Forensic Sciences 60:S11– S20.

Bertacchi, Alex, Potiphar Kaliba, and Jessica C. Thompson. 
2025. Short-distance hunting strategies of Late Quaternary 
foragers in the miombo woodlands of Malawi. Journal of An-
thropological Archaeology 77:101656. https://doi.org/10.1016/j 
.jaa.2025.101656

Blumenschine, Robert J., Curtis W. Marean, and Salvatore D. 
Capaldo. 1996. Blind tests of inter- analyst correspondence 
and accuracy in the identification of cut marks, percussion 
marks, and carnivore tooth marks on bone surfaces. Journal 
of Archaeological Science 23:493– 507.

Brandt, Steven A. 1988. Early Holocene mortuary practices and 
hunter- gatherer adaptations in southern Somalia. World Ar-
chaeology 20(1):40– 56.

Buikstra, Jane E. 1991. Out of the appendix and into the dirt: 
Comments on thirteen years of bio archae ol o gi cal research. In 
What Mean These Bones? Studies in Southeastern Bioarchaeol-
ogy, edited by Mary L. Powell, Patricia S. Bridges, and Ann 
Marie Wagner Mires. The University of Alabama Press, Tus-
caloosa, pp. 172– 188.

Buikstra, Jane E., and Douglas H. Ubelaker. 1994. Standards for 
Data Collection from Human Skeletal Remains: Proceeding of 
a Seminar at the Field Museum of Natural History. Arkansas 
Archeological Survey Research Series No. 44, Fayetteville.

Cerezo- Román, Jessica I. 2015. Unpacking personhood and fu-
nerary customs in the Hohokam area of Southern Arizona. 
American Antiquity 80:353– 375.

Cerezo- Román, Jessica I., Anna Wessman, and Howard Williams, 
eds. 2017. Cremation and the Archaeology of Death. Oxford  
University Press, Oxford.

Chapman, John, and Bisserka Gaydarska. 2007. Parts and Wholes: 
Fragmentation in Prehistoric Context. Oxbow Books, Oxford.

Charles, Douglas K., and Jane E. Buikstra. 1983. Archaic mortu-
ary sites in the central Mississippi drainage: Distribution, 
structure, and behavioral implications. In Archaic Hunters and 
Gatherers in the American Midwest, edited by James L. Phillips 
and James A. Brown. Academic Press, New York, pp. 117– 145.

Chávez Balderas, Ximena. 2018. Fire, transformation, and bone 
relics: Cremated remains at the Templo Mayor of Tenochtitlan. 
In Smoke, Flames, and the Human Body in Mesoamerican  
Ritual Practice, edited by Vera Tiesler and Andrew K. Scherer. 
Dumbarton Oaks Research Library and Collection, Washing-
ton, D.C., pp. 379– 409.

Chilver, E. M. 1965. Meta village chiefdoms of the Bome Valley 
in the Bamenda Prefecture of West Cameroon, Part 1. Nige-
rian Field 30(1):4– 18.

Clark, J. Desmond. 1956. Prehistory in Nyasaland. The Nyasa-
land Journal 9:92– 119. 

Clark, J. Desmond. 1973. Archaeological investigation of a 
Painted Rock Shelter at Mwana Wa Chencherere, north of 
Dedza, Central Malaŵi in July to September, 1972. The Society 
of Malawi Journal 26:28– 46. 

Crader, Diana Catherine. 1984a. Hunters in Iron Age Malaŵi: 
The Zooarchaeology of Chencherere Rockshelter. Malaŵi  
Government Ministry of Education and Culture, Lilongwe.

Crader, Diana Catherine. 1984b. Faunal remains from Chencher-
ere II Rock Shelter, Malawi. The South African Archaeological 
Bulletin 39:37.

Cunningham, Craig, Louise Scheuer, and Sue M. Black. 2016. 
Developmental Juvenile Osteology. 2nd  ed. Elsevier/AP, 
Amsterdam.

Darwent, Christalynn M., and R. Lee Lyman. 2002. Detecting the 
postburial fragmentation of carpals, tarsals and phalanges. In 
Advances in Forensic Taphonomy: Method, Theory, and Archae-
ological Perspectives, edited by William D. Haglund and  
Marcella H. Sorg. CRC Press, Boca Raton, FL, pp. 356– 377.

David, Nicholas. 1992. The archaeology of ideology: Mortuary 
practices in the central Mandara Highlands, northern Cam-
eroon. In An Africa Commitment, edited by Judith Sterner 
and Nicholas David, University of Calgary Press, Calgary, 
pp. 181– 210.

DeBusk, George H. 1998. A 37,500- year pollen record from Lake 
Malawi and implications for the biogeography of Afromon-
tane Forests. Journal of Biogeography 25:479– 500.

di Lernia, Savino. 2013. Places, monuments, and landscape:  
Evidence from the Holocene central Sahara. Azania: Archaeo-
logical Research in Africa 48(2):173– 192.

di Lernia, Savino, and Mary Anne Tafuri. 2013. Persistent death-
places and mobile landmarks: The Holocene mortuary and 
isotopic record from Wadi Takarkori (SW Libya). Journal of 
Anthropological Archaeology 32(1):1– 15.

Duday, Henri. 2009. The Archaeology of the Dead: Lectures in Ar-
chaeothanatology. Oxbow Books, Oxford.

Edlich- Muth, Miriam. 2022. Dissolving subjects in medieval reli-
quaries and twentieth- century mass graves. In Interdisciplinary 
Explorations of Postmortem Interaction: Dead Bodies, Funerary 
Objects, and Burial Spaces Through Texts and Time, edited by  
Estella Weiss- Krejci, Sebastian Becker, and Philip Schwyzer. 
Springer, Cham, pp. 189– 209.

Fagan, B. M. 1969. Excavations at Ingombe Ilede, 1960- 2. In Iron 
Age Cultures in Zambia, Volume 2, edited by B. M. Fagan, D. 
Phillipson, and S. G. H. Daniels. Chatto and Windus, Lon-
don, pp. 57– 161.

Fagan, Brian M., and Francis L. Van Noten. 1971. The Hunter- 
Gatherers of Gwisho (Zambia). Musee Royal de l’Afrique  
Centrale, Tervuren, Belgique.

Fernández- Jalvo, Yolanda, and Peter Andrews. 2016. Atlas of Ta-
phonomic Identifications: 1001+ Images of Fossil and Recent 
Mammal Bone Modification. Springer, Dordrecht.

Geller, Pamela L. 2012. Parting (with) the dead: Body partibility 
as evidence of commoner ancestor veneration. Ancient Meso-
america 23(1):115– 130.

Goldstein, Lynne. 1976. Spatial Structure and Social Organiza-
tion: Regional Manifestations of Mississippian Society. Ph.D. 
dissertation, Northwestern University, Evanston, IL.

Graham, Emma-Jayne. 2009. Becoming persons, becoming an-
cestors: Personhood, memory and the corpse in Roman rituals 
of social remembrance. Archaeological Dialogues 16(1):51– 74.

Graham, Emma-Jayne, Carrie L. Sulosky Weaver, and Andrew T. 
Chamberlain. 2019. Pars pro toto and personhood in Roman 
cremation ritual: New bio archae ol o gi cal evidence for the rite of 
os resectum. Bioarchaeology International 2(4):240– 254.

Haglund, William D. and Marcella H. Sorg, eds. 2002. Advances 
in Forensic Taphonomy Method, Theory, and Archaeological 
Perspectives. CRC Press, Boca Raton, FL.

Hall, Simon. 2000. Burial and sequence in the Later Stone Age of 
the Eastern Cape Province, South Africa. The South African 
Archaeological Bulletin 55(172):137– 146.

BI_0-0_08_Cerezo-Román_3pp.indd   26BI_0-0_08_Cerezo-Román_3pp.indd   26 27-05-2025   15:37:4727-05-2025   15:37:47



Cerezo- Román et al. 27

—-1
—0
—+1

Haverkort, Caroline M., and David Lubell. 1999. Cutmarks on 
Capsian human remains: Implications for Maghreb Holocene 
social organization and palaeoeconomy. International Jour-
nal of Osteoarchaeology 9:147– 169.

Hildebrand, Elisabeth A., Katherine M. Grillo, Elizabeth A. 
Sawchuk, Susan K. Pfeiffer, Lawrence B. Conyers, Steven T. 
Goldstein, et al. 2018. A monumental cemetery built by east-
ern Africa’s first herders near Lake Turkana, Kenya. Proceed-
ings of the National Academy of Sciences 115(36):8942– 8947. 
https://doi.org/10.1073/pnas.1721975115.

Hogg, Alan G., Timothy J. Heaton, Quan Hua, Jonathan G. 
Palmer, Chris S. Turney, John Southon, et al. 2020. SHCal20 
Southern Hemisphere calibration, 0– 55,000 years cal BP. Ra-
diocarbon 62:759– 778. 

Insoll, Timothy. 2015. Constructing ancestors in sub- Saharan 
Africa. In Death Rituals, Social Order and the Archaeology of 
Immortality in the Ancient World: ‘Death Shall Have No Do-
minion’, edited by Colin Renfrew, Michael J. Boyd, and Iain 
Morley. Cambridge University Press, Cambridge, pp. 153– 167.

King, Julia  A. 2016. Comparative colonialism and collections- 
based archaeological research: Dig less, catalog more. Mu-
seum Worlds 4(1):4– 17.

Kjellström, A. 2022. From saint to anthropological specimen: The 
transformation of the alleged skeletal remains of Saint Erik. In 
Interdisciplinary Explorations of Postmortem Interaction: Dead 
Bodies, Funerary Objects, and Burial Spaces Through Texts and 
Time, edited by Estella Weiss- Krejci, Sebastian Becker, and 
Philip Schwyzer, Springer, Cham, pp. 167– 188.

Knüsel, Christopher J., and Eline M. J. Schotsmans, eds. 2022. 
The Routledge Handbook of Archaeothanatology: Bioarchaeol-
ogy of Mortuary Behaviour. Routledge, London.

Knüsel, Christopher J. 2014. Crouching in fear: Terms of engage-
ment for funerary remains. Journal of Social Archaeology 
14:26– 58.

Kristeva, Julia. 1980. Pouvoirs de l’horreur: Essai sur l’abjection. 
Éditions du Seuil, Paris.

Kusimba, Sibel B. 2005. What is a hunter- gatherer? Variation in 
the archaeological record of Eastern and Southern Africa. 
Journal of Archaeological Research 13(4):337– 366.

Lam, Y. M., O. M. Pearson, Curtis  W. Marean, and Xingbin 
Chen. 2003. Bone density studies in zooarchaeology. Journal 
of Archaeological Science 30(12):1701– 1708.

Lipson, Mark, Elizabeth A. Sawchuk, Jessica C. Thompson,  
Jonas Oppenheimer, Christian A. Tryon, Kathryn L. Ran-
horn, et al. 2022. Ancient DNA and deep population structure 
in sub- Saharan African foragers. Nature 603:290– 296.

Liston, Maria A. 2007. Secondary cremation burials at Kavousi 
Vronda, Crete: Symbolic representation in mortuary practice. 
Hesperia: The Journal of the American School of Classical 
Studies at Athens 76(1):57– 71.

Madden, J. F. 1940. The exhumation of Latuka Rain Queen. Su-
dan Notes and Records 23:351– 354.

Malijani, Oris. 2019. Towards community inclusion: Chongoni 
world heritage site and sustainable cultural tourism in Ma-
lawi. South African Archaeological Bulletin 74(209):35– 41.

Marlowe, Frank W. 2005. “Hunter‐gatherers and human evolu-
tion.” Evolutionary Anthropology: Issues, News, and Reviews: 
Issues, News, and Reviews 14(2):54– 67.

McAnany, Patricia A. 1995. Living with the Ancestors: Kinship and 
Kingship in Ancient Maya Society. University of Texas Press, 
Austin.

McClelland, John, and Jessica I. Cerezo- Román. 2016. Person-
hood and re- embodiment in osteological practice. In Dealing 
with the Dead: Mortuary Archaeology and Contemporary So-
ciety, edited by Howard Williams and Melanie Giles. Oxford 
University Press, Oxford, pp. 39– 67.

Mgomezulu, Gadi Godfrey Yobe. 1978. Food Production: The Be-
ginnings in the Lithipe/Changoni Area of Dedza District, Ma-
lawi. 411. University of California, Berkeley.

Muller, Jean-Claude. 1976. Of souls and bones: The living and 
the dead among the Rukuba, Benue- Plateau State, Nigeria.  
Africa 46:258– 273.

Nilsson Stutz, Liv. 2003. Embodied Rituals and Ritualized Bod-
ies: Tracing Ritual Practices in Late Mesolithic Burials. 
Almqvist & Wiksell International, Stockholm.

Nilsson Stutz, Liv. 2016. The importance of “getting it right”: Trac-
ing anxiety in Mesolithic burial rituals. In The Archaeology of 
Anxiety: The Materiality of Anxiousness, Worry, and Fear, edited 
by Jeffrey Fleisher and Neil Norman. Springer, New York,  
pp. 21– 40.

Parker Pearson, Mike. 2008. The Archaeology of Death and 
Burial. Texas A & M University Press, College Station.

Parkington, John, Charles Cable, P. L. Carter, H. J. Deacon,  
Janette Deacon, A. J. B. Humphreys, et  al. 1980. Time and 
place: Some observations on spatial and temporal patterning 
in the Later Stone Age sequence in Southern Africa [with 
Comments and Reply]. The South African Archaeological Bul-
letin 35(132):73– 112.

Pearce, David. 2008. Later Stone Age Burial Practice in the East-
ern Cape Province, South Africa. Ph.D. dissertation, Univer-
sity of the Witwatersrand, Johannesburg, South Africa.

Pfeiffer, Susan, and Lesley Harrington. 2018. Regional continuity 
and local challenges to resilience among Holocene hunter- 
gatherers of the Greater Cape Floristic Region, South Africa. In 
Hunter- Gatherer Adaptation and Resilience. A Bio archae ol o gi cal 
Perspective, edited by Daniel H. Temple and Christopher M. Sto-
janowski. Cambridge University Press, Cambridge, pp. 26– 46.

Phillipson, D. W. 1976. The Prehistory of Eastern Zambia. British 
Institute in Eastern Africa, Nairobi.

Pickering, T. R. 2001. Carnivore voiding: A taphonomic process 
with the potential for the deposition of forensic evidence. 
Journal of Forensic Sciences 46(2):406– 411.

Pokines, James T., Ericka N. L’Abbe, and Steven A. Symes, eds. 
2021. Manual of Forensic Taphonomy. 2nd ed. CRC Press, Boca 
Raton, FL.

Pokines, James T., and Julian C. Kerbis Peterhans. 2007. Spotted 
hyena (Crocuta crocuta) den use and taphonomy in the Masai 
Mara National Reserve, Kenya. Journal of Archaeological Sci-
ence 34(11):1914– 1931.

Rakita, Gordon F. M. 2009. Ancestors and Elites: Emergent Com-
plexity and Ritual Practices in the Casas Grandes Polity.  
AltaMira, Lanham, Maryland.

Rangeley, W. H. 1952. Two Nyasaland rain shrines. Nyasaland 
Journal 5(2):31– 50.

Rangeley, W. H. J. 1963. The earliest inhabitants of Nyasaland. 
The Nyasaland Journal 16(2):35– 42.

Roberts, C., and S. Ingham. 2008. Using ancient DNA analysis 
in palaeopathology: A critical analysis of published papers, 
with recommendations for future work. International Journal 
of Osteoarchaeology 18:600– 613.

Roksandic, Mirjana. 2002. Position of skeletal remains as key to 
understanding mortuary behavior. In Advances in Forensic 
Taphonomy: Method, Theory and Archaeological Perspectives, 
edited by William D. Haglund and Marcella H. Sorg. CRC 
Press, Boca Raton, FL, pp. 99– 117.

Sandelowsky, Beatrice Henriette. 1972. Later Stone Age Lithic 
Assemblages from Malawi and Their Technologies. Ph.D. dis-
sertation, University of California, Berkeley.

Sandelowsky, Beatrice Henriette. 2004. Archaeologically Yours, 
Beatrice Sandelowsky: A Personal Journey into the Prehistory 
of Southern Africa, in Particular Namibia. Namibia Scientific 
Society, Windhoek.

BI_0-0_08_Cerezo-Román_3pp.indd   27BI_0-0_08_Cerezo-Román_3pp.indd   27 27-05-2025   15:37:4727-05-2025   15:37:47



Mortuary Practices in Hunter-Gatherers of Southern-Central Africa28

-1—
0—

+1—

Sandelowsky, B. H., and K. R. Robinson. 1968. Fingira: Prelimi-
nary Report, Publication (Malawi. Department of Antiquities) 
No. 3. Government Press, Zomba, Malŵi.

Sawchuk, Elizabeth A., Steven T. Goldstein, Katherine M. Grillo, 
and Elizabeth A. Hildebrand. 2018. Cemeteries on a moving 
frontier: Mortuary practices and the spread of pastoralism 
from the Sahara into eastern Africa. Journal of Anthropological 
Archaeology 51:187– 205. https://doi.org/10.1016/j.jaa.2018.08.001.

Sawchuk, Elizabeth  A., Elisabeth  A. Hildebrand, Austin Chad 
Hill, Daniel A. Contreras, Justus Erus Edung, Anneke Janzen, 
et al. 2022. The Jarigole mortuary tradition reconsidered. An-
tiquity 96(390):1460– 1477.

Sawchuk, Elizabeth  A., Susan Pfeiffer, Carla  E. Klehm, Mi-
chelle E. Cameron, Austin C. Hill, Anneke Janzen, et al. 2019. 
The bioarchaeology of mid- Holocene pastoralist cemeteries 
west of Lake Turkana, Kenya. Archaeological and Anthropo-
logical Sciences 11(11):6221– 6241.

Saxe, Arthur A. 1970. Social Dimensions of Mortuary Practices. 
Ph.D. dissertation, University of Michigan, Ann Arbor.

Siiriäinen, Ari. 1977. Later Stone Age investigation in the Laikipia 
Highlands, Kenya: A preliminary report. Azania 12:161– 186.

Skoglund, Pontus, Jessica C. Thompson, Mary E. Prendergast, 
Alissa Mittnik, Kendra Sirak, Mateja Hajdinjak, et  al. 2017. 
Reconstructing prehistoric African population structure. Cell 
171:59– 71.e21.

Smith, Benjamin W. 1995. Rock Art in South- Central Africa: A 
Study Based on the Pictographs of Dedza District, Malawi and 
Kasam District, Zambia. Ph.D. dissertation, University of 
Cambridge, Cambridge.

Suckling, Joanna K., M. Katherine Spradley, and Kanya Godde. 
2016. A longitudinal study on human outdoor decomposition 
in central Texas. Journal of Forensic Sciences 61:19– 25.

Villa, Paola and Eric Mahieu. 1991. Breakage patterns of human 
long bones. Journal of Human Evolution 21:27– 48.

Volavka, Zdenka. 1998. Crown and Ritual. University of Toronto 
Press, Toronto.

Watts, Rebecca, Maurice Mugabowagahunde, André Ntagwabira, 
and John Giblin. 2020. Deposition of modified human remains 
as evidence for complex mortuary treatment in East Africa 
during the first millennium AD. International Journal of Osteo-
archaeology 30:824– 834.

Woodburn, James. 1982. Egalitarian societies. Man 17:431.
Wright, David K., Sarah J. Ivory, Jago J. Birk, Jeong-Heon Choi, 

Benjamin Davies, Sabine Fiedler, et al. 2024. Palaeoenviron-
mental data indicate late quaternary anthropogenic impacts 
on vegetation and landscapes in Mzimba, northern Malawi. 
Frontiers in Environmental Archaeology 2:1– 20.

Zubieta, Leslie F. 2016. Learning through practise: Cheŵa wom-
en’s roles and the use of rock art in passing on cultural knowl-
edge. Journal of Anthropological Archaeology 43:13– 28.

BI_0-0_08_Cerezo-Román_3pp.indd   28BI_0-0_08_Cerezo-Román_3pp.indd   28 27-05-2025   15:37:4727-05-2025   15:37:47

https://doi.org/10.1016/j.jaa.2018.08.001



