Content uploaded by Toby Heron
Author content
All content in this area was uploaded by Toby Heron on Feb 06, 2021
Content may be subject to copyright.
1884 S/1, 2020Tobias Heron, Headhun�ng in Iron Age Europe : An evolu�onary adapta�on to a con�ngent environment
1884 Rhifyn arbennig
Special issue
Cyfnodolyn Yr Ysgol Hanes, Athroniaeth a Gwyddorau Cymdeithas
Journal of the School of History, Philosophy and Social Sciences
1
2020
Tobias Heron
Headhunting
in Iron Age Europe
An evolutionary adaptation
to a contingent environment
1884
About the author:
Tobias Heron holds a BA and MA in archaeology from Bangor University, where he is currently
studying for his PhD. He has a particular interest in later prehistoric social archaeology, evolutionary
psychology, and economics and has already published articles in these �ields. The study presented in
this special issue was submitted as his MA thesis, which he passed with distinction.
About this book:
The primary aim of this study is to attempt experimentally how evolutionary theory
can be applied to ethnographic analogy in order to construct a model with predictive
power to derive insights about other aspects of societies which practice the same
behaviour, which can then be applied to archaeology. This study will attempt to resolve
this by looking at the cultural practice of headhunting in Iron Age Celtic Europe, and
how a general understanding of the practice – that being, a cross-cultural investigation
of ethnographic and ethnohistorical case studies supported by evolutionary explana-
tions – can be useful not only in creating new potential knowledge in this speci�ic topic,
but with other cultural practices, while also as an experiment in the potential develop-
ment of archaeological theory by applying an independent scienti�ic and objective
basis (evolutionary psychology) to an imperfect method of archaeological theory (ana-
logy).
Rhifyn arbennig
Special issue 1
R
K
R
K
(Print) ISSN 2634-2057
(Online) ISSN 2634-2065
ISSN (print) 2634-2057
ISSN (online) 2634-2065
1884
Rhifyn arbennig / Special Issue No. 1
Tobias Heron
Headhunting
in Iron Age Europe
An evolutionary adaptation
to a contingent environment
2020
Bangor / Gwynedd
1884 Special Issue No. 1
Editors: Mari E. Wiliam, Molly Southward, and Tom Wilkinson-Gamble
First published 2020
by
RK Publications
102 Penrhos Road
Bangor / Gwynedd, LL57 2BQ
United Kingdom
office@rkarl.org
Printed and bound in Germany by
epubli – a service of neopubli GmbH, Berlin
Published under a
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
(CC BY-NC-SA 4.0) license.
ISSN (print) 2634-2057
ISSN (online) 2634-2065
Special Issue No. 1 1884
i
Contents
List of Figures ................................................................................................................................... iii
List of Tables ...................................................................................................................................... v
Introduction ....................................................................................................................................... 1
Iron Age Headhunting in Europe ................................................................................................ 3
The Classical Sources .................................................................................................................................... 3
Polybius ........................................................................................................................................................................ 3
Archaeological Evidence .............................................................................................................................. 6
Head Display .............................................................................................................................................................. 6
Post Conflict Decapitation .................................................................................................................................... 8
Heads and Horses .................................................................................................................................................. 10
Modification and manipulation of heads ..................................................................................................... 11
Southern France ........................................................................................................................................... 12
The Early Statuary ................................................................................................................................................ 13
The Later Statuary ................................................................................................................................................ 14
Headhunting Across the World ................................................................................................ 18
The Irian Jaya ................................................................................................................................................. 18
Settlement and subsistence .............................................................................................................................. 18
Social structure ...................................................................................................................................................... 18
Reproductive system ........................................................................................................................................... 19
Conflict and Warfare ............................................................................................................................................ 19
Headhunting Practice .......................................................................................................................................... 20
Changes ..................................................................................................................................................................... 21
The Naga .......................................................................................................................................................... 21
Settlement and Subsistence .............................................................................................................................. 21
Social structure ...................................................................................................................................................... 22
Reproductive System ........................................................................................................................................... 22
Conflict and Warfare ............................................................................................................................................ 23
Headhunting ............................................................................................................................................................ 23
Changes ..................................................................................................................................................................... 27
The Montenegrins ....................................................................................................................................... 27
Settlement and subsistence .............................................................................................................................. 27
Social structure ...................................................................................................................................................... 27
Reproductive system ........................................................................................................................................... 28
Conflict and Warfare ............................................................................................................................................ 28
Headhunting practice .......................................................................................................................................... 30
1884 Special Issue No. 1
ii
Changes ...................................................................................................................................................................... 30
The Jivaro ........................................................................................................................................................ 30
Settlement and subsistence ............................................................................................................................... 30
Social structure ....................................................................................................................................................... 30
Reproductive system............................................................................................................................................ 31
Conflict and warfare ............................................................................................................................................. 32
Headhunting practice........................................................................................................................................... 32
Changes ...................................................................................................................................................................... 34
The Maori ........................................................................................................................................................ 35
Settlement and subsistence ............................................................................................................................... 35
Societal structure ................................................................................................................................................... 36
Reproductive system............................................................................................................................................ 37
Conflict and warfare ............................................................................................................................................. 37
Headhunting practice........................................................................................................................................... 38
Changes ...................................................................................................................................................................... 39
Analysis............................................................................................................................................................ 40
Evolutionary Psychology and Headhunting ......................................................................... 44
Headhunting as an evolved adaptation? ........................................................................................... 45
Status .......................................................................................................................................................................... 46
Hierarchies: Dominance versus Prestige .......................................................................................... 48
Dominance hierarchies ....................................................................................................................................... 49
Prestige Hierarchies ............................................................................................................................................. 50
The evolutionary explication of headhunting ................................................................................. 52
Applying the Model: Headhunting in the Iron Age through an Evolutionary
Psychological Lens ........................................................................................................................ 56
The Lower Rhône Basin in the Early Iron Age ................................................................................ 56
Southern Britain at the Bronze/Iron Age Transition .................................................................. 59
Headhunting in Central and North-western Iron Age Europe ................................................ 61
Conclusion ........................................................................................................................................ 65
References ........................................................................................................................................ 66
Special Issue No. 1 1884
iii
List of Figures
Fig. 1: Perforated skullcap from Hillhead, Caithness (after Armit 2012, ill. 1.3). ................................... 7
Fig. 2: Perforated cranial fragments from Roissy-en-France (Val d’Oise) 4th century BC (after
Rousseau 2012, fig. 18). ...................................................................................................................................... 7
Fig. 3: Nailed skulls from Puig de Sant Andreu (after Rovira-Hortalà 1999, lam. 2). ............................ 8
Fig. 4: Skull peforated by a nail discovered near Koblenz, Germany (after Rousseau 2012, fig. 24).
....................................................................................................................................................................................... 8
Fig. 5: Upper skull from Glanum with nail piercing in the frontal bone (after Rousseau 2012, fig.
5). ................................................................................................................................................................................. 8
Fig. 6: Nailed skulls (4, left; 2, right) from Entremont (after Mahieu 1998, figs. 1, 2). ......................... 8
Fig. 7: The skulls and fastening devices from the oppidum of La Cloche, c. 2nd century BC (after
Chabot 1982, fig. 12 & Salviat 1972, fig. 12). ............................................................................................. 9
Fig. 8: General plan of Gournay-sur-Aronde (after Brunaux et al. 1985, fig. 35). ................................... 9
Fig. 9: Vertebrae with cut marks indicative of decapitation from Gournay-sur-Aronde, Oise (after
Brunaux & Malagoli 2003, fig. 10). ................................................................................................................. 9
Fig. 10: Plan of Ribemont-sur-Ancre indicating the location of human remains (after Fercoq du
Leslay 2000, fig. 17). ......................................................................................................................................... 10
Fig. 11: (left) Excavation photograph of one of the ossuaries, (right) detailed drawing of one
square metre from another showing iron weaponry (after Cadaux 1984; Fercoq du Leslay
1996 fig. 6; fig. 7). ............................................................................................................................................... 10
Fig. 12: (A) Coin depicting Dumnorix, 1st century BC; (B), Coin of Cunobelinus, 1st century AD.
(after Allen 1980, pl. 22, 317; Creighton 2000, 129). .......................................................................... 10
Fig. 13: Celtic coinage depicting the horse and severed head motif (left) with rider from Limousine,
(right) from Transdanubia (after Allen 1980, pl. 3, 24; pl. 33, 491). ............................................ 10
Fig. 14: Lintel from Nages demonstrating association of dead heads and horses (after Arcelin et al.
1992, fig. 22). ........................................................................................................................................................ 11
Fig. 15: (left) Stone block fragments of the suggested three-sided pillar from Entremont, Provence,
(right), close-up of the fragment with mounted warrior and severed head (after Patrice et
al. 2003, fig. 114; Arcelin et al. 1992, fig. 18). ......................................................................................... 11
Fig. 16: Celtic horseman with possible severed head from Kärlich, Kr. Koblenz, Germany, 3.8cm
(after Megaw & Megaw 1990, fig. 94). ....................................................................................................... 11
Fig. 17: Pottery fragment from Clermont-Ferrand, Aulnat, depicting a mounted warrior with head
hanging from his horse’s neck (after Rousseau 2012, fig. 23). ........................................................ 11
Fig. 18. (left) The bronze sceptre terminal from a tomb in Numantia (Soria), 2nd century BC;
(right) Fibula of mounted rider with severed head from Numantia (Soria) (after Lorrio &
Ruiz Zapatero 2005, figs. 22, 23). ................................................................................................................ 12
Fig. 19: Heinrich Wankel’s skull cup (after Peter-Röcher 1998, abb. 4). ................................................. 12
Fig. 20: Skull ‘masks’ from Montmartin, Oise (left) and the sanctuary at Corent (right) (after
Aldhouse-Green 2002, pl. 15; Rousseau 2012, fig. 12). ...................................................................... 12
Fig. 21: Early Iron Age human head carvings from Entremont, (left) the ‘head pillar,’ (right) two
angles of the bloc aux épis (after Arcelin et al. 2003, fig. 111). ....................................................... 13
Fig. 22: The two complete pillars from Saint-Michel-de-Valbonne, Hyéres, Var (after Brun 1999,
figs. 481, 483). ...................................................................................................................................................... 13
Fig. 23: The carved stone block from Badasset (after Armit 2012, ill. 4.11). ......................................... 14
Fig. 24: Artist’s rendition of how the portico structure at Roquepertuse possibly existed (after
Coignard & Coignard 1991, fig. 11). ............................................................................................................ 14
Fig. 25:Pillar C from Roquepertuse showing the three head niches alongside decorative carvings
and traces of colour (after Barbet 1991, fig. 17). .................................................................................. 14
Fig. 26: Head niches from Roma Villa, Nimes (after Barbet 1992, fig. 1). ............................................... 15
Fig. 27: Carved head niches with carved embossed ‘knuckles’ from Glanon, c. late 2nd/early 1st
century BC (after Arcelin et al. 2003, fig. 116). ...................................................................................... 15
1884 Special Issue No. 1
iv
Fig. 28: Lintel fragment believed to belong to a portico pillar from Entremont, Provence, with
carved head-niche simulacrum (after Arcelin et al. 2003, fig. 112)............................................... 15
Fig. 29: Distribution map of Portico sites in southern France, with niches that could receive real
heads (black dots), or carved simulacra (white dots) (after Arcelin et al. 2003, fig. 110). .. 16
Fig. 30: The seated warrior statues (left) Roquepertuse; (right) Glanon (after Megaw & Megaw
1990, fig. 272; Areclin et al. 1992, fig. 28). ............................................................................................... 16
Fig. 31: Distribution map of ‘Roquepertuse/Glanon’ style seated warrior statues in southern
France (after Armit 2012, ill. 5.10). ............................................................................................................. 16
Fig. 32: Seated statue from Bouriège, Aude with sever-ed head (after Arcelin et al. 2003, fig. 122).
.................................................................................................................................................................................... 17
Fig. 33: A reconstructed composite of one of the seated warrior statuary from Entremont (after
Armit 2010, fig. 9.10). ....................................................................................................................................... 17
Fig. 34: Examples of the individual character afforded to the Entremont warrior statuary
(modified after Arcelin & Rapin 2003, figs. 33, 34, 35). ...................................................................... 17
Fig. 35: Close up of the trophy head statuary from Entremont clearly illustra-ting their individual
nature (after Arcelin et al. 2003, fig. 123). ............................................................................................... 17
Fig. 36: A man wearing head hunter’s boar tusk collar (after Furness 1902, PL. XLIV, 1). .............. 24
Fig. 37: Two young Konyak men with fresh tattoo marks on their cheeks indicating their
achievement of successful headhunters. 1937. (after Jacobs 1990, 115). .................................. 24
Fig. 38: Hairpin used as a tally of number of heads taken. Konyak 1937 (after Jacobs 1990, 120).
.................................................................................................................................................................................... 25
Fig. 39: A rack of 151 skulls from a Konyak village. 1912-23. (after Jacobs 1990, 116). .................. 25
Fig. 40: A ‘head tree’ from a Chang village. 1947 (after Jacobs 1990, 123). ............................................ 26
Fig. 41: Skull trophies displayed on wooden shelves in the morang of a Konyak village. 1918-45
(after Jacobs 1990, 123). .................................................................................................................................. 26
Fig. 42: Prepared human head tsantsa trophy (after Karsten 1923, pl. 6). ............................................. 33
Fig. 43: An early engraving of a pre-colonial Pa demonstrating the concentric defensive palisades
(after Knight 2009, 4). ...................................................................................................................................... 35
Fig. 44: Aerial view of the pre-European Pa Karangaumu, illustrating the terraced hillslopes and
separate platforms encircled by ditches (after Knight 2009, 12). .................................................. 36
Fig. 45: Cross section of the multiple palisaded ditch and bank defences of a traditional Pa (after
Knight 2009, 14). ................................................................................................................................................ 36
Fig. 46: A Mokomokai (after Robley 1896, fig. 148). ........................................................................................ 39
Fig. 47: Middle-range theory in relation to present statics and past dynamics (after Johnson 2000,
fig. 4.1). .................................................................................................................................................................... 44
Fig. 48: Mate preferences for males (left) and women (right) (after Grammar 1996, 81). .............. 47
Fig. 49: Male status variables and lifetime fitness benefits of 21 simple societies (after von Rueden
2014, table 9.1). ................................................................................................................................................... 47
Fig. 50: Model of status and its determinants (after von Ruden 2014, fig. 9.1). ................................... 48
Fig. 51: The chieftain – state continuum (after Otterbein 2009, fig. 6.3). ................................................ 55
Fig. 52: Map of sites in Provence with stelae (after Golosetti 2014, fig. 5.6). ......................................... 58
Fig. 53: Midden sites in the Vale of Pewsey (after Waddington 2009, fig. 4.2). .................................... 60
Fig. 54: Gold Scythian ornament dated to the end of the 4th century BC (after Armit 2012, ill. 2.6).
.................................................................................................................................................................................... 63
Fig. 55: Skull from cemetery at Reims, Marne (after Bonnabel & Boulestin 2008, fig. 2). ................ 64
Special Issue No. 1 1884
v
List of Tables
Tab. 1: (top) The periods where references to headhunting are made, compared to (bottom) the
periods they are written (after Armit 2012, table 2.2). ......................................................................... 3
Tab. 2: Principle Iron Age sites in southern France yielding evidence for Headhunting (excluding
statuary without explicit severed head iconography) (after Armit 2012, table 6.1). ............ 17
Tab. 3: Settlement types. .............................................................................................................................................. 40
Tab. 4: Societal structure. ............................................................................................................................................ 40
Tab. 5: Reproductive systems. ................................................................................................................................... 41
Tab. 6: Conflict systems. ............................................................................................................................................... 41
Tab. 7: Headhunting, Status, and Prestige. ........................................................................................................... 42
Tab. 8: Cultural Integration of Headhunting. ...................................................................................................... 43
1884 Special Issue No. 1
vi
About the Author:
Tobias Heron holds a BA and MA in archaeology from Bangor University, where he is currently
studying for his PhD. He has a particular interest in later prehistoric social archaeology,
evolutionary psychology, and economics and has already published articles in these fields. The
study presented in this special issue was submitted as his MA thesis, which he passed with
distinction.
Special Issue No. 1 1884
Introduction
The primary aim of this study is to attempt experimentally how evolutionary theory can be
applied to ethnographic analogy in order to construct a model with predictive power to derive
insights about other aspects of societies which practice the same behaviour, which can then be
applied to archaeology. This study will attempt to resolve this by looking at the cultural practice
of headhunting in Iron Age Celtic Europe, and how a general understanding of the practice – that
being, a cross-cultural investigation of ethnographic and ethnohistorical case studies supported
by evolutionary explanations – can be useful not only in creating new potential knowledge in this
specific topic, but with other cultural practices, while also as an experiment in the potential
development of archaeological theory by applying an independent scientific and objective basis
(evolutionary psychology) to an imperfect method of archaeological theory (analogy).
The practice of head-taking in human populations is a pandemic phenomenon across time and
space for which we have evidence going back at least 600,000 years ago (White 1986) to modern
day Borneo (Metcalf 1996). Evolved adaptations are generally expected to be universal, and when
one discovers a recurrent theme like headhunting which independently emerges across time and
space, it is highly indicative that a head-taking behaviour is an evolutionary adaptation (Buss
2008, 79). However, unlike smiling/crying, or morning sickness in pregnancy (Buss 2008 78), this
does not occur everywhere all the time which indicates that the practice of head-taking is likely
an evolved adaptation to a particular set of contingent environmental circumstances (Workman
& Reader 2004, 23). Therefore, it should be possible to identify the common environmental
factors from which a headhunting behaviour is adaptive, and can thus emerge. In investigating
headhunting from an evolutionary psychological perspective in this manner, it should be possible
to gain a deeper understanding of the behaviour itself, such as its function, effects, and
development over time concurrent with societal complexity etc., as well as to construct a general
model of the components headhunting societies have, which can be applied to past societies
suspected of headhunting where evidence is ambiguous, where hopefully new inferences can be
made, and/or contribute or detract from existing interpretations/social models about aspects of
said societies which would otherwise remain obscure.
Headhunting in the Celtic Iron Age presents a highly suitable area for which to test this
approach,
1. because of the vast corpus of ethnographic/ethnohistorical literature on headhunting
which provides a solid foundation for comparison and application of evolutionary
psychological theory, and
2. because, while the nature and prevalence of headhunting in Celtic Europe has been a highly
debated topic (Collis 2003, 215; James 1999; contra Laing 1981, 113; Newman 1993, 21;
Enright 2006, 55), discourse has generally been limited to the question of the existence of
the recondite ‘head cult’ and consequent hegemonic belief systems (exception – Armit
2006; 2010; 2012) and not as an avenue of understanding Celtic society, mostly because
this is (to the best of my knowledge) an original methodological approach.
This Study will not therefore try to add to the belaboured deliberation of this clandestine Celtic
head cult, in fact, it is the intent to circumvent this intellectual caltrop and instead explore the
extent to which the claim of headhunting in Central and Western Europe during the Iron Age is
true, its patterns, and what else this can tell us about societies. ‘Headhunting’ has been defined by
Janet Hoskins (1996b, 2) as ‘an organised, coherent form of violence in which the severed head is
given a specific ritual meaning and the act of head-taking is consecrated and commemorated in
some form.’ Ian Armit, understanding that headhunting activity can be secular as well as religious,
defined headhunting as ‘a form of group-sanctioned, ritualised violence, in which the removal of
the human head plays a central role’ (Armit 2012, 11). I find both of these definitions
unsatisfactory, at least for the purpose of this study, because they are inherently tautological and
Heron, Headhunting in Iron Age Europe
2
descriptive, derived from loose observations of the effects common to (some or most)
headhunting societies. One of the primary aims of the study is to identify what these are, and I
would be remiss to commit to such definitions which would ultimately establish an intellectual
foundation drenched in bias. Thus, for the purpose of this study, headhunting will be defined more
broadly and neutrally as: ‘the repeated act of taking human heads by communities for a purpose
other than of itself.’
The structure of this study will be as follows. Chapter one will establish a foundation and
context to what is known of headhunting in the Iron Age of Europe, by reviewing the classical
literary sources, and then supporting them with the archaeological evidence. The archaeological
evidence, as it stands for itself, will also be explored. Chapter two will essentially systematically
investigate headhunting from the ethnographic literature as a pseudo- and micro-cross cultural
study from societies around the world, in order to identify the key themes and patterns of societies
which practice headhunting, and then compare these themes in order to extract the necessary
data, and analysis of it, from which a model can be constructed. Keeping these insights in mind,
chapter three will introduce evolutionary psychological theory. First it will explore how it can be
useful in working with analogy in archaeological and ethnographic interpretation, and then be
applied to the data from chapter two, in order to explain and understand headhunting from an
evolutionary perspective, from which a theory and model will be constructed. Chapter four will
then attempt to test the models validity as well as hopefully garner new knowledge by applying it
back onto ‘Celtic’ Iron Age society.
Special Issue No. 1 1884
Iron Age Headhunting in Europe
From the outset it must be stated that the archaeological evidence of preferential treatment of
the human head and its curation, manipulation, and frequency of recovery therein, does not
necessarily indicate that what was found in the archaeological record derives from the practice of
headhunting. This is an open question which will remain to be debated in perpetuum. To attempt
as such would be as foolhardy as it is futile. Instead the most practical method, I believe, is to first
delineate the most important classical literary sources pertaining to Celtic headhunting and
identify the general and specific claims made, and then look to the archaeological record to
support said claims. Using this approach of keeping in line with the literary sources, any
ambiguous evidence which may suffer the conniption of the most ardent cynic should be at least
be placated by reason to accept that headhunting is, at least, a possible, if not probable
interpretation. The evidence from southern France however is much more certain. Along with it
being the location whence some of the most important sources were written (Benoît 1975, 248),
it contains extraordinary statuary evidence which make the case for headhunting here difficult to
deny. As such the evidence from southern France will be reviewed in its own section.
The Classical Sources
Generally, knowledge of the Celts, or Gauls, from the classical literature come first hand from
the Greek and Roman authors; Polybius, Poseidonius, and Caesar, whom had direct contact with
them (Collis 2003, 25). The information from other authors is often recycled, repeated, and
sometimes plagiarised from these three sources (Collis 2003, 25), which thankfully have
preserved some of the knowledge of the primary sources which did not survive themselves.
Moreover, as Tab. 1 demonstrates, many of the sources are divorced by quite significant time
periods, often writing on events that occurred several centuries prior. This, alongside the usual
admonitory discipline when dealing with such sources, always leaves an avenue for doubt on the
validity and accuracy of the information, however references to the practice of
headhunting/taking are fairly recurrent.
4
th
BC
3
rd
BC
2
nd
BC
1
st
BC
1
st
AD
2
nd
AD
3
rd
AD
∎-----
∎∎∎∎∎∎
∎---∎-
∎
−
−
−
−
−
Tab. 1: (top) The periods where references to headhunting are made, compared to (bottom) the periods they are written
(after Armit 2012, table 2.2).
Polybius
Polybius’ Histories, writing in the second century BC, is the earliest author which describes the
Celts and their practice of headhunting. His works recording from the beginning of the second
Punic War contain three accounts on the matter of Celtic headhunting. The first is the battle
between Rome and the confederation of Celtic tribes in Etruria c. 225 BC, where the Roman consul
Gaius Atilius Regularis is decapitated:
“… in this action Gaius the consul fell in the mellay fighting with desperate courage, and his
head was brought to the Celtic kings” (Polybius, Histories, 2.28.10, trans. Paton 1922).
This terse passage offers little aside from what it actually suggests, that the head was brought
to ‘kings’ as proof of identity of an important individual. The next passage from Polybius regarding
headhunting comes during Hannibal’s march into northern Italy c. 218 BC, where a band of Celts
defects from the Roman army to join Hannibal’s forces:
Heron, Headhunting in Iron Age Europe
4
“The Celts let more than half the night pass, and just about the time of the morning watch
armed themselves and fell upon the Romans who were quartered nearest to them; killed a
considerable number, and wounded not a few; and, finally, cutting off the heads of the slain,
departed with them to join the Carthaginians” (Polybius, Histories, 3.67, trans. Shuckberg
2002).
What can be taken note of here is that there would be no conceivable necessary reason to cut
off the heads of the Roman’s, particularly during a mutiny while they are still within the territory
of their now-affectionately informed enemies, unless this is somewhat of well-established
custom? Conversely it could be Polybius trying to convey the ferocity that the Celts were well
known for in Greece and Rome (Davidson 1993, 3). A passage from Diodorus Siculus writing in
the 1st century BC, also testifies to this practice as being a custom of the Celts, except this time
writing over 300 years earlier in 387 BC on the events of the Battle of Allia where the Senonian
Gaul’s defeated the Roman Army:
“… for the Celts spent the first day cutting off, according to the custom, the heads of the
dead.” (Diodorus Siculus, Bibliotheca Historica, 14.115.5, trans. Oldfather 1946).
Additionally, spending an entire day, if this detail is to be believed, to the cutting of heads
during a campaign, certainly would demand this practice must have been of some significant level
of importance. The last passage from Polybius comes from an account of a Roman punitive
campaign on the Galatian Tolistobogii of Asia Minor, and their King, Ortiagon, in 190 BC:
“It chanced that among the prisoners made when the Romans won the victory at Olympus
over the Gauls of Asia, was Chiomara, wife of Ortiago. The Centurion who had charge of her
availed himself of his chance in soldierly fashion, and violated her. He was a slave indeed both
to lust and money: but eventually his love of money got the upper hand; and, on a large sum of
gold being agreed to be paid for the woman, he led her off to put her to ransom. There being a
river between the two camps, when the Gauls had crossed it, paid the man the money, and
received the woman, she ordered one of them by a nod to strike the Roman as he was in the act
of taking a polite and affectionate farewell of her. The man obeyed, and cut off the centurion’s
head, which she picked up and drove off with, wrapped in the folds of her dress. On reaching
her husband she threw the head at his feet; and when he expressed astonishment and said:
“Wife to keep faith is a good thing,” she replied: “Yes; but it is a better thing that there should
be only one man alive who has lain with me!” (Polybius, Histories, 29.38, trans. Shuckburgh
1889).
This detailed account of individual revenge by a woman of high standing, indicates head-taking
may not be only limited to the context of battle, and if so, suggests the custom is more prolific
throughout Gallic culture. Again we see the utilisation of the head as proof of death and justice
served.
Livy, writing around the turn of the 1st century AD, writes of an episode around seventy years
earlier than Polybius, of the victory of the Senonian Gaul’s over general Scipio and his army at the
siege of Clusium in Etruria in 295 BC:
“Thus the legion was attacked in the rear and found itself surrounded, with the enemy
assailing it on every quarter. Some writers say that the legion was even annihilated there, so
that none survived to bear away the tidings, and that the consuls, who were not far away from
Clusium, got no report of the disaster till some Gallic horsemen came in sight, with heads
hanging at their horses’ breasts or fixed on their lances, and singing their customary song of
triumph.” (Livy, History of Rome, 10.26.11, trans. Foster 1926).
This passage reinforces the post-battle harvest of heads that have been ascribed to the Celts,
but also it denotes the particular details of the hanging of heads from their horses, or impaled on
European Iron Age Headhunting
5
lances, while exultant in celebration. Another important passage from Livy regarding head
treatment describes the decapitation of the Roman general Postumius by the Celtic Boii in 216 BC:
“The Boii stripped his [Postumius’s] body, cut off the head, and carried their spoils in
triumph to the most hallowed of their temples. There they cleaned out the head, as their custom
is, and gilded the skull, which thereafter served them as a holy vessel to pour libations from
and as a drinking-cup for the priest and the temple-attendants.” (Livy, Ab Urbe Condita, 23.24,
trans. De Sélincourt 1965).
This passage is especially informative as it introduces as a custom, the curation and ritual use
of severed heads in combination with decapitation. Moreover it can be inferred that this special
treatment of a severed head may be due to the important status of Postumius as a general, perhaps
to go as far as a symbol of eternal enslavement or humiliation, as is known of the chiefdoms in the
Cauca Valley, Colombia (Chacon & Dye 2007a, 619). The credibility of Livy has been questioned
due to discrepancies between his work and Polybius (Collis 2003, 21-2), and in this case Polybius
mentions nothing more of this event than ‘the Praetor[’s]’ forces being attacked and ‘utterly
annihilated by the Celts” (Histories 3.118), and instead focusing on the events southwards in
Apulia. However Polybius is not the only source Livy was privy to use, and likely drew from
copyists of Poseidonius’ works such as Strabo and Diodorus Siculus (Webster 1996, 118) or even
Fabius Pictor (Armit 2012, 22-3). Moreover accounts testifying to the usage of enemy skulls as
drinking vessels add support to this claim, such as those of the Celtic Scordisci of Thrace by Florus
of the 2nd century AD (Epitome 1.39.2-3), and repeated in the 4th century AD by Ammianus
Marcellinus (Rae Gestae 27.4.4).
The best known authority on the Celts was Poseidonius (c. 135-51 BC) (Dowden 2000, 18). He
was a Stoic philosopher who believed Greece and Rome were in a state of moral decline, and that
the culture of the Celts was closer to that of the ‘Golden Age’ (Cunliffe 1999, 8). This detail is of
potential importance as his admiration of the Celts makes it difficult to levy claims of the usual
barbaric bias in his accounts as he saw them in a positive light, despite finding their use of severed
heads repugnant (Piggott 1968, 82-5). Poseidonius visited southern Gaul after the campaign of
Marius (104-102 BC), where it has been argued he roamed the still un-Romanised tribes of the
Celtic hinterland, much like an ethnographer (Benoît 1975, 248). Unfortunately none of
Poseidonius’ works survive themselves, and only through the works of other authors can his
material be salvaged. One version of Poseidonius’ account on heads from Diodorus Siculus is as
follows:
“They cut off the heads of enemies slain in battle and attach them to the necks of their
horses. The blood-stained spoils they hand over to their attendants and carry off as booty,
while striking up a paean and singing a song of victory, and they nail up these first fruits upon
their houses just as do those who lay low wild animals in certain kinds of hunting. They embalm
in cedar-oil the heads of the most distinguished enemies and preserve them carefully in a chest,
and display them with pride to strangers, saying that for this head, one of their ancestors, or
his father, or the man himself, refused the offer of a large sum of money. They say that some of
them boast that they refused the weight of the head in gold, thus displaying what is only a
barbarous kind of magnanimity; for it is not a sign of nobility to refrain from selling the fruits
of one’s valour, it is rather true that it is bestial to continue one’s hostility against a slain fellow-
man” (Diodorus Siculus, Bibliotheca Historica, 5.29.5, trans. Tierny 1960).
Strabo also has a near identical account (Geographia 4.4.5), with added details of Poseidonious’
disgust which he eventually became used to, that head-takers will not let their heads be redeemed,
even for their weight in gold, and that the Romans put an end to all these customs, as well as their
sacrificial and divinatory practices. Again, we see the specific claim of heads being hung from
horses, but are also referred to as booty, suggesting some token of reward, perhaps material.
Celebrations are again referenced in the context of severed heads, but now we have the detail of
public display of heads upon the houses of individual head-takers. The reference to heads as ‘first-
Heron, Headhunting in Iron Age Europe
6
fruits’ is interesting, as well as the observation of equivalence to head-taking and hunting, which
might allow with some basis the proper usage of the term ‘headhunting.’ Again, we have claims,
and very detailed ones, of a different method of head preservation and curation, and even an
explicit reference to support the tentative interpretation of Polybius’ passage above of
preservation only being worthy of distinguished individuals. Reference to ‘attendants’ also
implies some formal hierarchy. The empennage of the passage may be particularly significant and
presents an interesting interpretative quandary. At face value, it denotes that curated heads are
highly valued objects of prestige and will not be ‘sold’ (or ‘redeemed’, in Strabo, Geographia 4.4.5),
even for their weight in gold. It has to be inferred then, that there were people trying to acquire
these heads. Brunaux (1988, 110) contends that there must have been a market for preserved
heads, perhaps having a ‘magical’ power too. Jackson (1964, 37) offers an alternative
interpretation, that the people wanting the heads were perhaps relatives of the victims. Which
interpretation is more appropriate depends on if one takes Diodorus’ account where heads are
sought to be ‘sold’, or Strabo’s with where they are attempted to be ‘redeemed.’
Collating the sources presented, we can distinguish several common themes regarding the
Celts and the practice of taking heads:
General claims:
• The custom of taking heads post-conflict.
• Ritual and religious use or heads, including songs, dances, and celebrations, and association
in ‘temple’ contexts by ‘priests.’
• The public display of severed heads.
• Differential treatment of heads - Preservation and curation, particularly afforded to high
status individuals.
• Heads having a prestige value worth more than material wealth, and victims status tied to
their value.
• Identity of heads important, whether for proof of death or valour.
Specific claims:
• Hanging heads from their horses’ necks.
• Turning enemies head’s into drinking vessels.
The next section will attempt to validate the claims that are possible from the archaeological
evidence.
Archaeological Evidence
Head Display
From the British Isles, evidence for display of heads has been interpreted at the Iron Age camp
at Bredon Hill, Gloucestershire (Hencken 1938). Three crania and six mandibles were recovered
from the debris of the collapsed gate following most likely Roman capture of the settlement,
suggesting that these were remains of trophy heads displayed above the gateway (Hencken 1938,
57). A similar situation existed at the enormous Iron Age complex of Stanwick, Yorks. (Wheeler
1954), which is believed to be the main stronghold of the Brigantes in c. 1st century AD. On
Wheeler’s site B, a waterlogged ditch terminal adjacent to one of the gateways produced the
cranial remains of a middle-aged man, including the mandible and cervical vertebrae which is
suggestive that the head was fleshed at the time of deposition. Major cut marks suggestive of
wounds inflicted by a bladed instrument at time of death support this further, which altogether
led Wheeler (1954, 53-5) to interpret this deposition as the head of a warrior killed in battle,
whence his head was displayed as a trophy on the adjacent gateway before it later tumbled into
the ditch. This is mirrored at the Iron Age site of Dinorben, where along with cranial fragments
found in three houses and one of the guard chambers at the main entrance, a mandible fragment
European Iron Age Headhunting
7
was found from the ditch adjacent to the entrance, and was
interpreted as part of a trophy head which was displayed
over the gateway, and subsequently fell off (Gardner &
Savory 1964, 221). At Glastonbury Lake village, besides a
general domination of skull deposits in the human bone
assemblage, two cases were recovered of crania and
mandibles together, one of which had damage to the
foramen magnum which is suggestive of being placed on a
spear or stake (Coles & Minnitt 1995, 170-4). A similar
example was found, despite a 2nd century AD context, from
Folly Lane, St. Albans, where the head of an adolescent boy
was deposited in a pot which sustained a mortal injury to
the head and damage to the foram magnum, again
suggesting display on a pole (Aldhouse-Green 2002, 107).
From the oppidum of Buffe Arnaud, Provence, the exclusive burnt cranial remains of two
adult/teenage males were recovered from the destruction layer of the main gate (Duday 1995,
38). These were interpreted by the excavators to be the remains of heads displayed in the
entranceway to the oppidum (Garcia & Bernard 1995, 122).
Less ambiguous evidence comes from excavations of the Iron Age Broch complex at Hillhead,
Caithness, where a calvaria was recovered from the entrance passage floor, suggesting it was
displayed by hanging (Fig. 1; Armit 2012, 4). This method of suspension for display is mirrored at
the Roissy-en-France, Val d’Oise, where cranial fragments with perforations for suspension were
recovered, which also had evidence for occipital hole enlargement and was found in context with
a javelin and shield boss (Fig. 2; Rousseau 2012, 128).
Examples of heads suspended for display by iron nails are more common. From Catalonia
skulls pierced with large nails are known from oppida from Puig Castellar and Puig de Sant
Fig.
1:
Perforated skullcap from Hillhead,
Caithness (after Armit 2012, ill. 1.3).
Fig.
2: Perforated cranial fragments from Roissy-en-France (Val d’Oise) 4th century BC (after Rousseau 2012, fig. 18).
Heron, Headhunting in Iron Age Europe
8
Andreau (Fig. 3; Rovira-Hortalà 1999, 19-25). Similarly a skull
that had been nailed right through was recovered from
Koblenz, Germany (Fig. 4). Many examples come from
Languedoc and Provence in southern France. At the oppidum
of Glanum, Provence, skulls with iron nailing were found in the
enclosing ditches dating to the 2nd century BC (Fig. 5; Arcelin
et al. 1992, 215). From the oppidum of Pech Maho in
Languedoc, fragments of skulls from at least five individuals
have evidence of nailing, from the 3rd century BC (Dedet &
Schwaller 1990, 148). At the oppidum of Entremont,
Provence, at least fifteen skulls with iron nails were found in
the final destruction layer, conditions too fragmentary for sex
and detailed age ascriptions, except all adults or adolescents,
and have been interpreted as being displayed on the
‘hypostyle’ building within the oppidum (Fig. 6; Armit 2012,
191). Perimortem injuries in the form of cuts, presumably
from bladed weapons, suggests a number of these individuals
endured violent deaths, and perhaps their heads were de-
fleshed (Armit 2012, 191). The most exceptional example
comes from the oppidum of La Cloche, occupied between the
3rd and 2nd centuries BC (Chabot 2004, 19). Near the entrance,
among many other cranial fragments, two partial skulls were
recovered, one with suspensory iron hooks, and the other
with a peculiar clasp that grasps the occipital and the orbits
(Fig. 7; Chabot 1982, 50). Another identical clasp without a
skull was also found nearby (Chabot 1982, 50). These
examples from southern France are all from the second and
first centuries BC, but perforated skulls are also known as
early as the sixth and fifth centuries BC at Mont Lassois (Cȏtes
d’Or), Illfurth (Haut-Rhine) and Rosheim (Bas-Rhin) from
Alsace Lorraine, Saint-Etienne-de-Lisse at Niord (Gironde)
(Rousseau 1999, 131). An important point to note about
nailed heads is that it is highly probable this was done while
the head was still fresh, as dry skulls would likely shatter on
impact (Armit 2012, 192). The evidence thusly presented
strongly complies with the claims made by the classical
sources.
Post Conflict Decapitation
Returning to Bredon Hill, Gloucestershire, some 50
individuals (of those identifiable were male) were found
outside the entranceway which appeared to have been left to
decompose (Hencken 1938, 21-2). There was evidence the
bodies had sustained significant mutilation, and very few
skulls were recovered, suggesting to the excavator that the
heads had been taken away as trophies (Hencken 1938, 22).
At Alken Enge, Denmark, in the Illerup River valley, the early
1st century AD post-battle site has produced the heavily
modified human remains of c. 380 individuals (Holst et al.
2018). The large degree of bone modification along with the
significant underrepresentation of crania (and of those that
exist, most are fragmentary), led the interpretation suggesting
deliberate selection of particular body parts (Holst 2018,
Fig.
3:
Nailed skulls from Puig de Sant
Andreu (after Rovira
-
Hortalà 1999, lam.
2).
Fig.
4: Skull peforated
by a nail discovered
near Koblenz, Germany (after Rousseau
2012, fig. 24).
Fig.
5:
Upper skull from Glanum with nail
piercing in the frontal bone (after
Rousseau 2012, fig. 5).
Fig.
6:
Nailed skulls (4, left; 2, right) from
Entremont (after Mahieu 1998, figs. 1, 2).
European Iron Age Headhunting
9
5924) – in other words, along with many other practices
of post-conflict mutilation, it appears heads were
removed. Evidence for decapitations of corpses are also
known from the middle La Tène sites, La Tène itself, and
nearby Cornaux and Port (Jud 2007, 377), as well as
from the late Iron Age site of Basel-Gasfabrik, and the
oppidum at Manching (Jud 2007, 381), however it has
been suggested, particularly with the latter later Iron
Age sites, that these were results of funerary rites than
direct conflict (Jud 2007, 383). The middle La Tène site
at Gournay-sur-Aronde, Oise, appears to have been an
important place of worship, where within a
quadrangular enclosure a series of temple buildings
were set dating to the 3rd and 2nd centuries BC (Fig. 8,
Brunaux 2003a). The main activity at the site were high
amounts of deposition of animal bone and enormous
quantities of weaponry that were deliberately broken
and bent, which was found in the enclosing ditch
(Brunaux 2003a, 59; Rousseau 2012, 118). The degree of
weaponry naturally encouraged the interpretation of a
military context which is also known in the
Mediterranean world by the Greek city states (Armit
2012, 197). This interpretation was reinforced by the
recovery of c. 60 fragments of human vertebrae and long
bones, representing around a dozen individuals
(Brunaux & Malagoli 2003, 26). Many of the vertebrae
had cut marks clearly indicative of decapitation, and
some cranial fragments of the foramen magnum were
enlarged suggesting display on stakes, which were all
found clustered within the entrance terminal (Fig. 9;
Brunaux et al. 1985, 162 et seq.). Later research has
questioned the purely military interpretation however,
where it was discovered three of the twelve individuals
were female, and thus unlikely as warriors, but perhaps
represented victims of sacrifice, however they also note
the possibility of a plurality of observances that does not
preclude the original interpretation (Brunaux &
Malagoli 2003, 26).
At the nearby sanctuary site of Ribemont-sur-Ancre,
Somme, the connection to decapitation and a military
context is inarguable. Dating from the beginning of the
3rd century BC, three enclosures covering c. three
hectares were constructed (Brunaux 2003b, 64-5).
During a series of excavations (Fig. 10; Cadoux 1984;
Fercoq du Leslay 1996) several ‘ossuaries’ were found
within one of the ditched enclosures; square
constructions over a metre in dimensions built
predominantly of human long bones with some horse
bones, on top of a base of human coxa, along with large
amounts of iron weaponry, particularly spearheads and
shield bosses (Fig. 11; Rousseau 2012, 118; Bruneaux
2003b, 65). The first ossuary excavated contained the
remains of at least 300 individuals and a dozen horses,
Fig.
7:
The skulls and fastening devices from the
oppidum of La Cloche, c. 2nd century BC (after
Chabot 1982, fig. 12 & Salviat 1972, fig. 12).
Fig.
8: General plan of Gournay-sur-
Aronde
(after Brunaux et al. 1985, fig. 35).
Fig.
9:
Vertebrae with cut marks indicative of
decapitation from Gournay
-sur-
Aronde, Oise
(after Brunaux & Malagoli 2003, fig. 10).
Heron, Headhunting in Iron Age Europe
10
with many signs of peri-mortem trauma (Cadoux 1984, 74). The
ossuary from the latter excavations similarly produced the bones
of c. 40 individuals along with 200 items of iron weaponry
(Brunaux 2003b, 65-6). Stylistically the weaponry denotes a date
to around the transition of La Tène B-C, c. 260 BC (Brunaux &
Malagoli 2003, 34). Nearby in the outer ditch, a ‘mass grave’ dated
to the beginning of La Tène C, comprising c. 120 individuals, over
300 iron weapons, also contained vertebrae with marks indicative
of decapitation, bones with evidence of combat damage, deposited
in more-or-less anatomical positions (Brunaux 2003b, 66-7).
Interestingly a large gold torque and purse containing Amorican
coinage from the Lisieux-le-Mans region was also deposited here
(Brunaux 2003b, 67). The main point of interest of this site, is that
of all these remains, no traces of crania were recovered from any
context (Dedet 2011, 287). Thus, the evidence here for post-battle
decapitation is entirely convincing, where at least 500 crania in
total must have been removed (Brunaux 2006, 107), but moreover,
the purposeful treatment of the inner ossuaries pertaining to a
form of specific funerary rite, contrasted with the careless
abandon of remains in the ‘mass grave’ along with the Armorican
coin find, led Brunaux (2003b, 67; Brunaux & Malagoli 2003, 34-6)
to suggest these were the remains of a battle between the native
Belgic tribe and Amorican invaders, the latter being treated to a
heroes funerary rite, and the latter lackadaisical negligence. These
latter two sites of Gournay-sur-Aronde and Ribemont-sur-Ancre
also supports the relationship claimed of decapitated heads and
religious/ritual sites. Aside from the few examples at Gournay of
display of heads on stakes, the general absence of skulls constrains
this link to circumstantial.
Lastly, to elaborate on numismatic evidence, two examples of
coinage echo the practice of post-conflict decapitation by the Celts
exist, one silver recovered from the mid-1st century BC in Alesia,
Burgundy (Fig. 12a), reads ‘Dubnocou-Dubnorex,’ and depicts a
man with a sword holding a carnyx boar standard in one hand, and
a severed head in the other (Aldhouse-Green 2002, 97). This refers
to Dumnorix, the Aeduan aristocrat killed by Caesar’s cavalry in 54
BC, who was in opposition to Rome (De Bello Gallico 1.19). This
coin, minted around the same time as Caesar’s campaigns,
importantly bespeaks to the practice of head-taking in central Gaul
during this period, whereas mentioned above, Caesar gives it no acknowledgement. The other
example comes from 1st century AD southeast England attributed to Cunobelinus (Fig. 12b), which
is more tentative, as it is thought to represent the importation of the myth of Perseus holding the
Medusa’s head (Creighton 2000, 129).
Heads and Horses
Representations of severed heads and horses come from a
number of sources throughout the Celtic Iron Age. A few examples
of early Celtic coinage depict mounted horsemen with severed
heads (Fig. 13). The lintel from the oppidum at Nages, Languedoc
(Gard), dated to the 1st century BC stylistically, depicts alternating
horses and human heads (Fig. 14; Arcelin et al. 2003, 204). The
heads, having closed eyes, and where the mouth survives on one is
downturned, are generally accepted as signifiers that they are
Fig. 10: Plan of Ribemont-sur-
Ancre
indi
cating the location of human
remains (after Fercoq du Leslay
2000, fig. 17).
Fig. 11:
(left) Excavation
photograph of one of the ossuaries,
(right) detailed drawing of one
square metre from another showing
iron weaponry (after Cadaux 1984;
Fercoq du Leslay 1996 fig. 6; fig. 7).
Fig. 12:
(A) Coin depicting
Dumnorix, 1st c
entury BC; (B), Coin
of Cunobelinus, 1st century AD.
(after Allen 1980, pl. 22, 317;
Creighton 2000, 129).
Fig. 13:
Celtic coinage depicting the
horse and severed head motif (left)
with rider from Limousine, (right)
from Transdanubia (after Allen
1980, pl. 3, 24; pl. 33, 491).
European Iron Age Headhunting
11
dead, and in this case severed heads (Aldhouse-Green
2010, 132). From Entremont comes further statuary
evidence from precariously reconstructed blocks, once
thought to constitute a three sided pillar (Fig. 15). These
fragments are stylistically dated to the late 2nd or early 1st
century BC (Arcelin & Rapin 2003, 212). Armit (2012,
159), argues this scene represents a narrative of an elite
horseman, whom is repeatedly represented. Whether this
is the case or not, one of the stone blocks, clearly depicts a
rider with a head hanging from a horses neck – in direct
support of the classical sources.
Other sources of evidence that directly conform to this
specific practice come from a 5th century BC chariot grave
at Kärlich, Koblenz, Germany, where a Celtic warrior is
represented on sheet bronze with a typical La Tène sword
with oval chape, and carrying in his left hand, either a
‘pilgrim flask,’ or a severed head (Fig. 16; Megaw & Megaw
1990, 82). Another direct reference comes from Aulnat in
Clermont-Ferrand, where a pottery fragment was
recovered from deposits dating to c. 120-80 BC (Pèrichon
1987, 678) which has graffiti depicting the same scene
(Fig. 17), this time the figure holding a spear ready, clearly
demonstrating in this case that it was a warrior. Finally
two examples of metalwork from Numantia, Iberia,
support the same motif (Fig. 18). The first, a bronze
sceptre terminal, came from a high status tomb depicting
mounted back-to-back figures with human heads under
the horses, which came from a 2nd century BC context
(Lorrio & Ruiz Zapatero 2005, 209). The other is a bronze
fibulae which similarly shows a human head under the
horse’s mouth, and warrior holding a sword (Lorrio & Ruiz
Zapatero 2005, 209).
Clearly then, the claim that the Celts’ practice of
hanging the decapitated heads of their enemies from the
necks of their horses is strongly supported, and that this is
so for such a specific claim suggests this practice was
unlikely to be a mere consequence of happenstance.
Moreover, the examples presented here support the view
that this was the case over vast areas of space and time, at
least spanning from northern Iberia to western Germany,
and from the 5th century to the 1st century BC.
Modification and manipulation of heads
At Býči Skála, Bohemia, a ritual massacre site was
discovered dated to the late Hallstatt period, where the
remains of many dismembered people and horses were
found (Aldhouse-Green 2002, 101-4). Amongst the many
finds, one skull found in a cauldron, was interpreted as
being made into a drinking cup (Fig. 19). The
interpretation has been questioned however, due to
certain fracture edges being new, as has the veracity of the
original publisher and finder of the skull, Heinrich Wankel,
Fig.
14: Lintel from Nages
demonstrating
association of dead heads and horses (after
Arcelin et al. 1992, fig. 22).
Fig.
15:
(left) Stone block fragments of the
suggested three
-
sided pillar from Entremont,
Provence, (right), close
-
up of the fragment
with mounted warrior and severed head (after
Patrice et al. 2003, fig. 114; Arcelin et al. 1992,
fig. 18).
Fig.
16:
Celtic horseman with possible severed
head from Kärlich, Kr. Koblenz, Germany,
3.8cm (after Megaw & Megaw 1990, fig. 94).
Fig.
17: Pottery fragment from Clermont
-
Ferrand, Aulnat, depicting a mounted warrior
with head hanging from his horse’s neck (after
Rousseau 2012, fig. 23).
Heron, Headhunting in Iron Age Europe
12
who was accused of creating other forgeries such as the
two dismembered female hands (Peter-Röcher 1998, 12,
8). Peter-Röcher (1998, 12) does however state the
possibility of another skull that could fit this telos. Yet
more tentatively on my own conjecture, one wonders
that the ‘cap of a human skull up to the brow bone’
(Wavre & Vouga 1909, 230) found at La Tène may have
also served the same purpose, as it conforms very closely
with the detailed description of the preparation of
human skulls into cups by the Scythians via Herodotus
(Histories IV, 63-6).
At a high status Iron Age residential site at
Montmartin (Oise) dating to the 3rd or 2nd century BC, a
highly modified front part of a skull, fashioned into a
mask was recovered from a ‘cult house’ (fig. 1.20;
Aldhouse-Green 2002, 104). The skull appears to have
been subjected to some form of colouring agent, as well
as the inside being polished (Rousseau 2012, 123).
Another example of this practice was found from the
enclosing ditch at the sanctuary at Corent, Veyre-Monton
(Pur-de- Dôme), where the front part of the skull and
part of the mandible of a young adult male dating to the
2nd or 1st centuries BC was recovered (Fig. 20; Rousseau
2012, 123).
While terse, this section provides fairly limited
support for the claim of modifying skulls into drinking
cups, however at least renders it possible. The latter
examples of the modification of skulls into masks may be
more convincing, despite being afforded no mention in
the classical sources. Both were recovered from
ritual/religious contexts, and while perhaps not
providing adequate enough support alone, combined
with the sanctuary sites of Gournay-sur-Aronde and
Ribemont-sur-Ancre, they may go some way to ‘fill in the
gaps’ to a certain extent. This rather diffident connection
becomes all the more resolute when one considers that
the example at Montmartin is not only from the same time period as the sanctuary sites, but is
also roughly 5km from Gournay. Regardless, much more convincing evidence of
manipulation/modification and curation of human heads in religious/ritual contexts exists in
southern France, which will be presented below.
Southern France
The southern regions of France, Languedoc, and particularly Provence, provide some most
conclusive corpus of archaeological material associating severed heads and the Celtic people of
the Iron Age. Quite a few examples have already been presented above with regard to skeletal
remains at a number of these sites, but hitherto all that has been presented is supporting evidence
to the claims made by the classical sources. The archaeology of southern France, while also
supporting the classical literature, allows one to go beyond, where the statuary evidence in
particular and its context therein demonstrate that in this region the severed head was clearly an
important symbol in religious and ritual life. These reasons are why this region was fundamental
to the conception of the Celtic ‘head cult’ during the 20th century (Lambrechts 1954, 20). The
practice of sculpting stone has been argued to have been adopted by the indigenous peoples from
Fig.
18.
(left) The bronze sceptre terminal from a
tomb in Numantia (Soria), 2nd century BC;
(right) Fibula of mounted rider with severed
head from Numantia (Soria) (after Lorrio & Ruiz
Zapatero 2005, figs. 22, 23).
Fig.
19: Heinrich Wankel’s skull cup (after Peter
-
Röcher 1998, abb. 4).
Fig.
20:
Skull ‘masks’ from Montmartin, Oise
(left) and the sanctuary at Corent (right) (after
Aldhouse
-
Green 2002, pl. 15; Rousseau 2012, fig.
12).
European Iron Age Headhunting
13
the Greek’s who founded the colony of Massalia c. 600 BC, but
who retained their Gallic iconography, particularly the motif
of the human head (Aldhouse-Green 2010, 132). Over time the
style and ideological function behind the stone
representations quite evidently changed from the c 6th to the
mid-1st centuries BC, and this section will thus explore these
changes.
The Early Statuary
The 6th and 5th centuries BC was a period where for the first
time at a significant scale, southern French settlements in
Languedoc and Provence started to become substantially
enclosed and internally organised (Garcia 2004, 53). A few
large enclosed settlements such as Saint-Blaise, a few miles
west of Massalia, appears to have been founded at the end of
the 7th century (Chausserie-Laprée 2005, 42). It is also this
period we see the first examples of statuary associated with
human heads. The two earliest examples come from the site
of Entremont, Provence, which lies just north of the modern
city of Aix-en-Provence, and 30km north of Marseille. The
first, tentatively dated to c. 700-450 BC, is named the bloc aux
épis (Fig. 21) after the two ears of wheat on one side of its
decorated faces (Armit 2010, 92). The second is the less-
elegantly named ‘head pillar’ (Fig. 21) which depicts a ‘tree of
skulls’, and thought to have been carved c. 500 BC (Aldhouse-
Green 2010, 132). Both examples display similar
representations; simply drawn blank faces with plainly
incised lines for eyes and noses, lack of mouths, with the
bottom row of heads are inverted, have been interpreted as
representing that they are dead (Benoît 1975, 245) – in other
words, severed heads. It has also been noted that no
semblance of individuality is discernible, but where quantity
seems more important (Armit 2010, 92). Both pieces were
recovered from secondary 2nd century BC contexts whence
the site became an oppidum and capital of the Saluvii – a
powerful confederacy of indigenous tribes (Armit 2012, 89,
92), meaning little else can be ascertained of their
significance.
Two more examples from a sanctuary site Saint-Michel-de-
Valbonne, Hyéres, Provence, also demonstrate similar
depictions of human heads on stone pillars (Fig. 22). The
depictions are less clearly defined than the examples at
Entremont, and badly weathered. What can be identified are
similar blank faces to those of Entremont, arranged in a
vertical fashion, but the larger pillar also depicts a mounted
figure, which perhaps pushes the association of severed heads
and mounted warriors back even further to c. 800-650 BC,
where these stone pillars are tentatively dated to (Armit
2012, 87). While no other trace of human activity existed
there during this period, the site later in the Gallo-Roman
period became a shrine dedicated to Rudianus (Mars)
(Aldhouse-Green 2002, 100). On this fact, it is possible the
iconography of these pillars implies that head-taking,
Fig.
21:
Early Iron Age human head
carvings from Entremont, (left) the ‘head
pillar,’ (right) two angles of the bloc aux
épis (after Arcelin et al. 2003, fig. 111).
Fig.
22:
The two complete pillars from
Saint
-Michel-de-Valbonne, Hyéres,
Var
(after Brun 1999, figs. 481, 483).
Heron, Headhunting in Iron Age Europe
14
violence, and the worship of a war god existed back into the early
Iron Age. A final example was recovered as a chance find from a
deforested plateau above Badasset, c. 20km northwest of Entremont,
which consists of a c. 0.3m2 stone block (Fig. 23; Armit 2012, 95).
Little remains of the sculpture itself except four heads in
diametrically opposed pairs, similar to the bottom two rows of the
Entremont examples.
The linearly arranged heads in these examples have led some
archaeologists to interpret them as representing some form of
genealogy associated with worship of ancestors (Benoît 1975, 248).
Ian Armit (2012, 92) has put forward a link to fertility and
regeneration, based on representation of wheat on the bloc aux épis,
along with the similarity of arrangement to ‘skull trees or racks’
known from many societies in the ethnographic literature of
headhunting societies, whom also have an association between
headhunting and fertility (Jacobs 1990, 120; Hoskins 1996a, 217).
Both interpretations seem reasonable and are not mutually
exclusive, but what is clear at least, particularly in contrast to the
later statuary, representation of severed heads in the early Iron Age
of southern France were not concerned with individuality, of the
victims, or the aggrandisement, or any indication whatsoever of the
individual.
The Later Statuary
Archaeological evidence demonstrates that from the 4th-2nd centuries the settlement record of
southern France indicates increasing centralisation and urbanisation (Garcia 2004, 80). Literary
evidence also confirms the formation of large indigenous groups at this time, the Volcae Arecomici
in Languedoc, and the Saluvii in Provence (Strabo Geographia 4.1.11), whose capitals were at
Nimes and Entremont respectively (Py 1990, 243). Strabo also details the Saluvii where a
confederation of ten groups who were in confederation of a common army composed of infantry
and cavalry (Geographia 4.1.5). Moreover, Strabo also informs us in the same passage that the
Massiliotes established multiple fortresses (Olbia, Antipolis, Tauroeis, and Nicaeae) in response
to Ligurian (Saluviian) aggression, suggesting a significant degree of violence must have been
occurring at this time. Excavations at Olbia, near the modern city of Hyéres, have determined its
earlier construction within the 4th century BC (Ebel 1967,
27) allowing one to place this upturn of violence into
chronological context. Massiliote amphorae are also found
at many of the sites in southern France during this time
(Ebel 1967, 65), which conveys not all interaction was
violent. This context is important for understanding the
significance and change of the statuary and therefore
headhunting, where, along with cranial remains and heads
carved in reliefs eluded to above, two main themes emerge
within ritual/religious areas which undeniably attest to
the a headhunting practice in this region – warrior
statuary and carved (conveniently head-sized) niches.
The site of Roquepertuse, Provence, was
monumentalised in the early 2nd century BC, and at its
centre a two-storey portico structure was constructed
(Fig. 24; Armit 2010, 94). Three carved and decorated
stone pillars and two lintels bear niches that suggest that
they displayed real heads. The pillars were painted with
Fig.
23:
The carved stone block
from Badasset (after Armit 2012,
ill. 4.11).
Fig.
24:
Artist’s rendition of how
the portico structure at
Roquepertuse possibly existed
(after Coignard & Coignard 1991,
fig. 11).
Fig.
25:
Pillar C from Roquepertuse showing the
three head niches alongside decorative
carvings and traces of colour (after Barbet
1991, fig. 17).
European Iron Age Headhunting
15
associated carved depictions of horses and carrion birds, but only
one face bore the head niches (Barbet 1992, 99). Pillar C (Fig. 25)
is the best example providing certain proof that these cavities
displayed real human heads. During excavations one of the niches
still contained human crania, which had apparently been coated
and filled with clay which suggests how they fastened the heads
into their niches (Gérin-Ricard 1927, 23). Moreover, a fragment
of an atlas vertebra was also recovered which indicates that soft
tissue had to have existed and therefore these severed heads were
not fleshless skulls (Armit 2012, 148). It was noticed that there
were significant differences in size and shape of the niches, and
were also incorporated into the design of the pillars, being
outlined by incised lines and black and red colouration (Coignard
& Coignard 1991, 28). This suggests that 1) these niches were
shaped according to the size of particular heads, and 2) these
were unlikely later additions to the pillars, but carved to fit pre-
existing heads when the pillars were created. That at least ten of
these niches were recovered, suggests these must have been
created for the display of a pre-existing set of heads, which along
with the evidence for soft tissue, also indicates that they were
very likely preserved and curated for some time. It also follows
that if they were preserved, then it is plausible the identities of
the heads were also retained, and perhaps important. This raises the question whether they were
venerated ancestors or respected enemies, which must remain open, but these heads were on
display for over a century until the sanctuaries violent destruction c. 190 BC (Barbet 1992, 99).
These carved niches were not limited to Roquepertuse. At site of Villa Roma, Nimes, a similar
lintel fragment with evidence for coloured and incised decoration (again horses), bears two
individually sculpted niches (Fig. 26). The niches on this block demonstrate an alternative method
for the potential display of heads, where the hollows are only c. 5cm deep, but one shows evidence
that it was overlaid with a horizontal bar (Barbet 1992, 96), and if they displayed heads, was likely
only the front part of the skull (Arcelin et al 2003, 206). The display of the anterior part of heads
is also mirrored from a pillar found at the oppidum of Glanon (Fig. 27), where a different method
seems to have been employed. Here, within the niches, embossed ‘knuckles’ were left which seem
to suggest they were designed to receive ‘head-masks.’ One of the niche boss’s also shows
evidence that whatever was hung there, was likely fastened by an iron nail, although most lack
these.
The display of human heads and faces within stone pillars and lintels was a dominant feature
of the southern French sanctuary sites. That these were integrated into elaborate architectural
settings alongside symbolism suggests strongly for a religious purpose from at least the 3rd
century BC, if not earlier. Some sites also have carved simulacra of heads, such as the lintel
fragment from Entremont (Fig. 28). Fig. 29 shows the full extent of these portico sites with carved
niches, real or simulacra.
Another development in statuary in southern France, and
associated with sanctuaries, was the introduction of life-sized (or
near) seated ‘warrior statues’ as seen in the representation from
Roquepertuse (Fig. 24). At Roquepertuse only two of these seated
warriors survive to any sufficient degree, but it was thought from
the many fragments as many as ten could have existed (Arcelin &
Rapin 2003, 185). At Glanum, c. 45km northwest of
Roquepertuse, the remains of at least three statues in similar
seated position and costume exist. In both of these examples, the
dress clearly indicates they are warriors, with armour and
Fig.
26:
Head niches from Roma Villa,
Nimes (after Barbet 1992, fig. 1).
Fig.
27:
Carved head niches with
carved embossed ‘knuckles’ from
Glanon, c. late 2nd/early 1st century
BC (after Arcelin et al. 2003, fig. 116).
Fig.
28:
Lintel fragment believed to
belong to a portico pillar from
Entremont, Provence, with carved
head
-
niche simulacrum (after Arcelin
et al. 2003, fig. 112).
Heron, Headhunting in Iron Age Europe
16
indications on some for sheathed swords (Armit
2012, 134). This cross-legged seating position
was known to be a common behaviour among the
Gauls (Strabo Geographia, 4.4.3). Most of the
remains of these statues across southern France
are incredibly fragmentary, and come from
secondary contexts which make secure dating
impossible (Arcelin & Rapin 2003, 205). The best
surviving examples from Roquepertuse and
Glanum (Fig. 30) interestingly have been dated to
the earliest 5th century BC, based on stylistic
affinities of their clothing (Arcelin & Rapin 2003,
205). This means they predate many of these
sanctuaries by possibly several centuries, and
that they were displayed at their respective
sanctuaries for possibly another one or two
centuries meant they had incredibly long ‘lives’
(Arcelin & Rapin 2003). These examples, known
as the ‘Roquepertuse/Glanon’ type (Fig. 31 for
distribution map), do not however illustrate any
direct reference to headhunting themselves.
A few examples do however, such as the statue
from Bouriège, Aude, (Fig. 32) which was
recovered from an occupation layer dating
between the 5th and 4th centuries BC, clearly
depicting a seated figure with a severed head in
his lap (Arcelin et al. 2003, 212). The best known
examples come from Entremont. Here, at least
nine near-life size statues have been identified in
very fragmentary condition, representing, as with
Roquepertuse and Glanum style figures, seated
warriors whom in this case bear trophy heads in
front of them, and in their hands (Fig. 33). As with
the other warrior statues, they were recovered
from secondary contexts, but based on their
weaponry and attire, their date of creation has
been assigned no later than the first half of the 3rd
century (Armit 2010, 95), making them
significantly later than the other warrior statuary
and marginally postdating the monumentali-
sation of the sanctuary at Roquepertuse by a
generation or so. On the same basis, it also
suggests they were all made around the same time
(Armit 2012, 188). Interestingly it appears each
warrior seems to have been purposefully styled
(Fig. 34), each with individual hairstyles, facial
details, and unique brooches or insignia on their
torsos. This individuality was also afforded to
their severed trophy heads too (Fig. 35), who
adorn different hairstyles or headgear, one
donning a large moustache. These figures are
believed to have originally been displayed within
the ‘hypostyle’ building within the oppidum,
Fig.
29:
Distribution map of Portico sites in southern
France, with niches that could receive real heads (black
dots), or carved simulacra (white dots) (after Arcelin et
al. 2003, fig. 110).
Fig.
30:
The seated warrior statues (left) Roquepertuse;
(right) Glanon (after Megaw & Megaw 1990, fig. 272;
Areclin et al. 1992, fig. 28).
Fig.
31:
Distribution map of ‘Roquepertuse/Glanon’ style
seated warrior statues in southern France (after Armit
2012, ill. 5.10).
European Iron Age Headhunting
17
which was the large communal building. Armit (2012, 188) sees on the
one hand these statues representing elites (either living or remembered),
as there is little to suggest they were a pantheon of gods.
Strabo mentions that the Saluvii were a confederation of ten tribes
(Geographia 4.6.3), and he tentatively suggests perhaps the nine here;
perhaps the last being lost, represent these leaders. On the other hand he
also argues (Armit 2010, 95-6) these statues represent an
ideological/religious shift in the human head from the early iron age
statuary based on collective notions of fertility, to where now they have
come to symbolise the celebration of the individual, where the
individuality of his trophies (and their status) also contribute to his
prestige.
The evidence from southern France (summarised in Tab. 2),
particularly from the 3rd through to the 1st century BC, present a very
powerful case, not only for the existence of headhunting in this region and
period, but that it was also a practice that was profoundly culturally
integrated and expressed in ritual/religious contexts which evolved over
time. Moreover, I contend this chapter has clearly demonstrated the
veracity of the claims made by the classical authors, including the highly
specific claims, not just from southern France, but from Iberia, Britain,
northern France/Belgium, and western Germany.
Repeating the sentiment written at the beginning of the chapter to be
sure one does not commit the transgression many have been criticised for
with regard to the ‘head cult’ (Collis 2003, 216), this is not to suggest a
formalised headhunting practice was homogenous throughout Europe
during the Iron Age. Much of the supporting evidence is selective from
many different places and time periods across Europe, as well as much of
it easily open to alternative interpretations which do not necessitate
headhunting. What the evidence does suggest however is that
headhunting practices across Celtic Iron Age Europe was certainly a
possibility, and in southern France, practically undeniable.
Site Name
Region
Statues with
severed heads
Heads carved or
in relief
Head niches in
pillars and lintels
Skull
fragments
Entremont
P
X
X
X
X
La Courtine à Ollioules
P
X
Bringasses
P
X
Badasset, Vernègues
P
X
Le Pied de L’aygues, Rustrel
P
X
Le Puech, Noves
P
X
Castellet Fontvieille
P
X
Saint-Pierre, Martigues
P
X
X
X
Saint-Michel-de-Valbonne
P
X
X?
Buffe Arnaud
P
X
Constantine
P
X
Saint-Blaise
P
X
Cadenet
P
X
Roquepertuse
P
X
X
Glanon
P
X
X
La Cloche
P
X
Íle de Martigues
P
X
Bouriège
MP
X
X
Rodez
MP
X
Les Castels, Nages
L
X
Nimes
L
X
X
La Ramasse
L
X
X
Pech Maho
L
X
Puech de Lascours
L
X
L’Agréable
L
X
Carsae
L
X
Vié-Cioutat
L
X
Camp Redon
L
X
Ensérune
L
X
La Lagaste
L
X
La Rallongue
L
X
Tonnere I
L
X
Roque-de-Viou
L
X
La Liquière
L
X
Ambrussum
L
X
Lattes
L
X
Cayla de Maillhac
L
X
Fig.
32:
Seated statue from
Bouriège, Aude with sever-
ed head (after Arcelin et al.
2003, fig. 122).
Fig.
33:
A reconstructed
composite of one of the
seated warrior statuary
from Entremont (after
Armit 2010, fig. 9.10).
Fig.
34:
Examples of the
individual character
afforded to the Entremont
warrior statuary (modified
after Arcelin & Rapin 2003,
figs. 33, 34, 35).
Fig.
35:
Close up of the
trophy head statuary from
Entremont clearly illustra-
ting their individual nature
(after Arcelin et al. 2003,
fig. 123).
Tab.
2:
Principle Iron Age sites in southern France yielding evidence for Headhunting
(excluding statuary without explicit severed head iconography) (after Armit 2012, table 6.1).
Special Issue No. 1 1884
Headhunting Across the World
This chapter presents the research on headhunting societies derived from modern
ethnographic literature. Each society has been split into domains (settlement and subsistence,
social structure, reproductive system, conflict systems, and the headhunting practice itself) in
order to identify relationships which may indicate the particular contingent environmental
factors which would allow an evolutionary explanation for headhunting (Workman & Reader
2004, 23). Originally twelve societies had been researched for this study, however due to
limitations only five will be presented here in detail with accompanying notes. Data from the
remaining societies will be included in the analysis.
The Irian Jaya
They Irian Jaya, or The Jaqaj (human being) as they refer to themselves, are a tribe of West
Papua New Guinea which occupy the inland highlands of predominant rainforest but also low
rolling hills (Boelaars 1981, 2). They are surrounded by similar tribes, to the North and East, two
groups of Awju, and to the West the Qondu (Boelaars 1981, 1). In 1956 the estimated population
for the Jaqaj was 8,742 (Boelaars 1981, 3) and their language belongs to a wider regional language
group which includes the Marind on the coast and Boazi tribes, and within the Jaqaj two dialects
exist.
Settlement and subsistence
The Jaqaj subsist on hunting (by the men; pigs, cassowaries, birds, cuscus, rats, and mice) and
foraging/fishing (by the women; Sago pounding being the staple, but supplemented by leaves,
stalks, roots, young bamboo shoots, and nibung palms). They practiced no horticulture except for
relatively large scale tobacco planting (Boelaars 1981, 54). They live in scattered villages
subdivided into groups, and these were further subdivided into areas belonging to different
settlement groups (Boelaars 1981, 12-3). Every settlement consisted of one or more men’s
houses, and nearby one or more women’s houses, which were always in eyeshot of the men’s
house so that men could keep an eye on their wife (wives) (Boelaars 1981, 47). As such the sexes
were always separated, they never stood and talked with each other, in dance and singing, and
even in activities which required cooperation (Boelaars 1981, 49). The daily routine for women
was to gather food (mostly Sago) and the men occasionally went hunting, but often stayed in the
men’s house and smoked pipes and discussed and devised plans for forthcoming headhunts or
retaliation raids (Boelaars 1981, 48).
Social structure
The Jaqaj have somewhat of an atypical social system that roughly follows the tribe/clan, and
also not based strictly on blood relations (although most individuals likely have some degree of
kinship). Due to their communal living nature, no mention is made of a household unit, which is
to be expected, and instead at the bottom of the social system is the qari which is perhaps best
thought of as an extended family, which contains nuclear members and ‘newcomers’ (Boelaars
1981, 15, 18). The next division up is the imu, which was the consequence of several qari based
on kinship or intermarriage or shared common boundaries, to associate into a larger unit
(Boelaars 1981, 15). The function of these units is primarily for defence against external threats,
but also internally for the distribution of communal rights to plots of land (Boelaars 1981, 20-1),
however in practice Boelaars observed (1981, 24), the rule of law tends to function in the manner
described by Thrasymachus – power tends to overrule the notion of law.
Each qari has a designated leader, which was based on strength and/or age, and each imu also
had a leader (the poqoj-radé – ‘big man’), who was one of the heads of the associated qari, the
Headhunting Across the World
19
other leaders constituting an informal governing body (Boelaars 1981, 20). Authority ordained to
these individuals was not hereditary but based on personal prestige, founded on physical
strength, readiness to fight, and amont of heads one has taken (Boelaars 1981, 58, 43). Heads
taken also dictated hierarchy between leaders, such as in the village of Képi, the Kamaqajmu leader
Jabaqaj did not pull the same authority as Jaro or Jaèndé as he had no heads to his name (Boelaars
1981, 59). The Jaqaj also had two other positions of authority (which could also be filled by
women), the akiaq-radé (advisor/councillor) and the joqbera- radé (shaman).1
Reproductive system
The Jaqaj enact what Boelaars’ (1981, 20) calls an ‘optional descent-group system,’ where
matri- , patri- and ‘extra-locality’ was possible and could be changed at will.2 Strict prohibitions
on marriage exist with any relations, however remote (Boelaars 1981, 33), and so exogamy is the
general rule (which was frequently violated by powerful leaders) (Boelaars 1981, 34). Polygamy
occurred, and was the privilege of men who were successful in headhunting – most men had to
make do with only one, while the maximum Boelaars (1981, 41) heard of were 6 wives, and in the
distance past up to 10.3 Marriage was often conducted within a system of sister-exchange (and for
secondary marriages and above, this was a requirement) (Boelaars 1981, 35, 42), and female child
captured on headhunting raids were often kept for this purpose as a future second wife for
themselves or their son (Boelaars 1981, 33). Women (wives) had an economic function in
subsistence, and the more wives the more food could be procured, as well as the less work each
individual wife would have to entail (Boelaars 1981, 89). Women wanted their husbands to be
good head hunters, and would expect her husband to bring home many heads for their children,
as well as for her security in his ability to defend himself and them, and that her prestige is directly
associated with her husband’s (Boelaars 1981, 89).
Conflict and Warfare
Adult life (and loss thereof) was never free of tensions. Women could become the objects of
attack at any time and men might be off taking revenge elsewhere. Within the community quarrels
(mim) due to all kinds of reasons, the most often cited being cases of theft and adultery, could
break out (Boelaars 1981, 125, 57). In turn these could easily turn into homicidal feuds (nu) as
there was a distinct lack of trust between individuals, and any opportunity to renew existing feuds
were enveloped as former defeats or misgivings were seldom forgotten, and ‘revenge was always
foremost in their minds’ which ‘had to be avenged’ (Boelaars 1981, 125, 57, 155). These conflicts
within settlements did not involve headhunting and cannibalism (Boelaars 1981, 125). Intra-
tribal warfare between different settlements (tok) was however, in which they could evolve into
full-scale wars with groups of allies (Boelaars 1981, 125). Revenge is constantly cited as the
motive for continuing hostilities.4 Given the multitude of settlements engaged, shifting alliances,
and that men individually would often take part in fighting of other settlements in order to help
relatives there (Boelaars 1981, 126), this was an unlikely event, and even if possible, did not last
1 The akiaq-radé (advisor/councillor) were experts in landholding rights, marriage relations, lore, the
proselytization of headhunting; and the joqbera- radé (shaman) who among general shaman-type
functions regarding spirits, healing and birth, were important in ceremonies before headhunting raids
to bless warriors and capture souls of future victims (Boelaars 1981, 60).
2 The ‘optional descent-group system,’ described by Boelaars is where a child or wedded couple may choose
which parent they may want to follow and live with, or even possibly a friend or neighbour who are
neither consanguineals nor affines, and may decide to switch at any point to move to another (Boelaars
1981, 17-8) making them patri- and matri-lineal and ‘extra-lineal.’
3 At the village of Kepi in 1953, out of 75 men, 44 had more than 1 wife, and of these 44, 35 had second wives
from neighbouring villages (Boelaars 1981 42).
4 The various imu were locked in feuds for long periods (Boelaars 1981, 128), and indeed peace ceremonies
could not be held unless both sides had seized an equal number of heads (however it seems the giving
of children was also accepted) (Boelaars 1981, 7, 140).
Heron, Headhunting in Iron Age Europe
20
long, and scorekeeping was importantly remembered.5 Open battles also occurred.6 The most
common tactics were ambushes at known fishing grounds and village raids and parties often fled
after suffering one death (Boelaars 1981, passim). The directive of revenge was not on individual
killers but on any member of a settlement that had killed a member of one’s settlement (Boelaars
1981, 147), and as such women and sometimes children were liable for homicide (Boelaars 1981,
78). Qari involved in feuds would immediately suspend hostilities and stand together if invaded
by another imu (Boelaars 1981, 20). Headhunting raids proper existed outside of feuding, and
were as such different. They demanded lengthy preparation, planning, and scouting and were
directed at other neighbouring tribes (Boerlaars 1981, 134).7 All neighbouring tribes of the Jaqaj
have been the objects of headhunting raids (Boelaars 1981, 1-2). The central importance of
warfare in Jaqaj society is exemplified by the training of young children in killing from an early
age, where victims would often be tied up the children made to kill them by the other men
(Boelaars 1981, 145-8).
Headhunting Practice
Headhunting for the Jaqaj was not just a ‘practice’ but was the key cultural nexus which
overarched the rest of Jaqaj social life, which can be demonstrated by the three ritual feasts.8 The
taking of heads also performs an important role in the establishment of permanent social relations
such as the ménaqaé and nakaèri relationships.9 Heads taken would have the fleshy parts were
either eaten or thrown away, the brain was taken out and mixed with sago into cakes, and then
the head was roasted or smoked on a fire (Boelaars 1981, 167). If a pregnant woman was killed,
the foetus was also decapitated and considered a captured head (Boelaars 1981, 168). An earlier
ethnographer (Father Meuwese) maintained, the prepared skulls were hung on a ‘head tree’
which he personally saw (described as a pole with branches from which the heads had been hung)
in the settlement of osso (Boelaars 1981, 169), but when Boelaars visited the skulls were simply
thrown away after the period of mourning (Boelaars 1981, 271). Boelaars writes “the Jaqaj ...
jealously guards his prestige and honour” (1981, 270). This hints to the primary motivation and
individual importance of head taking, demonstrated by over a dozen marks of honour that can be
5 An example of this frequency comes from an informant of Boelaars (1981, 126), who recounts the conflicts
the village Kepi had with other villages over 15 years which included 26 battles, with peace made
between battles of the same group, and, if the numbers are to be believed, resulted in 97 deaths. The
particular attention to ‘scorekeeping’ is of note at least.
6 Casualties of open battles were low due to mêlée fighting with spears and shields where men were free to
flee, the two sides generally knew each-others tricks and ruses, and efforts to fight or kill were
repeatedly thwarted due to individuals often being related on either side (Boelaars 1981, 126).
7 These events are a multi-tribe cooperative event and are planned by the leaders of the respective
settlements, and were highly ritualised with narcotic taking, displays, and shamanistic blessings
(Boelaars 1981, 137, 60, 141). The time they would last is not made clear, but in one example of an attack
on Toba, it included a preliminary ambush two days before the raid, and, that additionally they construct
bivouac camps (Boelaars 1981, 162, 143), it seems plausible these excursions could last up to at least a
week.
8 The honoured dead in Jaqaj society are placed in canoes on trestles, and as long as a body exists there is a
prohibition on coconut trees (Boelaars 1981, 170). For the Ndat-jamé (termination of mourning feast),
a head was required in order that the bodies be cut down and the women were able to remove mourning
garbs, which would often be the cause for women to encourage the men to go on a headhunting raid by
insulting their masculinity (Boelaars 1981, 148). The Taker-jamé feast (the age ceremony) was the
initiation of young boys into future head hunters which required a head (Boelaars 1981, 135). A head as
eluded to above, was also required for a man to be able to marry, from which the marriage feast Batik-
jamé was undertaken (Boelaars 1981, 149).
9 When an enemy is slain by an individual, another of his comrades will behead the enemy, and this
established a powerful bond between the two men called the ménaqaé relationship (Boelaars 1981,
149). Another similar relationship would be formed called the nakaèri relationship, which after a head
was being prepared at basecamp, someone would ask for the jaw, and if accepted they would enter this
relationship (Boelaars 1981, 167).
Headhunting Across the World
21
earned from one head up to fifteen or more.10 Jaqay religion also perpetuates the headhunting
practice.11
Changes
In the beginning of the 20th century, the Netherlands East Indies government started with what
is called the ‘pacification’ of the district. The first tasks of the Dutch government was to end
headhunting raids. This at first increased the headhunting range for the Jaqaj as they could move
into the territory of other tribes (the Marind) due to them being nearer the coast and were
controlled first (Boelaars 1981, 4). Eventually the Dutch government came to control the inland
areas, which in turn made the Jagaj start to instead headhunt their neighbours and former allies
(the Awju) (Boelaars 1981, 136). Trade also developed between the Marind and nearby tribes who
coveted Western iron-ware and often sold their children for it (Boelaars 1981, 4).
The Naga
The Naga live in the mountains at an elevation of 1200-3700 metres, which stretch from the
southeast borders of Tibet through almost due south to the coast which form a dividing wall
between the Assam and Burma (Furness 1902, 445). One of the salient characteristics of the
Naga’s are the multiplicity of languages, so different that one tribe is quite unintelligible to
another, and during the 19th century the Deputy Commissioner at the time estimated there were
no less than 30 languages (Godden 1897, 165-6). As such there does exist reasonable variety
between the different Naga tribes, however they are all head-takers (Godden 1898, 12).
Settlement and Subsistence
The Naga’s live in various scattered villages, some with 300 houses but the most exceptional
examples such as Kohima (an Angami village) contained 865 houses with a population of c. 4000
people (Goddard 1898, 18, 22). Their villages are often situated on the crest of hills to provide a
commanding view in case of attack and often have defensive gates and a stockade (Furness 1902,
446, 451). The central building of the village was the bachelor’s house, or morang.12 The Naga
10 The warrior ornaments of prestige go as follows:
1. A bunch of cassowary feathers hang from the left arm after taking his first head
2. The same but from hung from the right arm for the second head
3. The same but from the back for the third
4 and 5. The same but from the left hip (fourth head), and right hip (fifth head)
6. For 6-10 heads the same bunches of cassowary feathers would then be attached down the arms or
back but on a longer piece of rattan so they resembled tails behind the warrior as he walked (Boelaars
1981, 174).
As well as these general marks, a lath of bamboo adorned with birds feathers and beads (a poqoj-tok)
was allowed to be worn by a man who had taken 2-4 heads, and marked his membership into the circle
of the poqoj-wir leaders (Boelaars 1981, 174). The Ramoq was presented after eight heads, and was a
beaded net wrapped around a small gourd, which had bird of paradise feathers decorations and hung at
the back of a man’s neck from a ribbon (Boelaars 1981, 174-5). The Tok was presented after 12 heads,
which was an upgraded Ramoq with additional wickerwork nooses meant to represent the Kud (head
hunter’s basket). Men who had earned the right to three insignia were also allowed to wear a girdle of
human hair (Boelaars 1981, 175). The Oana was the final mark of prestige, for 15 or more heads, and
was a boar’s tusk necklace worn on the chest, which after this rank the warrior could retire (Boelaars
1981, 175).
11 Idöm, are evil spirits of which the Jaqaj are very afraid, and especially those of dead relatives who are
believed to not leave them alone until they had been revenged in death (Boelaars 1981, 71). Conversely,
souls of headhunting victims (toqojja) were believed to establish a special relationship with the souls of
the headhunter, and could be accumulated, where they would follow behind head-takers and capture
the souls of new victims for him which would increase his ability to obtain yet more heads (Boelaars
1981, 71).
12 ‘Nuclear families’ lived within usual houses, but at a certain age young boys had to move to the morang
to sleep (Furness 1902, 450). The morang doubled up as a bachelor house and a guard/watch house for
Heron, Headhunting in Iron Age Europe
22
subsisted primarily on farming13 and livestock such as goats, buffaloes or metnas are only kept
for meat (Godden 1898, 9). Trade with the ‘more civilised’ plains is a frequent and important part
of Naga economy,14 and each village usually had one blacksmith and manufactured Doas (a short
sword), spears, hatchets, hoes, and 2-3 types of cloth for trade (Godden 1898, 11).
Social structure
Naga society has 3 levels within the tribe. First is the Jhat which are extended family groups
claiming descent from particular ancestors or founders (Furness 1902, 447). One or more Jhats
form to constitute a Khel, or clan (Furness 1902, 446), which in turn constitute the village, which
was functionally the largest unit Naga tribes operated under (Jacobs 1990, 139). Some villages
consist of two Khels but others have four to five or even up to eight, whom work cooperatively
cultivating and building houses (Furness 1902, 447). Although that is often as far as their
amicability went, and different Khels within the same village could be highly antipathetic to one
another.15 At the head of every Khel was a headman who possess little authority aside from
settling the practical affairs of the tribe together with warriors (Furness 1902, 446; Goddard 1897,
168). Village governance was democratic and each independent Khel would demand a voice in any
common action, and the system of blood-feuds and headhunting would check the growth of any
one clan into a position of supreme power (Goddard 1897, 168). Each village in one tribe
commonly had one or two (one subordinate) chief’s which was hereditary, whose authority varied
between tribes.16
Reproductive System
Marriage was strictly exogamous outside the Khel (Furness 1902, 448). Bride price was
required and free will between young couples was exercised (Goddard 1897, 175). Divorce was
common also (Goddard 1897, 177). Intermarriage between feuding Khels existed (Goddard 1898,
13), and was important as one incentive mechanism to achieve truces (Jacobs 1990, 139).
the village when under attack (Godden 1897, 179), but they also functioned as assembly rooms for
councils and dances, and the two decorated front posts of the building (one ‘male’ and one ‘female’) were
of particular ritual significance (Furness 1902, 452).
13 Most Naga tribes practice the jhum system of cultivation whereby land is tilled for 2-3 years and allowed
to lie fallow for 7-10 years (Godden 1898, 7). Some Naga such as the Angami practice permanent
terraced cultivation (Godden 1897, 183). Rice, pea, several varieties of small grain, indian corn, yams,
chillies, ginger, garlic, pumpkins, cotton, sugar cane, and other vegetables are grown (Godden 1898, 10).
14 During the 19th century trade with the plains was undertaken by all observed Naga tribes (Godden 1989,
10). For the eastern Naga’s it was so important that any threat to close markets (normally due to a
murder) was immediately capitulated and surrendered to (Godden 1898, 10). Border Naga’s who traded
regularly were known to ally in order to exclude other Naga tribes of the upper hills (Godden 1898, 10).
They traded Doas (a short sword), spears, cloths etc. commonly for salt, and iron was also highly valued
and as such only used for weapons (Godden 1898, 10-1). It has been recorded in 1854 that the Kachur
imported weapons, handbills and hoes from Manipur, and that the Lugupa Naga had a developed salt
industry (Godden 1898, 11).
15 Furness (1902, 447) gives an account from the Assam census report for 1891 where a Naga village was
attacked where one man, five women and twenty children were slaughtered in one of the Khels, whereby
the adjoining Khel only stood by and watched and did not help.
16 Villages commonly had one chief and a subordinate chief (Goddard 1897, 168-9). Captain Brodie in 1841
recorded one example of the chief of Changneye who commanded some considerable authority where
he was observed as the head by all Naga’s between the Deko and Jeypore, and as such they paid him
tribute (Goddard 1897, 169). Another example from Mr. Davies in 1891 speaks of the Semas Naga where
the privileges of being chief entailed their subjects cultivating their crops free of charge, receiving a free
portion of every animal killed, and having more-than-average number of wives and large families
(Goddard 1897, 170). Normally though hereditary chieftainship resulted in nothing more than a larger-
than-average house, and all decisions still had to be carried out through a council (Goddard 1897, 169-
70).
Headhunting Across the World
23
Polygamy was common but varied.17 Marriage was prohibited to men who had not taken a head
(Goddard 1897, 173).
Conflict and Warfare
The Naga’s have been described “as the abode of fierce and intractable tribes, living in a state
of incessant intertribal warfare … and asserting their presence on our border by savage raids….
Blood feud and relentless raiding were ripe” (Goddard 1897, 161-2). Every tribe and almost every
village was known to be at war with its neighbour (Goddard 1897, 166). Violence mostly occurs
between villages who know each other, are related, and have marriage links.18 The type of conflict
can only be described as feuding,19 as no evidence exists of a tribe united against another tribe,
but only village against village (sometimes involving political alliances of multiple villages), but
the feuding unit was no larger than the village (Jacobs 1990, 139). Revenge and balance was the
primary motivations for the feud, but it was impersonal, and instead directed at the level of the
community. This meant killers or their immediate family were not sought out as objects of
revenge, and instead aimed at their Khel, Morang or village, which in turn meant women and
children were fair game (Jacobs 1990, 142).
Ambush and raids were most common, treacherous attacks on guests was known as well as
random killings of strangers for their head, and pitched battles occurred also, where often the
latter were arranged via an intermediary (Jacobs 1990, 138). The Naga use spears, shields, and a
heavy short sword (Dao) (Goddard 1898, 14). Violence was given a positive cultural validation in
Naga society and fighting was seen as a sport and way of life, especially among young men eager
in competition for heads (Jacobs 1990, 136). Feuds could last very long times with a reasonably
low loss of life, however some events recorded demonstrate large losses.20
Headhunting
The tangible results of feuds were heads as trophies, and the emotion derived from feuding
was primarily channelled into celebrating the return of heads to villages and thus the status of
successful warriors (Jacobs 1990, 142). The main ritual arenas in Naga society were head-taking
17 Among the Sema and Lhota it was common, the latter averaging three wives (Hutton 1921, 28; Mills 1922,
31), whereas the Aos and Angami do not practice it (Goddard 1897, 177).
18 The relationship between villages often involved positive exchange of women, and negative of heads, as
an account by W. G. Archer in 1946 attests with regard to a feud between Konyak villages of Longmein
and Chonui, “[they] began haed-taking 9th March – before that on very friendly terms – intermarriage –
war broke out – casualties: Chonui – 6 killed, 4 wounded. Longmein – 2 killed, 2 wounded, and 1 likely
to die. Longmein could only take away 1 head out of the 6 killed. Chonui took 2 heads from Longmein
later in the day by surprising 2 of the Longmein.” (cited in Jacobs 1990, 138).
19 Feuds often started with material triggers, such as property damage, or infringing on trade routes, but
more often it was symbolic, such as the failure of a ‘formal friend’ to honour his partner with a feast, an
insult, or a failure of the wife-receiving clan to honour bride price obligations (Jacobs 1990, 140). Feuds
were handed down from generation to generation, the original causes frequently being forgotten
(Goddard 1898, 13). Balance was the goal between feuding villages and they were roughly of equal
strength (Jacobs 1990, 140). In 1873, H.H. Godwin-Austen (cited in Jacobs 1990, 140) wrote “a sort of
debit and credit account in lives goes on year after year.” For the Manipuri Angami, so long as the heads
of one village were kept by the opposite party the feud remained active, and the surrender of heads could
suspend hostilities for a while, peace was impossible until each party had suffered equal loss, however,
hostilities tended to continue indefinitely due to the necessity to go on fighting until an even balance was
arrived at (Goddard 1898 , 40).
20 An example of a feud between Tamlu and Namsang lasted for sixteen years, with casualties on both sides
totalling four, and a village which lost one head per year “would consider itself subjected to a series of
stunning blows” (Jacobs 1990, 140). In 1902-5, an average of c. 90 heads per year were taken over a
border area of 30 miles (Jacobs 1990, 142). In 1936 in an the unadministrated area, Ukha, Pangsha, and
Yangkao took together 96 heads from Agching (Jacobs 1990, 142). In 1939 it was reported one village
was completely exterminated with 400 lives lost, and an identical event occurred again in 1948 (Jacobs
1990, 142).
Heron, Headhunting in Iron Age Europe
24
and feasting and a man requires a head to marry, and to be married
enabled one to conduct a feast and sit on councils (Jacobs 1990, 131;
Goddard 1897, 185). It was not the heads themselves which bestowed
prestige to warriors but the markings and ornaments of progressive rank
which heads afforded (Fig. 36-Fig. 38).21 After a successful headhunt, the
respective warriors had to undergo a purification ritual (Jacobs 1990,
120). Captured heads were not just important for individual status of
warriors, but also for the whole village and were necessary for a variety of
rituals culminating in a feast (Goddard 1898, 16), and were publically
displayed in; the morang (figs. Fig. 39 and Fig. 41), to a log-gong (a symbol
of the morang itself), onto a ‘head tree’ (Fig. 40) or paraded around the
village perimeter.22
Fig. 37: Two young Konyak men with fresh tattoo marks on their cheeks indicating their achievement of successful
headhunters. 1937. (after Jacobs 1990, 115).
21 The different trophies among Naga tribes varied. For the eastern Naga, on acquiring a first head they were
permitted to wear a special collar called the Kapentali, a collar of boars tusks was worn after the second,
and after a third a man could wear an apron covered in cowrie shells where a tally of successive heads
could be recorded (Goddard 1898, 10). Successful headhunters of the Ao tribe wore a boar’s tusk
necklace where a new tusk was added for every head taken (Furness 1902, 454). Tattoos for becoming
a successful headhunter were commonplace as well, called the Ak (Goddard 1897, 185).
22 The heads are often mangled and insulted by villagers (usually the women (Jacobs 1990, 131)) who throw
rice and liquor on them (Goddard 1898, 16). Women of the Chang with blood relations on the male side
who took a head were also privy to receiving a particular head hunters tattoo if they cooked the head
with chillies (Jacobs 1990, 121). Another example is when a village made a new log-gong, it required a
head to ‘christen’ it (Jacobs 1990, 121). Warfare and head taking were important also for general beliefs
regarding fertility and good harvest (Jacobs 1990, 119-21).
Fig.
36:
A man wearing
head hunter’s boar tusk
collar (after Furness 1902,
PL. XLIV, 1)
.
Headhunting Across the World
25
Fig. 38: Hairpin used as a tally of number of heads taken. Konyak 1937 (after Jacobs 1990, 120).
Fig. 39: A rack of 151 skulls from a Konyak village. 1912-23. (after Jacobs 1990, 116).
Heron, Headhunting in Iron Age Europe
26
Fig. 40: A ‘head tree’ from a Chang village. 1947 (after Jacobs 1990, 123).
Fig. 41: Skull trophies displayed on wooden shelves in the morang of a Konyak village. 1918-45 (after Jacobs 1990, 123).
Headhunting Across the World
27
Changes
After 1900 it became evident that villages in the unadministrated area were making attacks
(often into the administrated area) resulting in very high losses in life (Jacobs 1990, 142). This
was due to British interference on several levels, not all intentional. Firstly the introduction of
guns made large scale killing very easy, where warfare with spears and daos would not (Jacobs
1990, 142-4). Secondly the arbitrary division of land into ‘administrated’ and ‘unadministrated’
cut off some villages from their allies who protected them, allowing them to be decimated (Jacobs
1990, 146), as well as growing antagonisms by annexed villages who could no longer headhunt,
while their neighbours could (often at them) (Jacobs 1990, 144). W.G. Archer (cited in Jacobs
1990, 148) also notes positive outcomes however, where the Sangtam villages in 1927 were very
happy with British administration suppressing violence as now the men could work instead of
wasting half of their time on sentry duty, and as such developed a rich economy and trade with
the plains.
The Montenegrins
The tribes of Montenegro have lived for over a millennia within contact with centralised state
systems (Boehm 1984, 222), and are the last example in Europe of a wholly tribal nation that
developed into a state (Durham 1928, 34). As such, most of the evidence comes from the last half
of the 19th century as after King Nikola’s constitution in 1905 and the subsequent formation of the
former Yugoslavia, tribal conflict was largely suppressed (Durham 1928, 52), however in 1966
tribal identities were still held with primary importance (Boehm 1984, 46). In 1900 there were
thirty tribes (Durham 1928, 34).
Settlement and subsistence
The Montenegrin tribes were primarily pastoralists, but also practiced horticulture for the
grains and vegetables which they were dependent on (Boehm 1984, 45). Traditionally they lived
in stone 1-2 roomed houses (Katuns) with roofs either thatched or covered with slabs of stone,
built on mountain slopes, often on slabs of native rock (Durham 1928, 37). During the summer
they moved to communally (tribally) owned lowland pasture, which as a result meant they were
quite mobile (Boehm 1984, 46). They produced their own woollen clothes and bedding, and sold
their surplus and products to each other, but also, along with plunder gained in raids, used Turkish
or Venetian markets to acquire firearms, swords, munitions, sugar and coffee (Boehm 1984, 8,
46).
Social structure
The Montenegrins live in a three-tiered social system. The largest unit was the plemé (tribe)
which were believed to be descended from a common male ancestor (Durham 1928, 34). It
functioned as the basic territorial unit which collectively protected the communally held pasture
from other tribes, mediated peace within, and made decisions on war, peace, or other
accommodations with other tribes (Boehm 1984, 48). Each Plemé was divided into several clans
(Bratstvos) based on extended kinship, not farther than second cousins around 1900 (Durham
1928, 34). The Bratstvos was a socially entangled, sharing the same moral reputation and military
prestige, it functioned as a semi-localised group of political and social solidarity, and was the basic
military and feuding unit (Boehm 1984, 48-9). Bratstvos’ could number from a few houses to
several hundred (Boehm 1984, 49). Clans were subdivided into households (kuća) whom owned
their own winter houses and pastures, fields, and forests around them, and constituted the most
basic cooperative economic unit, involving as many as 30-40 people (Boehm 1984, 48). At all
levels there was a headman which was generally hereditary, including the Plemé which had their
own chief, and traditionally there were other positions of merit.23 The leaders (glavari) conduct a
23 The head of house was a domachin or stareshin; the head of each clan was the knezh, and the general term
for leading men was glavari (heads) who were nominally elected posts (chosen by other glavari) which
Heron, Headhunting in Iron Age Europe
28
tribal sud dobrih ljude (court of good men) whenever it is required to prevent feuds or create
truces (Boehm 1984, 48), but other than that they had little authority.24
Reproductive system
Montenegrin tribal society was traditionally strictly exogamous between tribes, except when
the tribe descended from two or more separate stocks, which could then intermarry (Durham
1928, 34). The general belief was no blood was inherited from women, only through the male line
(Durham 1928, 34, 164), and given the fact the cultural ideal was that kuća’s should never divide
due to greater cooperative strength, they were patrilineal (Boehm 1984, 48). Marriages were
arranged by heads of houses, and the young rarely were consulted or had any choice, often
betrothals occurring at infancy (Durham 1928, 205). Until the 19th century it was common for
boys to be married at sixteen and girls as young as thirteen (Durham 1928, 205). Child mortality
was exceptionally high, particularly for females (infanticide highly likely), as was childbirth and
post-natal mortality.25 While marriages functioning as mediatory and mitigative social ties
between groups/tribes, and strengthen one’s own house, is common in tribal societies, this
function seems to have been particularly important for the Montenegrins, and contributed to an
already short supply of women.26 Polygamy was attested to in historical times but two wives
appears to have been the maximum.27
Conflict and Warfare
The Montenegrin tribesmen spent much of their energy externally in warfare, headhunting,
and raiding, while concomitantly continually engaged in vicious blood feuds amongst themselves
(Boehm 1984, 3). Bravery (junaštvo) and honour (čast) were the most salient values for
usually went down hereditary lines to the eldest son (unless in cases of cowardice or incompetence)
(Durham 1928, 36). Most Pleména had an esteemed ‘noble’ clan from which the chief of the Plemé would
come (Durham 1928, 36). Other traditional positions of merit were ‘serdar’ (noble rank), ‘Voyvoda’
(military commander), and ‘Bariaktar’ (Standard bearer and word of Turkish origin) (Durham 1928,
36).
24 Several accounts in Mary Durham’s ethnography illustrate the general lack of authority within
Montenegrin society despite having a clear hierarchical structure. An example by an Austrian envoy,
Colonel Pautlitch, writes in 1782 to Vienna, “the nation has no police, no laws. A kind of equality rules
among them. None in his heart thinks himself less than his chief or headman. These have only a certain
authority for the common business and for settling blood-feuds.” and he continues showing how consent
is required from the population for every function (1928, 77). Another account in 1841 by a certain
Mickiewicz wrote “Montenegro has no species of government. It consists of 24 tribes, each with a chief
who has no governmental power. The Vladika [(bishop), who was Vladika Peta I at the time] is regarded
as the head of the country, but his political authority is nil. He calls the nation to war if attacked by the
Turks and sometimes presides at councils, but as no authority except over the clergy.” (1928, 81).
25 For the first two years of life, the ratio of males:females at one point was 1.6:1 (Boehm 1984, 177). Many
women that did not die during childbirth, often died from haemorrhages or displaced wombs as it was
cultural practice to submit them to heavy labour immediately after birth for forty days due to being
considered ‘unclean’ or impure (Durham 1928, 188).
26 Old Serbian ballads reveal, as well that exogamy was the rule, that marriages were designed to strengthen
the house, as the following example where a youth laments to his father that he has not received a wife,
where his father replies, “I have long sought for one, but where I find a maiden for thee, I do not find a
friend [ally] for myself’ and where I find a friend, I find no maiden” (Durham 1928, 205-6).
27 Mary Durham frequently heard from individuals the practice of having two wives in ‘the good old times’
(1928, 211), and also relays two accounts, one from Medakovitch in the 1840’s where he states
Montenegrins often used to have two wives, “but do not now”, and another by Joseph Kermpotitch in
1788 who made a similar statement (1928, 211). While Durham could get no definitive information, but
she hypothesises that this was likely the result of levirate, as marriage customs permit the bridegroom’s
next brother and uncle as djeveri to the bride (1928, 212), and in historical times may have even gone so
far as wife sharing between brothers of the bratstvos (1928, 208). Additionally the practice of levirate
was common all through the neighbouring Albanian tribes and many of the Montenegrin tribes are of
Albanian descent (1928, 212).
Headhunting Across the World
29
Montenegrins (Boehm 1984, 46), which were key virtues historically for maintaining their
independence from the nation states and empires. Blood feuds existed within the tribe, and
generally started between households and could escalate to the clan but no farther (Boehm 1984,
172), they followed strict public rules, women were completely exempt from harm and blood-feud
obligations (despite being cited most often as the cause in several respects), but children over the
age of roughly two years were liable (Durham 1928, 165), as with any male kinsmen/clan
members of the respective killers, thus indicating communal morality/responsibility.28 Attacks
were generally ambush by firearms (Durham 1928. 165). Men who fail to take blood are
dishonoured and ostracised (Durham 1928, 170).
Inter- or extra-tribal conflict was where clans and individuals could compete for heroic status
(Boehm 1984, 167). The Montenegrin tribal territories were historically densely packed with
chronic problems of overpopulation (Boehm 1984, 222), and inter-tribal conflicts over pastures
commonly resulted in feuds and warfare, whence allies could be forged and brought in (Boehm
1984, 166). They were also vociferous raiders, of cattle, heads, and women (Boehm 1984, 43),
whether against other tribes, their Albanian neighbours, or invaders such as the Ottoman’s
(Boehm 1984, 43-4). Aside from the material benefits, raiding was considered a kind or ‘sport’
(Boehm 1984, 46), and even known to be led by their priests (Durham 1928, 77). Different
methods of warfare were recognised by the Montenegrins.29 At any time due to external threat,
such as resistance against the Ottomans, tribes could quickly suspend transgressions and
confederate (Boehm 1984, 48).
28 The Montenegrins literally describe themselves being in feud as being ‘in blood’ (Durham 1928, 77), and
as descendence only passes through the male line, this explains why women are exempt, and to kill one
was a great source of shame for the entire clan (Boehm 1984, 112). As leaders of clans and tribes were
generally all bound by pobratimstvo (blood brothers) feuding clans within the tribe were usually left to
the respective clans themselves (Boehm 1984, 172). Abductions and seduction of women, as well as
adultery, runaway wives, or refusal of girls to uphold betrothal arrangements were common sources of
blood-feuds (Durham 1928, 171), as well as quarrels between women of different clans (Boehm 1984,
48). Additionally, women themselves were fierce upholders of the blood feud, and a widow left with a
young son was often brought up by the mother to avenge his father when he comes of age, often keeping
a blood-soaked piece of cloth as a reminder (Durham 1928, 164). This was often done despite the fact
‘blood-gelt’ had been paid and the feud ‘settled’, and risking the life of her son, thus passing down feuds
inter-generationally, leading to feuds that were once settled commonly being reopened (Durham 1928,
164).
The general way a feud was conducted was after two clans became engaged ‘in blood’, usually after a
murder, the affected clan would siege the house of the perpetrator’s and if strong enough, the
perpetrator, or his whole clan if the imbalance was significant enough, could be forced to flee, and houses
and property were sometimes destroyed (Boehm 1984, 168, 166). If they were too evenly matched, then
they would turn to the traditional rules of feuding, which involved turn based killing where scorekeeping
exact and public, and only the clansmen of the lower score went hunting (Boehm 1984, 165).
Interestingly the exemption of women from harm was so highly regarded that during a siege, the women
of the household would often go out at night with a torch to find the attackers’ locations and relay this
information to her kin so they could be shot (Boehm 1984, 112). The rule of hospitality from an account
by Mary Durham between her guide and a house she visited, demonstrates that individuals who were
‘in blood’ could be temporarily suspended (and even joked about) if a house head received an enemy
member as a guest, from which he would be safe until the sunset of the day he leaves, and this custom
was even extended to her, where if anything were to befall her, they would be obliged to ‘take blood’ for
her (Durham 1928, 169). On her travels Durham heard of a story of a household head killing his own
brother for violating this rule, and was judged to have acted righteously (Durham 1928, 169).
29 The rat was a full tribal army that was mobilised to fight another army, where heads were taken as
trophies and the objective was territorial acquisition or the political domination (Boehm 1984, 193).
The četovanje was a smaller raiding party of 15-30 men who stealthily entered enemy territory in order
to size moveable property, including women, and quickly escape without sustaining losses (Boehm
1984, 193).
Heron, Headhunting in Iron Age Europe
30
Headhunting practice
Headhunting only occurred in inter- or extra-tribal warfare (Boehm 1984, 166), and was most
often practiced during the 19th century against the Turks and during the middle-ages.30 Taking an
enemies’ head was the most prized trophy of manly accomplishment as a warrior, where they
would be taken with their traditional sword, the handzhar (Durham 1928, 173), Historically,
heads were sought after by men as young as 16 who attended raids (Boehm 1984, 46). Heads are
attested to have been used as tokens in the sharing of loot (Durham 1928, 173). Heads were also
displayed on stakes in victory celebrations at home base, with important heads being taken for
display and preservation in monasteries (Durham 1928, 175-6). Heads were known to be
preserved by salting (Durham 1928, 179).
Changes
Due to the extensive and continuous nature of contact and influence with empires and states,
despite the Montenegrin’s extraordinary resilience, there are far too many changes to commit to
space here, and due to the continuous nature of contact provide little diagnostic benefit. All that
will be said is, that as one might expect, headhunting substantially increased during the Ottoman
threat (Durham 1928, 52), and the movement to statehood suppressed headhunting and internal
violence in general. As headhunting was gradually suppressed, there was a switch to nose-taking
instead for a time before and during World War I (Durham1928, 177), but after the Great War it
had all but died out, save for a few isolated cases (Boehm 1984, 6).
The Jivaro
The Jivaro live in the wet, mountainous, heavy rainforest of the Amazon of eastern Ecuador
(Harner 1972, 1, 12). In 1956-7, five tribes existed, the Achuara, Huambisa, Arguaruna, Mayna and
the Shuar, and together their estimated population at this time was 7,830 (Harner 1972, 14). The
following data of this tribe concerns primarily the Shuar, unless otherwise specified.
Settlement and subsistence
Before the 19th century it appears the Jivaro lived in large communal houses, which typically
contained several families (Steel 1999, 752), but from the 20th century onwards, the Jivaro lived
in dispersed loosely grouped nuclear households, which has been hypothesised due to exposure
of infectious disease brought about by European contacts from the mid-19th century (Bennett Ross
1981, 56). In 1956-7 the interior Jivaro were scattered in c. 245 houses over an area of 4775km2,
giving approximately 0.46 people per km2 (Harner 1972, 77). The Jivaro subsist mostly on slash
and burn agriculture, with each house having several gardens which provide 65% of their diet
(Harner 1972, 47). Hunting is a chief source of protein (about 20%), but some fish, and virtually
all households raise chickens, some ducks, and very few pigs (Harner 1972, 55, 60, 68). Due to
slash and burn technique, houses are usually occupied for 5-9 years after which they family
relocates to a new area (Harner 1972, 44). Division of labour is strictly defined.31
Social structure
The nuclear family within the individual house is the centre of Jivaro life, and lack any villages
or formal corporate kin groups (Harner 1972, 41, 170). The family father is the absolute ruler of
his house (Karsten 1923, 7). A Jivaro man could aspire to achieve three positions of prominence;
30 See Mary Durhams (1928) section IV, 6, for an extensive account of accounts and ballads of head taking,
mostly against the Turks, where often thousands of heads were taken after successful hostilities.
31 While men are responsible for some garden activities, such as harvesting maize, plantain, and
‘gold’bananas, and peanuts which both sexes plant and harvest together, women do all the laborious
garden work including planting, harvesting, and weeding (Harner 1972, 52). Men go hunting, fish, and
prepare skins from wild boar, jaguar, ocelot, and other felines for trade, and both sexes manufacture
braided rope, thread, and string (Harner 1972, 68). Men are also responsible for protecting their
families, clearing forest for gardens, building houses and collecting firewood (Harner 1972, 79).
Headhunting Across the World
31
amigri (a paired social bond based on trade relations); uwišin (Shamans), and kakaram (‘powerful
one’ an accomplished warrior (Harner 1972, 125, 116, 112).32 The Jivaro only elect temporary
chiefs in times of war, which can only be ordained to those who have killed many enemies and
captured many heads (Karsten 1923, 8). They possess great authority and they alone arrange and
decide everything for the planned expedition, the time, the mode of attack, and have absolute
obedience from the younger warriors (Karsten 1923, 8). The most honoured men of Jivaro society
are untä (big/old men) who are a combination of being renowned killers in feuds and war, and/or
are shamans (most old men and/or kakaram are initiated in some degree with shamanistic arts
(Karsten 1923, 9)), old enough to have grandchildren, are magnanimous and honest with friends,
and are successful hosts for feasts (Harner 1972, 110).
Reproductive system
The Jivaro are polygamous and the typical household composition is one man, two wives, and
seven children (Harner 1972, 79). Levirate is practiced (Karsten 1923, 12). The Jivaro have the
atypical combination of endogamy and temporary matrilocal residence (Harner 1972, 95-6).33
This is temporary in that it was customary that a husband would move in with his father-and-law
and wife and help him in his general toils until after their first child (or second or third in earlier
times), whence the couple would establish their own household (Harner 1972, 199). The number
of wives a household had was of utmost economic importance for the procurement and
maintenance of social prestige.34 If a man has taken a head but has no wife for the tsantsa feast, it
is considered obligatory he take a wife first (Karsten 1923, 38)
32 With common conflict and violence, and lacking formal corporate kin groups for security, amigri
relationships are a formalised relationship of an egalitarian nature formed through relationships of
trade and/or exchange (Harner 1972, 125). Amigri pairs are the strongest social bond between men, and
partners become mutually obliged to one another over that of closest kin relations. As such it is not
unusual for brothers and fathers/sons to become amigri (Harner 1972, 125). Shamans and Kakaram
have a hierarchical nature and constitute what little leadership exist in Jivaro society.
Shamans fulfil general healing and bewitching services, and as sorcery is highly feared in Jivaro society,
they often wield considerable power in the neighbourhood (Harner 1972, 116). Shamans are paid for
their services the most highly valued goods available (such as a muzzle loading shotgun for a single
treatment, often with an additional dog or machete for good measure thrown in), also receiving gifts of
ornaments, clothing, and chickens for only casual visits (Harner 1972, 117-8). At the same time, due to
fear of being bewitched, Shamans are allowed to circumvent reciprocal relations, and marriage
obligations, which all together results in them being the wealthiest in material terms in Jivaro society,
frequently utilising their wealth for social capital and influence in acquiring specific services, and often
boasted of their wealth indicating they were fully conscious of this (Harner 1972, 117-8). Shamans
constituted about one quarter of all Jivaro males (Harner 1972, 122).
Prestige is earned by being an outstanding killer, in turn helping the people of his locality in eliminating
their enemies (Harner 1972, 112). When a man has killed several people, and has possessed a significant
quantity of arutam soul power, he is regarded as a kakaram (Harner 1972, 112). The majority of young
men interviewed by Harner expressed strongly their desire to kill to avenge deaths and acquire arutam
so they may be become respected and feared kakaram (1972, 112). Particularly exceptional kakaram’s
may have so much prestige they may even be sought for help by his worst enemies in times of dyer need
(Harner 1972, 113-4.
33 Under further investigation the strange combination of endogamy and matrilocality follows an adaptive
logic. Due to their highly dispersed single household settlement pattern, with rarely another house
within a kilometre of each other (Harner 1972, 78), along with the lack of formal corporate kin groups
of any sort, one can expect endogamy to be the only functional reproductive strategy. As such cross-
cousin marriages are the preferred course (Harner 1972, 95). This however, creates conflicts within
households, as brothers naturally compete for the same females. Matrilocal residence splits up brothers
and therefore is a good component strategy for mitigating internal conflict.
34 The Jivaro place a high value on drinking beer (and Jivaro men alone drink an average of 11-15 litres per
day (Harner 1972, 52)) and food. One’s status is greatly affected by one’s ability to generously supply
guests with such goods during visits or hosting feasts, and “no man can expect to have many friends
unless he is a good host” (Harner 1972, 81). Therefore, given the women are responsible for the
Heron, Headhunting in Iron Age Europe
32
Conflict and warfare
Traditionally Jivaro violence was conducted with a spear (nanki) and shield (tandara) (Karsten
1923, 6), but by the 20th century every adult male possesses at least one firearm, which dominate
combat (Harner 1972, 67-8). Feuding between households is rife in intra-tribal Jivaro society, but
follows a set of rules which are observed in near unanimity, and any infraction is always followed
by retaliatory violence (Harner 1972, 170). Feuds commonly start over discretions over women
(Karsten 1923, 10), but most commonly due to ‘witchcraft’ as the Jivaro ascribe all non-epidemic
illnesses and deaths, and non-violent deaths to witchcraft and poisoning (Harner 1972, 170),
which also tends to provide an interminable pretext for further feuds (Karsten 1923, 13). The
Jivaro have a communally orientated morality (Karsten 1923, 11-2), where any relation of a slayer
is liable for revenge (Karsten 1923, 13), including women and children (Harner 1972, 172).
Feuding rules strictly dictate a ‘one life for one life’ system, and if a household does kill more than
one, they receive severe indignation from the whole tribe (Karsten 1923, 13). Feuds are often
settled after both sides have exchanged one alike, termed tumáshi ajérkama (blood guilt atoned)
(Karsten 1923, 13), but can also be settled with payment or if their untä is killed (Harner 1972,
182). Attacks consist of nearest male relations, and sometimes ambushes are set up secretly
(Karsten 1923, 17), but more often the feud is publically proclaimed, where three to five years can
sometimes pass before an assassination attempt occurs (Harner 1972, 181). After a successful
revenge killing, slayers must observe a purification ritual.35
Jivaros of different tribes (and other groups like Canelos Indians) exist in almost perpetual
enmity, and wars (meséta) are of complete extermination (Karsten 1923, 14). Warfare is often
conducted with several tribal alliances on either side, and preparation for attacks takes a