See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/318754333

The dogs of war: A Bronze Age initiation ritual

in the Russian steppes

Article in Journal of Anthropological Archaeology · December 2017

DOI: 10.1016/j.jaa.2017.07.004

CITATION READS

1 261

2 authors, including:

David Anthony
Hartwick College
34 PUBLICATIONS 1,583 CITATIONS

SEE PROFILE

All content following this page was uploaded by David Anthony on 02 January 2018.

The user has requested enhancement of the downloaded file.

 Journal of Anthropological Archaeology 48 (2017) 134–148

Contents lists available at ScienceDirect

Journal of Anthropological Archaeology

journal homepage: www.elsevier.com/locate/jaa

The dogs of war: A Bronze Age initiation ritual in the Russian steppes MARK
⁎
David W. Anthony , Dorcas R. Brown
Hartwick College, 1 Hartwick Drive, Oneonta, NY 13820, USA

A R T I C L E I N F O A B S T R A C T

Keywords: At the Srubnaya-culture settlement of Krasnosamarskoe in the Russian steppes, dated 1900–1700 BCE, a ritual
Canid sacrifice occurred in which the participants consumed sacrificed dogs, primarily, and a few wolves, violating normal food
Comparative mythology practices found at other sites, during the winter. At least 64 winter-killed canids, 19% MNI/37% NISP, were
Rituals roasted, fileted, and apparently were eaten. More than 99% were dogs. Their heads were chopped into small
Zooarchaeology
standardized segments with practiced blows of an axe on multiple occasions throughout the occupation. Two
Bronze Age
Eurasian steppe
adult men and two adult women from the nearby cemetery, possibly two generations of resident ritual spe-
cialists, showed unusual skeletal pathologies and post-mortem treatments. The repeated violation of the canid-
eating taboo, unique to this site, combined with the metaphor of human transformation into male canids,
suggests that the participants entered a liminal state typical of a rite of passage. Parallels from comparative Indo-
European (IE) mythology provide the indigenous narrative that gave meaning to this ritual: we argue that it was
an initiation into the widely attested IE institution of the youthful male war-band, symbolized by transformation
into a dog or wolf.

1. Introduction: Youthful war-bands in mythology and
archaeology
This essay describes an archaeological assemblage containing at
least 64 sacrificed dogs and wolves that, we argue, is best explained as
the remains of a Bronze Age coming-of-age ceremony in which boys
were transformed into warriors by symbolically becoming dogs and
wolves through the consumption of their flesh. The site where the canid
sacrifices occurred is 3.5 km north of Krasnosamarskoe (Kras-no-sa-
MAR-sko-yeh), a village in the northern Russian steppes 40 km south-
east of the city of Samara (Fig. 1). The Krasnosamarskoe settlement was
occupied between 1700 and 1900 BCE by people associated with the
Late Bronze Age (LBA) Srubnaya (or Timber-Grave) culture, and con-
sisted of two or at most three structures occupying an area of about
40 × 60 m located on the first terrace of the Samara River. Most of the
evidence that we discuss is archaeological, but in order to move from
bone counts and features to an interpretation of the canid sacrifices, we
also use evidence from comparative mythology, which is presented
first. The meanings of ancient rituals were multiple and depended on
agent and context, but all group rituals were supported by public nar-
ratives that encoded shared beliefs about the meaning of the ritual acts
(Harrison, 1991: 7–15; Puhvel, 1987: 15). Archaeologists are pessi-
mistic about the possibility of identifying narratives and beliefs from
material evidence alone, and therefore often turn to examinations of
performance, social adaptation, or other aspects of ritual (Rowan, 2012;
Whitehouse et al., 2014; Hull, 2014). Comparative mythology can be a
particularly helpful tool in this regard because the comparison of si-
milar myths can reveal narrative themes whose antiquity is suggested
by their geographic and chronological distribution, as well as by phi-
lological analyses of shared ritual phrases (Lankford, 2007, 2011;
Watkins, 1995). If we can recover ancient narrative themes we might be
able to identify at least some of the beliefs that gave meaning to ancient
rituals.
A report containing the primary data presented here was published
previously (Anthony et al., 2016), but that volume focused on under-
standing how the Srubnaya pastoral economy functioned across the
landscape of the lower Samara River valley in the LBA, examining the
economic roles played by agriculture and mobility in an ecotone setting
of steppe-and-river-valley resources. These economic questions were
the principal focus of the Samara Valley Project from 1995 to 2001
(Anthony, 2016). The canid sacrifices found at Krasnosamarskoe were
described in the final report in three archaeozoological chapters pre-
pared by two archaeozoological research teams with slightly different
approaches, Russian (Kosintsev, 2016), and American (Russell et al.,
2016); and a third team examined the seasons of death based on in-
cremental banding in animal teeth (Pike-Tay and Anthony, 2016).
These chapters described all of the fauna, not just the canids. Their
stratigraphic context was described in a fourth chapter on settlement
stratigraphy and features that also covered all of the features, not just
the aspects connected with the sacrificial deposits. A section on

⁎
Corresponding author.
E-mail address: anthonyd@hartwick.edu (D.W. Anthony).

http://dx.doi.org/10.1016/j.jaa.2017.07.004
Received 23 October 2016; Received in revised form 13 July 2017
0278-4165/ © 2017 Published by Elsevier Inc.
D.W. Anthony, D.R. Brown
Fig. 1. Krasnosamarskoe site shown within the Srubnaya-Andronovo horizon of LBA steppe cultures, ca. 1900–1200 BCE.
comparative mythology was added to Pike-Tay’s chapter on seasonality
with her cooperation, but the data to which it referred was scattered
through other chapters. This is the first essay that extracts the canid
sacrifices at Krasnosamarskoe and reviews them comprehensively.
Although the canid sacrifice assemblage at Krasnosamarskoe ap-
pears to be unique, the association between wolves, dogs, and youthful
war-bands is widespread and well-known in comparative Indo-
European (IE) mythology (see West, 2007 for an overview). Youthful
male war-bands, often symbolized as dogs or wolves, are thought to
underlie the IE groups designated as the luperci or suodales in Latin, the
kouros or ephebes in Greek, fian in Celtic, männerbünde or jung-
mannschaft in Germanic, and vrātyas or Maruts in Indic. References to
youthful war-bands in these mythic traditions shared a cluster of traits
recognized by Falk (1986) and McCone (1987), re-examined by
Kershaw (2000) and Das and Meiser (2002), and summarized by West
(2007: 448–451) and Mallory (2007: 93-94).
IE youthful war-bands were composed of adolescent (post-pub-
escent, pre-adult) males, sons of aristocratic or elite families in Vedic,
Celtic, and Latin sources. They were sent away to live in the ‘wild’
outside their own society for a number of years, where they were
specifically ‘landless’, lacking socially accepted rights to the land upon
which they lived, and without possessions other than weapons. Under
these conditions they could steal, raid, and take sexual license with
women of other groups without the ordinary legal penalties that would
have resulted from these behaviors at home (Kershaw, 2000: 114–117),
a legal inversion mentioned in Germanic, Latin, Vedic, Iranian, Celtic,
and Greek sources, and hinted at in Anatolian myths (Kershaw, 2000:
140–141). They returned to their homes to become adult men, after
being submerged in water in Celtic and Germanic myths, or after
burning their wolf-skin cloaks in the Volsunga Saga (Byock, 1990); or
they migrated and founded new settlements and attracted followers.
135
Journal of Anthropological Archaeology 48 (2017) 134–148
They were symbolically associated with death and symbols of death,
perhaps because they were represented as dogs, and dogs were them-
selves symbols of death (a linkage explored at the end of this paper).
They assumed dog or wolf names, garments, and symbols in Germanic,
Latin, Vedic, Iranian, Celtic, and Greek sources. Becoming like a wolf or
dog was a central metaphor for youthful war-bands, indicated by names
containing the element ‘dog’ or ‘wolf’ (Lincoln, 1991a: 134), the
wearing of dog or wolf skins (Bremmer, 1982: 136–137; Kershaw, 2000:
133, ff.), and even, in some late Vedic texts, the consumption of dogs
during initiation (White, 1991:72, 87).
Youthful war-bands are thought to have actually existed among
later IE-speaking groups, although whether they go back to Proto-Indo-
European (PIE) is debated (Zimmer, 2004). In Europe they seem to have
been absorbed into the military retinues of increasingly powerful pa-
trons and kings during the Iron Age (Bremmer, 1982; Bremmer and
Horsfall, 1987:39–42; Kershaw, 2000: 134; García Moreno, 2006). In
South Asia, Heesterman (1962) and Falk (1986) found that oath-bound
initiatory war-bands called Vrātyas were phased out and downgraded
with the rise of the Brahmin caste, a process that had started already
when the Rig Veda was compiled between about 1500–1200 BCE
(Witzel, 1995), leading to the eventual demise of the institution.
The initiation ceremonies connected with youthful war-bands were
not as widely referenced in comparative mythology as the war-bands
themselves. We know, for example, that young Spartan males sacrificed
a dog or dogs to an archaic god of war, Enyalios (mentioned in
Mycenaean Linear B documents), as they entered the age group of
ephebes, but no ceremonial details are preserved (Mazzorin and Minniti,
2002: 63). Perhaps the best-described ancient ritual that retained ele-
ments of a canid-related initiation ceremony was the February 15 Lu-
percalia (“wolf festival”) in ancient Rome. Described by Cicero, born
about 100 BCE, as the oldest Roman ritual, inherited from a pastoral era
D.W. Anthony, D.R. Brown Journal of Anthropological Archaeology 48 (2017) 134–148

“before civilization and the laws” (quoted in Dumézil, 1996: 347), the cemetery (Table 1). This was the beginning of the Late Bronze Age in
Lupercalia was dedicated to the obscure god Faunus, a deity of cattle the chronology of the steppes (Anthony, 2007; Kohl, 2007; Koryakova
and shepherds in the wild, uncultured places beyond the settled social and Epimakhov, 2007). Our interpretation of the canid rituals at
space (Hraste and Vuković, 2011; Wiseman, 1995). A dog (a symbol of Krasnosamarskoe begins with an overview of the site and the identifi-
death and war) and a goat (a symbol of generative fertility) were sa- cation of specific features that point toward a male-centered rite of
crificed and their blood was wiped on the foreheads of adolescent boys passage in which the participants were transformed symbolically into
representing two aristocratic lineages. They consumed a feast and then dogs and/or wolves. The core of the paper is a description of the faunal
ran naked around the base of the Palatine Hill, following the cultured/ assemblage that supports the assertions made in the introductory
uncultured border of the oldest settled part of Rome, and striking overview. Finally, we show that the IE warrior-as-dog/wolf narrative
women who desired to conceive with strips of the sacrificed goat skin theme offers better parallels with the archaeological evidence than do
(the strips were the februa, after which February is named) to impreg- other narrative themes from other mythological traditions.
nate them, a fertility aspect of the rite that evolved into modern Va-
lentine’s Day (Harrison, 1991[1903]:51–54; Dumézil, 1996[1966]:
2. Krasnosamarskoe as a central place for institutionalized rites of
346–350; Bremmer and Horsfall, 1987:25–48). The Lupercal cave,
passage
where the run began, was where the brothers Romulus and Remus, the
mythical first settlers of Rome, were suckled by a wolf after being
In this section the Krasnosamarskoe settlement is briefly described,
conceived by the war-god Mars and then being cast out into the wild by
and the faunal remains recovered there are summarized, emphasizing
an elite family. In the myth, the adolescent boys later recruited landless,
the aspects of the site that suggest that a canid-related, male-centered
unmarried, wandering young males (origins not described) to become
rite of passage occurred on multiple occasions at this place, uniquely,
Rome’s first citizens, whose first act was to raid the neighboring Sabine
and not in other known Srubnaya sites.
tribe for wives (the ‘rape of the Sabine women’). Rome’s origin myths
The excavated Srubnaya-culture structure at Krasnosamarskoe was
were deeply colored by narrative tropes apparently derived from the
a building 8 × 12 m in preserved dimensions, sectioned and partly
institution of IE youthful war-bands (Bremmer and Horsfall, 1987:
destroyed by erosion from a man-made lake that flooded the western
25–48), and the Lupercalia, linked directly to these origin myths both
half of the site, probably inundating the principal residential structure
topographically and symbolically, retained a dog sacrifice as one part of
(Fig. 2). LBA pottery and animal bones collected from the flooded
a complex ritual that included aspects of a boys’ initiation ceremony.
western part of the site yielded the same Srubnaya ceramic types and
Archaeological evidence for IE youthful war-bands has not been
radiocarbon dates (Table 1) as the excavated eastern portion. Srubnaya
identified before, so our interpretation of Krasnosamarskoe as the first
artifacts were found within a compact area on and offshore about
prehistoric site with evidence for initiation rituals linked to IE youthful
25 × 40 m in size, 207 m2 of which was excavated 20% of the esti-
war-bands will be greeted with skepticism. The dating of the faunal
mated occupied area. Excavations in 1999 and 2001 yielded 22,445
assemblage between 1900 and 1700 BCE is perhaps its least con-
animal bone fragments from the structure and the occupation surface
troversial trait, as it is well established by nineteen radiocarbon dates
around it, of which 7669 were identifiable to species (NISP), according
from the settlement and eight from the associated Srubnaya-culture
to Russian archaeozoologist Pavel Kosintsev of the Ekaterinburg

Table 1
Srubnaya radiocarbon dates from Krasnosamarskoe.

Kurgan or Unit # Grave #/Unit Quad Arizona Lab # AZ 14C age BP Cal BC Lower Cal BC Upper Probability Material

KRASNOSAMARSKOE IV CEMETERY HUMAN BONE

Kurgan 3 13 AA37044 3358 ± 48 1754 1514 95.40% 45+ yr. female
Kurgan 3 16 AA37045 3407 ± 46 1879 1612 95.40% 0–0.5 yr. unknown
Kurgan 3 6 AA37039 3411 ± 46 1879 1615 95.40% 15–17 yr. unknown
Kurgan 3 11 AA37043 3416 ± 57 1888 1608 94% 6–8 yr. unknown
Kurgan 3 23 AA37047 3425 ± 2 1883 1621 95.40% 35–45 yr. male
Kurgan 3 1 AA37038 3490 ± 57 1952 1664 95.40% 17–25 yr. female
Kurgan 3 17 AA37046 3545 ± 65 2039 1732 91.70% 35–45 yr. male
Kurgan 3 10 AA37042 3594 ± 45 2129 1777 95.40% 5–7 yr. unknown

KRASNOSAMARSKOE SETTLEMENT
O5 3 AA47790 3311 ± 54 1738 1456 95.40% Animal bone
M5 1 AA41023 3445 ± 51 1899 1643 95.40% Animal bone
N2 2 AA41032 3448 ± 37 1882 1682 94% Animal bone
O4 2 AA41028 3450 ± 57 1905 1626 95.40% Animal bone
M6 3 AA41024 3453 ± 43 1887 1662 95.40% Animal bone
O4 1 AA41027 3460 ± 52 1905 1641 95.40% Animal bone
N3 3 AA41025 3469 ± 45 1906 1681 94.50% Animal bone
P1 4 AA41029 3470 ± 43 1904 1683 95.40% Animal bone
R1 2 AA41031 3476 ± 38 1892 1692 95.40% Animal bone
S2 3 AA41030 3477 ± 39 1894 1691 95.40% Animal bone
N4 2 AA41026 3491 ± 52 1946 1686 95.40% Animal bone
M2 4 AA47794 3492 ± 55 1953 1682 94.90% Plant remains
L5 2 AA41022 3531 ± 43 2009 1744 95.40% Animal bone
M2 4 AA47795 3550 ± 54 2031 1744 95.40% Plant remains
M2 4 AA47793 3615 ± 41 2132 1883 95.40% plant remains

Lake Finds next to the structure

Lake find 2 AA47792 3492 ± 55 1953 1682 94.90% Animal bone
Lake find 1 AA47791 3494 ± 56 1957 1681 94.70% Animal bone

Y excavations south of settlement

Y2 2 AA47796 3416 ± 59 1889 1607 93.60% Animal bone
Y1 3 AA47797 3450 ± −50 1891 1638 95.40% Animal bone

136
D.W. Anthony, D.R. Brown Journal of Anthropological Archaeology 48 (2017) 134–148

Fig. 2. Aerial view of the Krasnosamarskoe site from about 400 m

elevation. For scale, the white area labeled Settlement Excavation is
26 m long, north-south. The Srubnaya cemetery was located in kurgan
3. The area of artifact finds on the lake bottom is shaded. The Y1 and
Y2 excavation was a low-density midden contemporary with the set-
tlement.

Table 2
Frequency of mammal species, domesticated and wild, at Krasnosamarskoe (after Kosintsev, 2016).

Mammals Total MNI % of all prehistoric mammals MNI Total NISP % of all prehistoric mammals NISP

Domestic
Cattle – Bos taurus 116 34.12% 2661 34.69%
Goat and sheep – Capra et Ovis 98 28.82% 1926 25.11%
Pig – Sus scrofa domestica 16 4.71% 123 1.60%
Horse – Equus caballus 23 6.76% 125 1.63%
Dog – Canis familiaris 51 15.00% 2770 36.12%
> dog; < wolf – Canis sp. 6 1.76% 6 0.08%

Wild
wolf – Canis lupus 7 2.06% 18 0.23%
Eurasian elk – Alces alces L. 6 1.76% 14 0.18%
roe deer – Capreolus pygargus 4 1.18% 10 0.13%
fox – Vulpes vulpes 3 0.88% 3 0.04%
steppe polecat – Mustela eversmannii 4 1.18% 7 0.09%
hare – Lepus sp. 6 1.76% 6 0.08%
sums 340 100% 7669 99.98%

137
D.W. Anthony, D.R. Brown
Table 3
Percentage of domestic mammal bones by NISP in LBA sites of the Middle Volga region (Kosintsev, 2016 Table 15.24).
Species Krasno-samar-skoe Moechnoe Ozero
Cattle – Bos taurus 35 53
Goat/sheep – Capra et Ovis 25 23
Horse – Equus caballus 2 15
Pig–Sus scrofa domestica 2 7
Dog – Canis familiaris 36 0.3
NISP 7655 4692
Institute of Ecology (Table 2).
The excavated structure is regarded as an outbuilding because it did
not have a hearth or fireplace but did contain at least one well (Pit 10)
that held open water during the occupation. Organics that fell into the
well were preserved in anaerobic mud at the bottom, and were a rich
source of botanical data. Botanical studies (Popova, 2016) showed that
no cultivated cereals were present at Krasnosamarskoe, a surprising
economic result in a permanent settlement occupied year-round for
multiple generations without an observable break. But a similar ab-
sence of cultivation was shown at other Srubnaya settlements with well-
studied flora, such as the Srubnaya mining settlement of Gorny in the
Kargaly copper ore field, 290 km east of Krasnosamarskoe (Lebedeva,
2005; Chernykh, 2004; Diaz del Río et al., 2006). Srubnaya settlements
in the Volga-Ural steppes exhibited no agriculture, a strong reliance on
meat and dairy foods, and some evidence for wild seed (Chenopodium)
collection (Anthony, 2016).
The Krasnosamarskoe settlement fauna presented an anomalously
high percentage of dogs, 15% by minimum number of individuals
(MNI), 36% by number of identifiable specimens (NISP), and 45% by
diagnostic zones (DZ). If we add all canids together, the percentages are
19% MNI and 37% NISP. These high percentages are unique. At six
other Srubnaya settlement components in the Samara oblast (an ad-
ministrative unit like a US state) with published fauna (Table 3), dogs
are < 1–3% (NISP) of total settlement fauna. Inter-site comparisons are
in NISP because this is the only number consistently published for other
Srubnaya sites in the Samara region. Similarly, in a survey of fauna
from 19 additional Srubnaya settlements distributed across the western
steppes from the Dnieper to the Ural rivers, dogs were universally less
than 1.5% NISP, and usually less than 1% (Morales Muniz and Antipina,
2003: Table 22.2).
The dog percentages at Krasnosamarskoe were so far beyond normal
that it was immediately apparent during the excavation. Speculative
conversation over the cataloguing table included the idea that dog skins
were processed at the site. Prior to the publication of the 2016 final
report this hide-processing hypothesis appeared in a conference paper
and a book (Russell, 2012: 290), but consideration of the entire canid
assemblage by Russel et al. (2016: 436–438) and Kosintsev (2016)
showed that it was not supported. The bones of the entire body were
present, they showed fileting cut marks typical of meat processing, they
were roasted, and they were chopped into small pieces that made no
utilitarian sense for hide processing or dietary needs. All three ar-
chaeozoological teams agreed that the canid bones were best explained
as a ritual assemblage, not a dietary or hide-processing assemblage, as
elaborated below.
In the settlement sample of 2794 canid bones, wolves (18 bones, less
than 1% of canids) and dogs (2770 bones, more than 99% of canids)
were differentiated by size. The standard LBA dog at Krasnosamarskoe
was medium-sized, or even medium-small-sized (Kosintsev, 2016:395),
and wolves were large, leaving only 6 canid bones of uncertain desig-
nation and intermediate size (Table 2).
More than 99% of the Krasnosamarskoe canid bones were from
dogs. Most were older individuals (83% were 6–12 years or older) and
therefore familiar companions, and had been treated well (few limb
injuries and no trauma to the head). The low frequency of dog-gnawing
marks on bones at the site, combined with the extraordinarily high
Journal of Anthropological Archaeology 48 (2017) 134–148
Lebyazhinka V, layers 4–6 Lebyazhinka V, layers 1–3 Suskanskoe I Poplav-skoe Sachkovo
62 67 52 60 52
20 18 28 27 25
12 9 16 10 15
4 4 2 2 5
0.1 0.5 0.5 0.7 3
4767 3183 1767 502 3795
number of dogs (51 MNI), suggests that many of the dogs that died at
Krasnosamarskoe lived elsewhere (Russell et al., 2016: 437). The dogs
were killed exclusively during the cold months. Three definite deaths
occurred in winter, eight indefinite deaths in late-autumn-or-winter
(could be either), and three more deaths occurred definitely in late
winter/early spring (Pike-Tay and Anthony, 2016). The dogs were
cooked, fileted, chopped, and apparently eaten. Most of the bones were
burned to a carbonized-calcined state, but some retained traces of
roasting, particularly on the heads. The heads were chopped intensively
with an axe, well beyond any practical dietary needs, using a standar-
dized, ritualized sequence of cuts that reduced the head to chunks 3–7
cm in size. Stratigraphy and radiocarbon dates in the excavated struc-
ture where the bones were discarded indicate that the dog sacrifices
occurred in all levels over a time span of decades so were repeated at
intervals throughout the history of the site.
Wolves were treated the same way. Although few in number com-
pared to dogs, wolves were the most frequent wild animal at
Krasnosamarskoe by MNI (7) and NISP (18), an unusual statistic for a
Srubnaya settlement. They might have been hunted or trapped in-
tentionally for use in ceremonies. Kosintsev and Russell both noted that
the wolf bones included at least one wolf skull and mandible that was
roasted, chopped and split like the dog skulls, and the other wolf bones
were chopped like the dog bones. Dog bones outnumbered wolf bones
150-1, so wolves were included quite rarely, or if they were included
regularly, they were eaten in much smaller quantities or by a single
participant, or both.
Several unusual characteristics of the canids at Krasnosamarskoe
suggest a rite of passage. A rite of passage is indicated by the repeated,
therefore institutionalized, inversion of the normal non-food treatment
of canids, including wolves. Dogs were not eaten at Krasnosamarskoe as
a starvation food by a population in extremis, because the cattle and
sheep/goat bones in the same archaeological deposits with the canid
bones show normal butchering patterns. Cattle were slaughtered
throughout the year at Krasnosamarskoe, with deaths in each season;
more young animals were culled in the fall, when they were fat, and
more old animals (10+ years) in the spring before the herd moved to
summer pastures, a normal form of herd management (Pike-Tay and
Anthony, 2016: 377). The dogs were killed only during the cold
months. Canids were intentionally and selectively chopped into small
pieces; cattle and sheep-goat bones in the same trash deposits were
butchered normally. The consumption and ritualized processing of ca-
nids therefore was intentional. Moreover, the ritual was not a single
exceptional event, but was repeated over the life of the site, which is
estimated at two generations based on cemetery demography and
radiocarbon dates in stratified floor deposits, described in more detail
below.
In places around the world where dogs are not eaten, as in other
Srubnaya settlements in this region, their consumption is often re-
garded with considerable revulsion and disgust (Russel et al., 2016:
438; Tambiah, 1969). A repeated dietary role for canids at Krasnosa-
marskoe would make it a place where normal dietary behavior toward
canids was suspended and inverted under unusual circumstances. The
shock attached to such an act in a culture that did not eat dogs was
increased by the intentional selection of older dogs for more than 80%
of the victims: familiar, well-treated, human-like companions and
138
D.W. Anthony, D.R. Brown
therefore perhaps stand-ins for human victims; rather than young dogs,
more suitable if starvation explained the behavior. Old, familiar dogs,
possibly even their own dogs, might have represented an emotionally
significant first death for boys learning to become killers of men. An
additional level of inversion was represented by eating wolves, sym-
bolic of anti-culture—a murderer ‘has become like a wolf’ in Hittite law,
and ‘wolf’ was used to refer to brigands and outlaws, people who stood
outside the law, in many other IE languages (West, 2007: 450). ‘Wolf-
like’ also described the berserk rage of the warrior in Greek, Germanic,
and Celtic myths, a fury that could make them invincible and extra-
ordinarily dangerous in battle, so wolves often were symbols of war and
warriors (Lincoln, 1991; West, 2007: 450–451; Cebrián, 2010). By
eating wolves and dogs, the consumers at Krasnosamarskoe were
transformed into the animals that they consumed. Personal transfor-
mation into a new status achieved through the inversion of customary
behavioral rules is typical in rites of passage and in almost no other
regularly occurring institutional context (van Gennep, 1909; Garwood,
2011).
If these inversions suggest that a rite of passage occurred at
Krasnosamarskoe, was it a male-oriented rite of passage? Typically rites
of passage are highly gendered, as they are one of the principal means
by which gender roles are publicly affirmed (Garwood, 2011: 4). One
critical indicator could be the sex of the sacrificed dogs. Since the bones
are too fragmented to separate by sex, we submitted 25 adult dog teeth
from Krasnosamarskoe to Pontus Skoglund and colleagues Erik Ersmark
and Love Dalén at Stockholm University and the Swedish Museum of
Natural History for DNA analysis. The results are presented here with
their permission. Of the 21 that passed screening, 15 (70%) were cer-
tainly male, four more (20%) were “uncertain but consistent with
male”, and two (10%) were female (Fig. 3). This is a significant excess
of males (chi2 = 9.9412, df = 1, P = 0.001616). The intentional se-
lection of male dogs for 70–90% of the sacrificed individuals suggests
that the canid deaths at Krasnosamarskoe were part of a male-directed
rite of passage.
Rites of passage are not often preserved recognizably in archae-
ological sites, so are comparatively neglected in the archaeological
study of ancient religion and ritual (but see Garwood, 2011;
Kristiansen, 2014; Langdon, 2010). If the widespread IE parallels cited
above can be used as a guide to interpret LBA behavior at this place in
the northern steppes, then the human-to-canid transformation ritual at
Krasnosamarskoe represented a public discourse on the ideal life path
for at least some boys, whose goal was to become wolves (symbolically)
and feed the dogs of war. This single-household settlement was the only
Fig. 3. A comparison of the ratio of X-chromosome (female) reads to autosomal reads in
canid DNA from Krasnosamarskoe. A result above 0.05 indicates female, below 0.05 in-
dicates male. 15 individuals are male, 2 are female, and 4 are undetermined but con-
sistent with male.
139
Journal of Anthropological Archaeology 48 (2017) 134–148
place in the region where the ritual was found, so if it was an initiation
site, then boys from other families presumably came here to be in-
itiated. It was argued above that at least some of the dogs also lived
elsewhere, perhaps with the boys—but the boys escaped and the dogs
did not. Krasnosamarskoe was strongly associated with dogs and death.
Below we describe details of the canid sacrifices to support some of
the assertions made above. At the end of the paper we consider some
alternate dog-related rituals and myths both within the world of com-
parative IE mythology and in other, non-IE northern Eurasian and
American cultures. These are interesting comparanda but do not pre-
sent as convincing parallels as the IE war-band myths. But before we do
either of those things, we briefly defend this attempt, still unusual
among European prehistoric archaeologists (but see Kristiansen, 2013),
to connect IE myths and languages with Bronze Age archaeology.
3. Indo-European languages and archaeological cultures
While archaeologists might be ambivalent about assigning a parti-
cular IE language to any particular archaeological entity, we can at least
accept—putting aside for the moment our poor understanding of the
social factors that mediate the relationship between language and
material culture (Anthony, 2017; Vander Linden, 2015)—that Indo-
European languages, rituals, and myths were an important part of the
Bronze Age cultural landscape that we study archaeologically. Non-IE
languages and language families survived and persisted into the Clas-
sical era particularly in southern Europe and the Mediterranean—in
eastern Iberia, southern France, the western Mediterranean islands,
northern Italy, Thessaly, and Crete (Ringe, 2013). Northern, central,
and western Europe might also have contained regions of non-IE
speech. But when the era of texts began, about 600–400 BCE, IE lan-
guages were the most commonly documented languages spoken over
much of Europe from modern France to Germany, Greece, the steppes,
Central Asia, Iran, and northern South Asia. Texts dated to the early and
middle 2nd millennium BC show that by 2000–1500 BCE the IE lan-
guages already were differentiated into at least three very divergent
branches: Anatolian, Greek, and Sanskrit (Burrow, 1973; Mallory and
Adams, 2006: 442–463; Anthony, 2007: 50–58; Bouckaert et al., 2012).
These mature 2nd-millennium BC daughters push the initial breakup of
Proto-Indo-European (PIE) into the Neolithic and/or Bronze Ages, even
without the computational phylogenetic linguistic studies of Bouckaert
et al. (2012), which dated PIE to 6500–6000 BCE (probably too early, in
our view). The breakup of PIE occurred in phases, not all at once, but
the initial differentiation, the separation of the earliest daughter branch
(probably Anatolian), must have occurred after the speakers of PIE
became familiar with agriculture and herding, so after 6500 BCE in
Europe. This post-quem date is required because many cognate word
roots securely assigned to the PIE vocabulary (Mallory and Adams,
2006:139–40, 167–169, 260–262; Benveniste, 1973: 17 & ff.), including
Anatolian, had meanings related to Neolithic economies (cow, bull,
calf, ewe, ram, lamb, wool, milk products, plow/ard, domesticated
grain).
Cognate terms for ‘wheel’ and ‘axle’ in all of the IE daughters except
Anatolian imply that all of the post-Anatolian IE languages are des-
cended from a single language community that existed after 4000–3500
BCE, when wheeled vehicles were invented, which pulls PIE itself to-
ward that date (Garret, 2006; Anthony, 2007; Anthony and Ringe,
2015; Ehret, 2015). PIE therefore expanded in at least two phases, the
first leading to the separation of what would later become Anatolian,
which uniquely preserved an archaic stage of PIE (Jasanoff, 2003); and
the second representing the expansion of the post-Anatolian PIE
daughters, after the invention of the wheel-and-axle principle, so after
4000–3500 BCE. Ancient DNA evidence (Haak et al., 2015; Allentoft
et al., 2015) now indicates that a massive migration flowed from the
wagon-dwelling Yamnaya culture of the western steppes into south-
eastern, central, and northern Europe after 3000 BCE, where a Yam-
naya assortment of genes accounted for at least 70% of Corded Ware
D.W. Anthony, D.R. Brown
genetic ancestry in Corded Ware graves across Germany, Scandinavia,
Poland, Estonia, and Latvia. The Yamnaya migration certainly in-
troduced a steppe-derived language with the steppe-derived Corded
Ware population, and there are independent reasons to identify that
language as a form of late PIE (Anthony, 2007; Mallory, 2008). If late
PIE spread with the Yamnaya culture and was spoken in the steppes
during the Yamnaya period (EBA, in steppe chronology), then the
people of the Volga-Ural steppes, where Krasnosamarskoe is located,
probably were IE-speaking long before the LBA. A geographic linkage
between the archaeological site and IE languages and institutions is
facilitated to the extent that the Yamnaya = late PIE hypothesis is
deemed persuasive. For those archaeologists who feel that this hy-
pothesis rests on too many ‘ifs’, we can at least agree that IE languages
must have been spoken across much of Bronze Age Europe and the
steppes. Institutions derived from IE comparative mythology therefore
are ‘on the table’ as possible sources of information about Bronze Age
behaviors documented archaeologically.
4. The dog and wolf bones at Krasnosamarskoe
In this section we provide data to support some of the preceding
descriptions of the Krasnosamarskoe canid sacrifices, particularly those
most relevant for a ritual interpretation generally, and for a rite-of-
passage interpretation specifically. Four topics are examined:
1. The repetition of the canid sacrifices over the duration of the set-
tlement, essential for interpreting the ritual as institutionalized,
depends on the stratigraphic distribution of canid remains, so their
stratigraphic context is discussed first.
2. The season of death can support a recurring seasonal ritual if deaths
cluster in one season or fail to support it if deaths show no pattern.
Here, as in the next two topics, we must also look briefly at the other
animals found in the same stratigraphic contexts to realize how the
canids compare with cattle and sheep-goat, the predominant do-
mesticated taxa.
3. The age at death and the body parts of the dogs compared with the
other fauna can reveal non-utilitarian and utilitarian patterns of
culling and usage.
4. Finally, cooking and butchering patterns are essential parts of the
argument for a ritual purpose involving feasting and ritualized
segmenting of the dogs’ heads and bodies.
4.1. Spatial and stratigraphic distribution
Faunal remains at Krasnosamarskoe were discarded in shallow pit
features both outside and inside the structure. Most of the catalogued
animal bones, about three quarters, were discarded inside the structure.
Clusters of animal bone were distributed in shallow features across the
dug-out floor (Fig. 4). Kosintsev noted that it was unusual to find so
many bones inside rather than outside a Srubnaya building. It is pos-
sible that the interior of the structure was used to discard ritually sa-
crificed animal remains to protect them from scavenging by living dogs.
Russell et al. (2016: 431) noted that bones from excavation units out-
side the structure exhibited more weathering and dog-sized gnawing
marks than those from inside, so the structure did perform this pro-
tective function, whether it was intentional or not.
The structure interior contained 60–80 cm of LBA cultural deposits,
which we divided into six to eight 10-cm horizontal levels. Filled-in
burrows of steppe rodents (Spermophilus major) about 10 cm in diameter
cut across all levels and were an obvious vector for moving bones and
artifacts both up and down between levels, a common problem in the
archaeology of the steppes, where burrowing rodents are ubiquitous.
However, the majority of the cultural deposit retained substantial
stratigraphic integrity (Fig. 5). Iron Age sherds reflecting ephemeral
activities in the fourth or fifth century BC were confined to the upper 30
cm, and the few Middle Bronze Age (MBA) sherds were concentrated
140
Journal of Anthropological Archaeology 48 (2017) 134–148
below the main LBA deposit. A small assemblage of Final Bronze Age
sherds, however, all moved downward into level 5, possibly down a
burrow. But the LBA deposit, with a much larger quantity of diagnostic
artifacts and therefore less vulnerable to small disturbances, exhibited
well-stratified earlier and later LBA ceramic types.
Early Srubnaya ceramics are divided into two slightly different
styles, Pokrovka and mature or developed Srubnaya, with Pokrovka
representing the transitional proto-Srubnaya style. In the stratigraphy
inside the excavated structure, Pokrovka styles were predominant by
both sherd count and weight in levels 7 and 8, and developed Srubnaya
sherds predominated by both sherd count and weight in the upper le-
vels, 3–6 (Fig. 5). This means that the brief occupation of the site
witnessed a phase shift in ceramics between primarily Pokrovka and
primarily developed Srubnaya types, well reflected in the LBA strati-
graphy (Anthony et al., 2016: 248–250). This is important for under-
standing what the two types represented culturally, but what interests
us here are the implications for the canid sacrifices. If the canid sacri-
fices at the site changed in frequency, or occurred only once as an ex-
ceptional event, we should be able to see that as clearly in the strati-
graphy as we can see the changes in ceramic styles.
Russell compared the animal taxa and body part percentages con-
tained in the upper and lower floor levels within the structure, and
inside and outside of the structure. She determined diagnostic zone
(DZ) counts for all taxa identified at the site, and compared species
percentages from levels 3–5 (upper floor levels) with those from levels
6–8 (lower floor levels), two stratigraphic sets that were about
equivalent in excavated volume. She found small differences in dog
bone frequencies between the upper and lower levels (Fig. 6, center and
right) but, unlike the ceramic sherd counts for the two Srubnaya styles,
the difference was not statistically significant (Russell et al., 2016: 422).
Dogs were 49% (DZ) of all fauna in the upper levels and 45% in lower
levels inside the structure. They were roasted, fileted, and chopped into
pieces in all levels, suggesting a recurring ritual that occurred episo-
dically throughout the life of the site. The canid sacrifices were not a
single exceptional act.
4.2. Seasonality
Anne Pike-Tay examined incremental banding in the roots of the
animal teeth under polarized light in a transmitting microscope (Pike-
Tay and Anthony, 2016). She analyzed 25 canid teeth, 21 Bos, and 6
Ovis from Krasnosamarskoe. In an effort to avoid using multiple teeth
from the same individuals we chose 25 canid teeth from 21 different
1 × 1 m excavation squares and levels (Fig. 4).
Taking the cattle and sheep-goat first, one Ovis had root surfaces
sufficiently intact for analysis. It was a 2–3-year old, slaughtered in the
Fall. The 13 successfully analyzed Bos teeth exhibited ages at death
distributed throughout the year, with at least one death in each season
(Table 4).
The canids showed a very different seasonal pattern. Fifteen had
root surfaces sufficiently intact to determine season of death. Fourteen,
or 93.3%, died in the cold half of the year. Three canid deaths were
definitely winter, eight were fall or winter (could be either), and three
occurred in late winter/early spring. One dog died in the late summer,
perhaps a normal settlement dog (Table 4).
The seasonal evidence for canids can be read as a minimum of two
clusters of deaths with the first one poorly resolved (winter? and late-
winter/early spring); or maximally distributed, it could be seen as a
sequence of canid deaths beginning in early winter or possibly late fall
and continuing at unknown intervals until late winter/early spring.
Winter witnessed three certain deaths and late winter/early spring
three more (perhaps in February, the month of the Lupercalia) but eight
deaths were assignable only to an undetermined cold season.
D.W. Anthony, D.R. Brown
4.3. Age, sex, and body part distribution
The body part distribution of the canids indicates that the whole
body was discarded (Table 5). Heads are somewhat over-represented,
particularly inside the structure, so might have played a special role,
but the discarded canid bones included vertebrae, ribs, and proximal
limb bones, suggesting that the whole body was present.1
1
A reviewer suggested adding MNI breakouts by body part to Table 5 in order to see if
MNI were similar for all body parts as an added check on body part representation. This is
141
Journal of Anthropological Archaeology 48 (2017) 134–148
Fig. 4. The excavated structure (dashed line is the edge of the dug-out
floor) with cultural pit features and artifacts found in the occupation
layers. Shaded squares show total animal bone weight in ranges for
each 1 × 1 m square. Circles are canid teeth with known season of
death. F10 was the well; F14 might have been a second well.
Most (83%) of the age-specific dog bones were from animals six
years old or older, including very old animals, 10–12 years old (Fig. 7).
Dogs have frequent litters of young, so usually exhibit a majority of
young animals, suggesting that at this site older dogs were intentionally
selected (Russell et al., 2016:438). Their teeth were noticeably worn by
an unusually abrasive (unknown) diet different from that of the human
site occupants, who did not display excessive tooth wear. They
(footnote continued)
a goal for a future study with P. Kosintsev, who created Table 5.
D.W. Anthony, D.R. Brown Journal of Anthropological Archaeology 48 (2017) 134–148

INSIDE STRUCTURE Fig. 5. Proportions of ceramic types in each level in-

LEVELS
side the structure by weight (gm s). Levels 9 and
CERAMIC TYPES 1 2 3 4 5 6 7 8 9+
deeper are limited to pit features. The maximum
IA/Gorodetskaya 15 3 weight for each type is highlighted in the table and
Final Bronze/Suskanskaya 216 occurs in expected stratigraphic order.
Developed Srubnaya 239 910 1668 3612 4251 5415 3134 1643 2851
Pokrovka/Early Srubnaya 11 308 498 2959 4286 6185 1662 297
MBA 11 12 32 73
Unidentifiable 70 63 137 212 919 1926 1134 542 334

0 2000 4000 6000 8000 10000 12000

IA/Gordetskaya
1

2
Final
Bronze/Suskansk
3 aya
Developed
4 Srubnaya

5 Pokrovka/Early
Srubnaya
6
MBA
7

8 Unidentifiable
9
+

Fig. 6. Distribution of animal taxa outside the structure (left) and in the upper and lower floor levels inside the structure (center and right).

Table 4 Table 5
Season of death for dogs vs bos and ovicaprids at krasnosamarskoe settlement. Cattle and Krasnosamarskoe settlement. Dog remains by body parts. (Canis familiaris) after
sheep-goat died year-round; canids in the cold season. (Pike-Tay and Anthony, 2016). Kosintsev (2016) Table 15.12.

Season of death Bos and ovicaprid Dog Skeletal unit %

Late Fall OR Winter 2 8 Head – skull and mandible 24%

Winter 1 3 Isolated teeth 5%
Late Winter/Early Spring 2 3 Trunk – vertebrae, ribs 25%
Spring 1 0 Proximal limb (scapula, pelvis, humerus, femur, radius, ulna) 25%
Late Spring/Early Summer 1 0 Carpalia, tarsalia, sesamoidea 5%
Summer 2 0 Distal limb (metapodia, phalanges 1–3) 17%
Late Summer/Early Fall 2 1 NISP 2770
Fall 3 0

the entirely pastoral Srubnaya economy) would represent an extra-

exhibited a very low number of fractures (2) on their extremities and no
trauma on their heads (Russell et al., 2016:434), so seem to have been
well treated during life.
In an exact contrast to the dogs, the cattle at Krasnosamarskoe were
culled at surprisingly young ages. Crabtree (2004) and Russell (2012:
386-87) argued that the slaughter of unusually young cattle in an
economic context where dairying was important for subsistence (as in
ordinary expenditure of resources, a social display probably reserved
for feasting. Fernández-Götz and Roymans (2015:27) identified Iron
Age ritual feasting sites in Belgium solely by the presence of unusually
large percentages of young cattle. Krasnosamarskoe fits this pattern.
Fully 50% of the Bos at Krasnosamarskoe were culled at 1.5 years
(18 months) or younger, much too young for milking (Table 6). Only
36% were 2.5 years (30 months) or older, potential milking age. At

142
D.W. Anthony, D.R. Brown
Fig. 7. Dog ages at death (Russell et al., 2016).
other LBA Srubnaya settlements in the region, 52–58% were 2.5 years
or older when they died (Table 6), consistent with mixed production for
both dairy and meat. Krasnosamarskoe exhibits unusual high counts of
what could be considered feasting food remains, supporting a ritual
interpretation of many (but not all) site activities. It is interesting that
Sachkovo, a Srubnaya settlement on the Volga River 140 km west of
Krasnosamarskoe, shows similarities in age structure and body part
representation (Tables 6 and 7), possibly an indicator of feasts there
also, at a settlement that showed a slight increase above average in
canid bones (Table 3).
The cattle bones at Krasnosamarskoe, again in contrast to the dog
bones, do not represent the whole body in near-expected proportions.
Instead the cattle bones are skewed toward an over-representation of
heads and distal front limbs (Table 7). The sheep-goat also exhibit an
abundance of distal front limbs, but not of heads over the site as a
whole, although sheep-goat head elements are overrepresented inside
the structure. The overrepresentation of head and distal limb bones
could suggest primary butchering refuse, although head-and-hoof ritual
deposits were used in this region throughout the Bronze and Iron Ages,
so that is another possible explanation. In head-and-hoof deposits the
flayed skin of the sacrificed animal with the head and hooves attached
was placed at a ritual site, a ritual practice documented ethno-
graphically across the Eurasian steppes (Piggott, 1962).
The bones of the proximal or upper limbs, the meatiest parts of the
animal, are significantly under-represented among the recovered bones
for sheep-goat and for cattle (Russell et al., 2016: 424). Moreover, for
both sheep-goat and cattle, the rear proximal limbs are under-re-
presented. Many of the thighs, the meatiest and most desirable part of
the carcass, seem to have been discarded elsewhere. The bias to fore-
limbs “…is interesting and does not appear to be taphonomic,” Russell
et al. (2016: 424) argued, as “…the softer hindlimb DZ such as the
patella are well represented.” In Homeric and later Greek sacrificial
rituals, the thighs of the sacrificed animal, ideally Bos, were removed,
and the bones were burned to ashes as an offering to the gods
(Bremmer, 2007: 137). Exactly this ritual is not likely to have been
practiced in the LBA steppes, but the removal of thigh bones for the
gods’ portion could explain their under-representation. Or perhaps they
were served to adult celebrants and discarded in another part of the
Table 6
Age structure (%) of cattle (Bos taurus) in Srubnaya settlements: Krasnosamarskoe, recently-excavated Sachkovo and Lebyazhinka V (Kosintsev, 2016 Table 15.26), and 26 LBA sites in
the region excavated before 1972 compiled by Tsalkin (from Kosintsev, 2016: Table 15.26).
Age (months) Krasno-samarskoe Sachkovo
<6 18 22
6–18 32 27
18–30 14 19
> 30 36 32
NISP 160 200
Journal of Anthropological Archaeology 48 (2017) 134–148
site, now under the lake. Either possibility only emphasizes the ritual
derivation of many of the faunal deposits located in the excavated
structure, not only the canids but also some (not all) of the cattle and
sheep-goats.
4.4. Butchery and cooking
Kosintsev recognized that all the animal bones at Krasnosamarskoe
were more fragmented than at other Srubnaya sites he had examined.
The high number of fragments was partly a result of the systematic use
of ¼″ screens for artifact recovery on 100% of the excavated sediments,
which occurred only at Krasnosamarskoe among excavated Srubnaya
settlements. But it was also apparent that the canids were more frag-
mented than the cattle and sheep-goats at Krasnosamarskoe. Kosintsev
(2016: 395) observed that dogs accounted for a much higher percentage
of the pieces (36% of NISP) than of the individuals (15% of MNI) be-
cause dogs were chopped into many more pieces than cattle or sheep-
goats. For him, the outstanding feature of the dog remains, “…besides
their sheer quantity, is their very strong fragmentation.” The head, the
trunk, and the long bones all were cleaved into many pieces. Re-
cognizing that MNI and NISP are different kinds of counts, we still
might compare them, as Kosintsev did, to indicate the relative degree of
fragmentation between taxa. Examined this way, dogs exhibit a mean of
54.3 pieces per animal (NISP/MNI, not an actual piece count), cattle
exhibited a mean of just 22.95, sheep-goat a mean of 19.65, and pig a
mean of 7.69. Although these are relative numbers only, not actual
pieces per animal, they do indicate that dogs were selectively frag-
mented, more so than cattle or other animals.
Russell et al. (2016) noted that most of the cooked bone at Kras-
nosamarskoe probably was boiled, which leaves no discoloration on the
bone. This might have been the normal secular cooking method. Traces
of burning were apparent on about one third of the bones (1226 of the
3765 NISP examined by Russell et al.). These range from slight toasting
through roasting and carbonization to calcination, with most in the
more heavily burned, carbonized-calcined range. Some bones, however,
show a pattern of burning that is likely to result from roasting. Roasting
is a cooking method often associated with feasting, as it is the fastest
way to cook large quantities of meat, but it is wasteful because fat and
other juices are lost to the fire (Russell, 2012: 377 & ff.). The roasted
specimens were chopped with an axe after roasting, as the chop mark
surfaces on roasted bones show no burning. Cattle (30%) and especially
canids (56%) were over-represented among the roasted specimens
compared to their proportions in the site fauna, and sheep/goat (12%)
and horse (2%) were under-represented. Roasting marks were seen
most frequently on canid skull parts, and secondly on cattle skull parts,
deposited inside the structure. Oddly, the roasted bones also included
the paws/hoofs of some dogs and cattle, which contain almost no edible
meat, indicating that the feet and heads were still attached to the car-
cass when it was roasted.
Unlike Bos and Ovis heads, the canid heads, including at least one
wolf, were carefully chopped into small, neat, geometrical segments
(Fig. 8). The pieces produced for each canid skull were somewhat
standardized. Although it is difficult to reconstruct the sequence of
blows with confidence after such intense chopping, the canid heads
apparently were split longitudinally into three parts with two cuts
Sites compiled by Tsalkin Lebyazhinka V layers 1–3 Lebyazhinka V layers 4–6
9 6 11
18 9 11
21 27 23
52 58 54
? 127 107
143
D.W. Anthony, D.R. Brown Journal of Anthropological Archaeology 48 (2017) 134–148

Table 7
Body parts (%) of cattle and sheep-goat in srubnaya settlements, middle Volga region. Tsalkin compilation is 26 LBA sites in the region excavated before 1972 (from Kosintsev, 2016:
Tables 15.25 and 15.28).

Skeletal unit Krasno-samarskoe Sachkovo Lebyazhinka V Tsalkin compilation

Cattle
Head – skull and mandible 42 43 24 22
Trunk – vertebrae, ribs 13 12 18 16
Proximal limb (scapula, pelvis, humerus, femur, radius, ulna) 16 23 31 31
Distal limb (metapodia, phalanges 1–3) 29 22 27 31
NISP 2661 1041 2239 3865

Sheep-goat
Head – skull and mandible 35 33 25 36
Trunk – vertebrae, ribs 15 21 16 8
Proximal limb (scapula, pelvis, humerus, femur, radius, ulna) 25 31 45 44
Distal limb (metapodia, phalanges 1–3) 25 15 15 12
NISP 1846 619 943 1886

segmenting of the canid skulls into standardized pieces ca. 3–7 cm in

size was the most interesting feature of the butchering and processing at
Krasnosamarskoe. There is very little meat on most of these pieces, so it
was not food-related butchering, but it was skilled, practiced, and
standardized, a redundant ritual act.
Moreover, many other dog bones were intensely chopped. Whole
canid bones were limited to the phalanges, carpals, tarsals, sesamoid
bones and metapodia. The trunk was cut into many pieces, and some
vertebrae were split down the middle, as if sectioning the canids into
two halves along the spine, and then were cross-cut to segment the
spine (Russell et al., 2016: 429). Many tarsal, carpal, and long bones of
the legs were cleaved into several parts. It looks as if the dogs and at
least one wolf were roasted, fileted, chopped, and eaten. The bodies and
particularly the heads were chopped into many small pieces. It is pos-
sible that many but not all were thrown into a fire, burning the bones to
a calcined state, before they were discarded in shallow pits inside the
roofed structure (perhaps a well-house) at Krasnosamarskoe. The dog
sacrifices seem to have been accompanied by scorched offerings of
cattle and perhaps sheep-goat, and rarely a horse, which were not
chopped as intensely as the dogs, but were discarded within the same
structure.

5. The people of the Srubnaya cemetery, Krasnosamarskoe IV

The Srubnaya cemetery at Krasnosamarskoe, like the settlement,

Fig. 8. Standardized chopped dog skull fragments (top) and one of the bone scatters
under excavation. Middle shows dog skull with common chopping sites drawn.
through the eye orbits on either side of the palate (see the fragments in
Fig. 7). These three longitudinal sections were cross-cut into three parts
by a second cut through the eye socket at a right angle to the first, a cut
through the nose, and another through the back of the head. The latter
two cross-cuts were recognized as recurring on multiple specimens by
Kosintsev (2016: 395). The pieces that resulted sometimes were again
chopped in half, particularly the mandible. The mandibles were re-
moved with a chop across the ramus or behind the molars, and the
detached mandible was then chopped in half at the midline and
sometimes chopped again. For Russell et al. (2016: 427), the careful
exhibited unusual features. Three kurgans were erected in a line about
30 m apart running NW-SE during the early MBA, 1000 years before the
Srubnaya settlement was established. Surveys by archaeologists from
Samara designated the cemetery we excavated as Krasnosamarskoe IV
because it was the fourth kurgan cemetery they found near the modern
village of Krasnosamarskoe. We discovered the nearby Srubnaya set-
tlement only after the Krasnosamarskoe IV cemetery attracted our GPS
survey team to the site in 1996. All three kurgans were entirely ex-
cavated in 1999, and radiocarbon dates from each of the three central
graves overlapped in a short period at the opening of the MBA, about
2800–2700 BCE (Kuznetsov et al., 2016: 448–458). The kurgans had
been untouched for centuries when, around 1900–1800 BCE, a Srub-
naya settlement was established 60 m southwest of the cemetery.
Kurgan 3 was chosen as the site for a Srubnaya cemetery at the same
time (Fig. 9). The cemetery contained 27 Srubnaya individuals in 22
graves dug into the northeast side of the kurgan, opposite the settle-
ment. Eight radiocarbon dates from the Srubnaya graves (Table 1)
clustered between 1900 and 1600 BCE, contemporary with the settle-
ment but perhaps continuing briefly after it was abandoned.
The first unusual aspect of the Srubnaya cemetery was the high
percentage of sub-adults. Fully 82% (22 of 27 individuals) were chil-
dren, most of them (15 of 27) quite young, aged 1–7 (Table 8). Three
other Srubnaya cemeteries in this part of the Samara River valley, each
containing 20–25 graves and paired with a settlement, exhibited sub-

144
D.W. Anthony, D.R. Brown Journal of Anthropological Archaeology 48 (2017) 134–148

Fig. 9. Plan of Kurgan 3 with secondary Srubnaya cemetery

of 22 graves.

Table 8 adult percentages of 17%, 48%, and 50%, significantly lower than
Krasnosamarskoe IV, Kurgan 3, Srubnaya individuals. Krasnosamarskoe (Kuznetsov et al., 2016: 469). Only one of the sub-
adults in our cemetery, an adolescent aged 12–14 in grave 2, exhibited
Grave # Age Age group Sex % of skeleton
clear signs of illness, in this case resulting in a very small and stunted
3 6–7 Younger than 7 years. 15/27, Unknown 100% stature incommensurate with the dental age. All of the young in-
3 6–7 55.55% Unknown 100% dividuals must have died of either disease or injury, but evidence for
4 5–7 Unknown 40–50%
mortal forms of either was rare. One explanation for the high percen-
7 5–7 Unknown 10–20%
7 5–7 Unknown 10–20% tage of sub-adults could be that some children were brought to the site
10 5–7 Unknown 10–20% for treatment for illnesses and died there. This would be more likely if
15 1.5–2.5 Unknown 40–50% the occupants of the settlement were identified as ritual specialists.
16 0–0.5 Unknown 1 femur The five adults in the Srubnaya cemetery also displayed unusual
18 1.5–3.5 Unknown 30–40%
features. The adults consisted of two males, two females, and a few leg
18 4 Unknown 1 femur
19 4.5–6.5 Unknown 30–40% bones of a fifth adult. Both adult men displayed pathologies that were
20 3–4 Unknown 40–50% rare among 88 Srubnaya individuals from 16 Srubnaya cemeteries in
21 3–4 Unknown 50–60% the Samara region examined by Eileen Murphy of Queen’s University/
22 5–6 Unknown 60–70%
Belfast (Murphy, 2016). The men from graves 23 and 17 at Krasnosa-
24 0–0.5 Unknown Partial
2 12–14 7–14 years, 5/27, 18.52% Unknown 70–80%
marskoe IV were 35–45 years old when they died. They had similar
11 6–8 Unknown 30–40% lower back injuries (Schmorl’s nodes on thoracic and lumbar vertebrae,
12 8–10 Unknown 50–60% which affected about 1/3 of Srubnaya males examined); inflamed
12 7–9 Unknown 50–60% Achilles tendons indicated by enthesopathies on calcanei (not seen on
16 7–8 Unknown 50–60%
any other Srubnaya individuals); and twisting knee injuries (os-
6 15–17 15–17 years, 2/27, 7% Unknown 100%
14 15–17 Male 50–60% teochondritis dissecans, seen on less than 1% of Srubnaya tibias ex-
1 17–25 Older than 17 years, Adults. 5/27, Female 100% amined excluding Krasnosamarskoe IV), suggesting unusual stress on
5 Adult 18.52% Unknown 10–20% the knees, ankles, and lower back (Murphy, 2016: 180, 186). Interest-
13 45+ Female 30–40%
ingly, the incomplete fifth adult at Krasnosamarskoe also had a twisting
17 35–45 Male 100%
23 35–45 Male 70–80%
knee injury. Both complete adult male skeletons also had one thigh li-
gament chopped post-mortem, a treatment applied to only 5% of ex-
amined Srubnaya femurs excluding Krasnosamarskoe IV, but to both
men here (Kuznetsov et al., 2016: 466–469). This was not done to
compress the legs against the chest of the deceased, as both men were

145
D.W. Anthony, D.R. Brown
buried in lightly contracted but relatively relaxed positions. It is pos-
sible that the unusually strained activities of these two males and their
equally unusual post-mortem treatments were related in some way to
unusual behaviors and statuses connected to the boys’ initiations in the
neighboring settlement.
Pollen of a medicinal plant, Seseli, was found in a single feature, Pit
1, in the excavated settlement structure. Seseli has sedative and muscle-
relaxing qualities that might have been used to sedate the sacrificed
animals or human patients (Popova, 2016: 342). It does not grow lo-
cally so probably was imported from a region with chalky soils, pre-
ferred by Seseli plants. Chalk regions can be found today on the west
bank of the Volga, 140 km west of Krasnosamarskoe, and in the steppes
about 100 km to the southeast. Imported medicinal plants are con-
sistent with the hypothesis that the adult occupants of Krasnosamarskoe
were part-time ritual specialists who oversaw occasional initiation
ceremonies.
6. Canid myths across Eurasia: Which is the best parallel?
Dogs and wolves appear in myths and rituals across Eurasia. Why
should we choose the IE youthful war-band myth as the best parallel for
the archaeology at Krasnosamarskoe? Might other canid-related myths
offer equally convincing parallels, narrative element for artefactual
element? An answer requires a brief evaluation of other canid-related
myths and rituals both within the IE world and outside it.
Within comparative IE mythology, dogs were not only connected
with war-bands and war, they were also, and perhaps more funda-
mentally, potent symbols of death. In Greek, Indic, Iranian, and
Germanic myths (and less clearly in later Baltic, Lithuanian, and
Albanian superstitions) a multi-headed (sometimes multi-eyed) dog or
dogs guarded the entrance to the afterlife and challenged the dead
before letting their spirits enter (Schlerath, 1954; Lincoln, 1991b;
Mallory and Adams, 2006: 439.). The name of the hell-hound in Greek
(Kérberos) and Indic (Śárvara) was a cognate name, derived by regular
sound changes from a shared PIE root name meaning ‘spotted’, a trait of
domesticated dogs, showing that the PIE hell-hound was specifically a
dog, not a wolf (Mallory and Adams, 2006: 439; West, 2007: 392). In
Mediterranean Europe, particularly in Greece and Italy, this chthonic/
death aspect of dogs probably explains the occasional sacrifices of dogs
in the bottom of ritual wells (Chenal-Velarde, 2002; Chilardi, 2002;
Mazzorin and Minniti, 2002), most of which were connected with the
worship of Hecate, goddess of ghosts and death.
It is interesting that a similar mytheme is found across North
America (in Souian, Algonquian, and Iroquoian myths), as well as in
Central America and South America (Lankford, 2011:245-246; 2007:
183-85). In these very widespread myths, a fierce guard dog is located
in the Milky Way, the path of souls, guarding the entrance to the
afterlife. Getting past it was a test for the souls of the dead, much as in
the IE myths about the hell-hound. The guardian-of-the-afterlife theme
is also found spottily among the myths of Siberia, particularly among
the Chukchi and Tungus, who believed in both the guardian-of-the-
afterlife dog and a dog that would absorb the dead man’s spirit and act
as guide-to-the-afterlife (Stutley, 2003: 102–105). Adult dogs and
wolves both were sacrificed in mortuary rituals of north Eurasian Me-
solithic foragers near Lake Baikal (Losey et al., 2011) and at Mergen-6
farther west in Siberia near the mouth of the Ishim (En’shyn, 2012).
Ancient human DNA from Mal’ta, an Upper Paleolithic cemetery near
Lake Baikal dated 24,000 years ago (Raghavan et al., 2014), showed a
cluster of genetic traits termed Ancient North Eurasian (ANE), found to
account for 40% of the genetic ancestry of Native Americans, and for
70% of the genetic ancestry of the indigenous ceramic-Neolithic for-
agers of the Samara region (Haak et al., 2015). The presence of a large
ANE component defines a genetic bridge of connected mating networks
that was located in northern Eurasia during the Upper Paleolithic and
extended into the Pontic-Caspian steppes on the west and into the
Americas on the east. Perhaps the dog-as-guardian-of-the-afterlife myth
146
Journal of Anthropological Archaeology 48 (2017) 134–148
was associated with this northern Eurasian network.
In all of these archaeological contexts, whether IE or non-IE, the
dog-as-the-guardian-of-the-afterlife was clearly separated from daily
life and associated with graves or sacrificial wells that penetrated into
the underworld. At Krasnosamarskoe, sacrificed canid remains were
discarded in a multi-purpose building in the settlement, also used for
secular tasks and trash. No dog remains or canid symbols were en-
countered in the cemetery, which was entirely excavated. The archae-
ological context at Krasnosamarskoe does not match the expectations of
the guardian-of-the-afterlife or dog-as-death themes, which should be
associated with death-related myths and rituals.
A second canid-related series of beliefs, myths, and rituals connects
dogs with healing rather than death, but also regards them as impure,
exactly because they had the power to absorb illness from humans. Near
Eastern and Turkic-Altaic myths associated dogs with healing rather
than death and generally also categorized dogs as impure (Ornan, 2004;
Golden, 1998: 186–192; Hobgood-Oster, 2014). In Mesopotamia, the
goddess Gula, associated with healing, was symbolized by a dog, and
near her temples many dog graves have been found (Ornan, 2004). In
the Hittite ritual of Zuwi, possibly influenced by these beliefs, a puppy
was held up to a sick person and was directed to lick the injured or
diseased parts of his body thereby licking up the illness and absorbing
it. The puppy was then killed in order to destroy the illness (Collins,
1990). Similar absorbent-puppy healing and sacrifice rituals were found
in Greece and Italy (Mazzorin and Minniti, 2002), again possibly in-
fluenced by Near Eastern traditions. Whole dogs buried under house
thresholds and in foundation pits beneath houses in the Eneolithic site
of Botai in Kazakhstan, dated ca. 3500 BCE (Olsen, 2000), might re-
present the dog as absorber of illness, a guardian of the household
against disease and evil.
In these contexts, where dogs were symbolized as healers and si-
multaneously as impure absorbers of illness, they were not eaten. At
Krasnosamarskoe, the dogs were fileted, roasted, and apparently were
eaten. In many of these Near Eastern/Mediterranean contexts, puppies
were preferred as illness-absorbers, but the sex of the puppy was irre-
levant. At Krasnosamarskoe, old dogs, and male dogs, were specifically
selected for sacrifice. Again, specific traits of the canid deposits at
Krasnosamarskoe contradict the expectations of a dog-as-healer ritual.
Several aspects of the archaeological record at Krasnosamarskoe
correspond with the expectations of the dog/wolf-as-warrior myth. One
is that wolves were included. Wolves were not regarded as healers or as
guardians of the afterlife in ancient mythologies, but they were widely
associated with violent death and war, and in IE mythology, were a
symbol of the savage fury that the warrior desired in battle (Spiedel,
2002). The dogs and wolves were eaten at Krasnosamarskoe in a clear
inversion of normal eating practices, a strong signal of a rite of passage
centered on transformation into dogs/wolves, exactly as the warrior is
symbolized in IE myths. The fact that male dogs were selected for sa-
crifice identifies this as a male-centered ritual, and the fact that old,
familiar dogs were selected makes it appear that the dogs symbolized
cultured, that is human, victims. All of this aligns with the dog-as-
warrior myth better than the other possibilities.
The dog-as-warrior myth might have evolved from a more ancient
Eurasian symbolic linkage between dogs and death. War and death,
while often separated, are also frequently connected. An evolution from
dog-as-death symbolism to dog/wolf-as-warrior symbolism seems to
have happened independently in parts of North America, where dogs
and wolves became symbols of warrior sodalities in Algonquian and
Souian war rituals (Comba, 1991; Cook, 2012). The dog/wolf-as-war-
rior myth probably was an independent New World development be-
cause it was not widespread in North America or in Siberia in-
dependently of the IE institution.
7. Conclusion
Krasnosamarskoe provides archaeological support for the Bronze
D.W. Anthony, D.R. Brown
Age existence of a winter-season initiation ritual involving transfor-
mation into a status symbolized by becoming a male wolf and/or dog.
The best explanation for the meaning of this ritual is that it was an
initiation into the IE institution of the youthful war-band, strongly
connected with wolf and dog symbols in comparative IE mythology. If
the LBA canid sacrifices at Krasnosamarskoe were motivated by the IE
mythic narrative that we suggest, then the institution of canid-themed
youthful war-bands can be projected into the Bronze Age as one aspect
warfare and social organization in the steppes, and perhaps in other
regions where Indo-European mythic narratives and institutions oper-
ated.
If initiations into youthful war-bands were taking place throughout
the Bronze Age, why is Krasnosamarskoe unique? Possibly comparable
sites have been found, but were not interpreted this way. Recent ex-
cavations on the lower Don River have found “large numbers of dog
bones” in a post-Srubnaya settlement of the Final Bronze Age,
1400–1000 BCE, but there is no additional information on the season of
death or the exact percentages (van Hoof et al., 2012: 73). In the lower
Rhine valley Fernández-Götz and Roymans (2015: 27-29) have mapped
several pre-Roman Iron Age open-air ritual-political centers of the
Germanic Batavians, where the presiding deity, Magusanus, had attri-
butes linked to war and the initiation of youths into warrior status.
Many multi-use ritual centers might have hosted youthful-war-band
initiations that are now difficult to sort out from other ritual behaviors.
Evidence for youthful warriors might be found in the famous stone
tomb at Breradör on Kivik in Sweden, dated about 1300–1000 BCE. The
human teeth and bones were discovered to be from adolescents aged
13–15 (Goldhahn, 2009: 369), not from kings, as was previously
thought. The puzzling images inside the tomb could show the initiation
rites of adolescent boys, and the tomb itself could be a monument to
youthful dog-warriors.
Our comparison of the archaeological remains at Krasnosamarskoe
with comparative IE mythological evidence for canid-related youthful
war-bands suggests that correlations between prehistoric behaviors
perceived archaeologically and IE institutions and myths perceived
through philology and comparative mythology can successfully unlock
the indigenous narratives that gave meaning to otherwise obscure
Bronze Age ritual archaeological assemblages.
Note
We thank Michael Witzel and two anonymous reviewers for their
very helpful comments on an earlier draft of this paper. Field excava-
tions at Krasnosamarskoe and for the Samara Valley Project generally
were funded by grants from the National Geographic Society, the US
National Science Foundation (NSF Project BCS-9818527), the Russian
Academy of Sciences (RFBR #15-06-01916 and RFH #15-11-63008), a
state grant from the Ministry of Education and Science of the Russia
Federation (33.1195.2014/k) and the facilities of the Institute for the
History and Archaeology of the Volga (IHAV) in Samara, Russia. We
also thank Nerissa Russell and Emmett and Audrey Brown in the US and
Pavel Kosintsev in Russia for their careful analyses and advice.
References
Allentoft, M.E., Sikora, M., Sjögren, K.G., Rasmussen, S., Rasmussen, M., Stenderup, J.,
Damgaard, P.B., Schroeder, H., Ahlström, T., Vinner, L., Malaspinas, A.-S.,
Margaryan, A., Higham, T., Chivall, D., Lynnerup, N., Harvig, L., Baron, J., Della
Casa, P., Dąbrowski, P., Duffy, P.R., Ebel, A.V., Epimakhov, A., Frei, K., Furmanek,
M., Gralak, T., Gromov, A., Gronkiewicz, S., Grupe, G., Hajdu, T., Jarysz, R.,
Khartanovich, V., Khokhlov, A., Kiss, V., Kolář, J., Kriiska, A., Lasak, I., Longhi, C.,
McGlynn, G., Merkevicius, A., Merkyte, I., Metspalu, M., Mkrtchyan, R., Moiseyev, V.,
Paja, L., Pálfi, G., Pokutta, D., Pospieszny, Ł., Price, T.D., Saag, L., Sablin, M.,
Shishlina, N., Smrčka, V., Soenov, V.I., Szeverényi, V., Tóth, G., Trifanova, S.V.,
Varul, L., Vicze, M., Yepiskoposyan, L., Zhitenev, V., Orlando, L., Sicheritz-Pontén, T.,
Brunak, S., Nielsen, R., Kristiansen, K., Willerslev, E., 2015. Population genomics of
Bronze Age Eurasia. Nature 522, 167–172.
Anthony, David W., 2007. The Horse, the Wheel, and Language. How Bronze Age Riders
from the Eurasian Steppes Shaped the Modern World. Princeton University Press,
147
Journal of Anthropological Archaeology 48 (2017) 134–148
Princeton, NJ.
Anthony, David W., 2016. The Samara Valley Project and the evolution of pastoral
economies in the western Eurasian steppes. In: Anthony, David W., Brown, D.,
Kuznetsov, P., Mochalov, O., Khokhlov, A. (Eds.), A Bronze Age Landscape in the
Russian Steppes: The Samara Valley Project. Cotsen Institute of Archaeology Press,
Monumenta Archaeologica 37, Los Angeles, pp. 3–36.
Anthony, David W., 2017. Archaeology and language: why archaeologists care about the
Indo-European problem. In: Crabtree, Pam J., Bogucki, Peter. (Eds.), European
Archaeology as Anthropology: Essays in Memory of Bernard Wailes. University of
Pennsylvania Museum, Philadelphia, pp. 39–69.
Anthony, David W., Brown, D., Kuznetsov, P., Mochalov, O., Khokhlov, A. (Eds.), 2016. A
Bronze Age Landscape in the Russian Steppes: The Samara Valley Project. Cotsen
Institute of Archaeology Press, Monumenta Archaeologica 37, Los Angeles.
Anthony, David W., Ringe, Don., 2015. The Indo-European homeland from linguistic and
archaeological perspectives. Annual Review of Linguistics 1, 199–219.
Benveniste, Emile., 1973. Indo-European Language and Society (Translated by Elizabeth
Palmer). University of Miami Press, Coral Gables, FL.
Bouckaert, R., Lemey, P., Dunn, M., Greenhill, S.J., Alekseyenko, A.V., Drummond, A.J.,
Gray, R.D., Suchard, M.A., Atkinson, Q.D., 2012. Mapping the origins and expansion
of the Indo-European language family. Science 337 (6097), 957–960.
Bremmer, Jan N., 2007. Greek normative animal sacrifice. In: Ogden, D. (Ed.), A
Companion to Greek Religion. Blackwell, London, pp. 132–144.
Bremmer, Jan N., 1982. The suodales of Poplios Valesios. Zeitschrift für Papyrologie und
Epigraphik 47, 133–137.
Bremmer, Jan., Horsfall, N.M., 1987. Roman Myth and Mythography. Institute of
Classical Studies Bulletin Supplement 52, London.
Burrow, T., 1973. The Proto-Indoaryans. J. Roy Asiatic Soc. New Series (2), 123–140.
Byock, Jesse L., 1990. The Saga of the Volsungs: The Norse Epic of Sigurd the Dragon
Slayer, Kindle Edition. Penguin Classics.
Cebrián, Reyes B., 2010. Some Greek evidence for I-E youth contingents of shape-shifters.
Journal of Indo-European Studies 38 (3 & 4), 343–357.
Chenal-Velarde, Isabelle, 2002. Food, rituals? The expolitation of dogs from Eretria
(Greece) during the Helladic and Hellenistic periods. In: Dogs and People in Social,
Working, Economic or Symbolic Interaction, edited by Lynn M. Snyder and Elizabeth
A. Moore, pp. 24–31. Durham: 9th ICAZ Conference.
Chernykh, E.V., 2004. Kargaly: the largest and most ancient metallurgical complex on the
border of Europe and Asia. In: Linduff, Katheryn M. (Ed.), Metallurgy in Ancient
Eastern Eurasia From the Urals to the Yellow River, Chinese Studies 31. Edwin Mellen
Press, Lewiston, pp. 223–279.
Chilardi, Salvatore, 2002. Artemis pit? Dog remains from a well in the ancient town of
Siracusa (Sicily). In: Dogs and People in Social, Working, Economic or Symbolic
Interaction, edited by Lynn M. Snyder and Elizabeth A. Moore, pp. 32–37. Durham:
9th ICAZ Conference.
Collins, Billie Jean, 1990. The puppy in Hittite ritual. J. Cuneiform Stud. 42 (2), 211–226.
Comba, Enrico, 1991. Wolf warriors and dog feasts: animal metaphors in Plains military
societies”. Eur. Rev. Native Am. Stud. 5 (2), 41–48.
Cook, Robert A., 2012. Dogs of war: potential social institutions of conflict, healing, and
death in a Fort Ancient village. Am. Antiq. 77 (3), 498–523.
Crabtree, Pam J., 2004. Ritual feasting in the Irish Iron Age: re-examining the fauna from
Dún Aillene in light of contemporary archaeological theory. In: O’Day, S.J., Van Neer,
W., Ervynck, A. (Eds.), Behavior Behind Bones: The Zooarchaeology of Ritual,
Religion, Status and Identity. Oxbow, Oxford, pp. 62–65.
Das, R.P., Meiser, G. (Eds.), 2002. Geregeltes Ungestüm. Brüderschaften Und
Jugendbünde Bei Altindogermanischen Völkern. Hempen, Bremen.
Diaz del Río, P., García, P.L., López Sáez, J.A., Martina Navarette, M.I., Rodrígues Alcalde,
A.L., Rovira-Llorens, S., Vicent García, J.M., de Zavala Morencos, I., 2006.
Understanding the productive economy during the Bronze Age through archae-
ometallurgical and paleo-environmental research at Kargaly. In: Peterson, D’L.,
Popova, L.M., Smith, A.T. (Eds.), Beyond the Steppe and the Sown: Proceedings of the
2002 University of Chicago Conference on Eurasian Archaeology. Leiden, Brill, pp.
343–357.
Dumézil, G., 1996. Archaic Roman Religion (Translated From Orig. 1966). Johns Hopkins
Press, Baltimore.
Ehret, Christopher, 2015. Agricultural origins: what linguistic evidence reveals. In: In:
Barker, Graeme, Goucher, C. (Eds.), A World with Agriculture, vol. v. II. The
Cambridge World History, Cambridge, pp. 55–92.
En’shyn, D.N., 2012. K voprosu o sakralizatsii zhiloi sredy naseleniya Nizhnego
Prishim’ya v epokhu neolita. In: Kul’turny Stepnoi Evrazii i Ikh Vzaimodeeistvie s
Drevnimi Tsivilizatsiyami, vol. 2. Institut Istorii Material’noi Kul’tury RAN, St.
Petersburg, pp. 76–79.
Falk, Harry., 1986. Bruderschaft Und Würfelspiel. Hedwige Falk, Freibug.
Fernández-Götz, Manuel, Roymans, Nico, 2015. The Politics of Identity: late Iron Age
Sanctuaries in the Rhineland”. J. North Atlantic 8, 18–32.
García Moreno, Luis A., 2006. Celtic place and personal names in Spain and the socio-
political structure and evolution of the Celtiberians. e-Keltoi 6, 675–688.
Garrett, Andrew, 2006. Convergence in the formation of Indo-European subgroups:
phylogeny and chronology. In: Forster, Peter, Renfrew, Colin (Eds.), Phylogenetic
Methods and the Prehistory of Languages. McDonald Institute for Archaeological
Research, Cambridge, pp. 139–151.
Garwood, Paul, 2011. Rites of Passage. In: Insoll, Timothy (Ed.), The Oxford Handbook of
the Archaeology of Ritual and Religion. Oxford University Press, Oxford, pp. 1–19.
Golden, Peter, 1998. Religion among the Qipchaks of Medieval Eurasia. Central Asiatic J.
42 (2), 180–237.
Goldhahn, Joakim, 2009. Breradör on Kivik: a monumental cairn and the history of its
interpretation. Antiquity 83, 359–371.
Haak, Wolfgang, Lazaridis, I., Patterson, N., Rohland, N., Mallick, S., Llamas, B., Brandt,
D.W. Anthony, D.R. Brown
G., Nordenfelt, S., Harney, E., Stewardson, K., Fu, Q., Mittnik, A., Bánffy, E.,
Economou, C., Francken, M., Friederich, S., Pena, R.G., Hallgren, F., Khartanovich,
V., Khokhlov, A., Kunst, M., Kuznetsov, P., Meller, H., Mochalov, O., Moiseyev, V.,
Nicklisch, N., Pichler, S.L., Risch, R., Guerra, M.A.R., Roth, C., Szécsényi-Nagy, A.,
Wahl, J., Meyer, M., Krause, J., Brown, D., Anthony, D., Cooper, A., Alt, K.W., Reich,
D., 2015. Massive migration from the steppe was a source for Indo-European lan-
guages in Europe. Nature 522 (7555), 207–211.
Harrison, Jane Ellen, 1991. Prolegomena to the Study of Greek Religion (Orig. 1903).
Princeton University Press, Princeton.
Heesterman, Jan, 1962. Vrâtyas and sacrifice. Indo-Iran. J. 6, 3–37.
Hobgood-Oster, Laura, 2014. A Dog’s History of the World: Canines and the
Domestication of Humans. Baylor University Press, Texas, Waco.
Hraste, Daniel Neças, Vuković, Krešimir, 2011. Rudra-Shiva and Silvanus-Faunus: savage
and propitious. J Indo-Eur Stud 39 (1 & 2), 100–115.
Hull, Kathleen L., 2014. Ritual as performance in small-scale societies. World Archaeol.
46 (2), 164–177.
Jasanoff, Jay., 2003. Hittite and Indo-European Verbs. Oxford University Press, Oxford.
Kershaw, Kris., 2000. The One-Eyed God: Odin and the (Indo-) Germanic Männerbünde,
Journal of Indo-European Studies, Monograph No. 36. Institute for the Study of Man
Inc, Washington D.C.
Kohl, Philip L., 2007. The Making of Bronze Age Eurasia. Cambridge University Press,
Cambridge.
Koryakova, Ludmila, Epimakhov, Andrie, 2007. The Urals and Western Siberia in the
Bronze and Iron Ages. Cambridge University Press, Cambridge.
Kosintsev, Pavel, 2016. Archaeozoological Report on the animal bones from the
Krasnosamarskoe settlement. In: Anthony, D.W. (Ed.), A Bronze Age Landscape in the
Russian Steppes: The Samara Valley Project. Cotsen Institute of Archaeology Press,
Monumenta Archaeologica 37, Los Angeles, pp. 385–419.
Kristiansen, Kristian, 2014. The dialectics of gender: Ritualizing gender relations in Late
Bronze Age southern Scandinavia. In: Alexandersson, H., Andreff, A., Bünz, A. (Eds.),
Med Hjärta Och Hjärna: En Vänbok Till Professor Elisabeth Arwill- Nordbladh.
Gothenburg Archaeological Studies 5, Göteborg, pp. 339–353.
Kristiansen, Kristian, 2013. Religion and society in the Bronze Age. In: Christensen, L.B.,
Hammer, O., Warburton, D.A. (Eds.), The Handbook of Religions in Ancient Europe.
Acumen Handbooks, Durham, pp. 77–92.
Kuznetsov, P., Mochalov, O., Anthony, D.W., 2016. The Bronze Age Kurgan Cemetery at
Krasnosamarskoe IV. In: Anthony, D.W. (Ed.), A Bronze Age Landscape in the Russian
Steppes: The Samara Valley Project. Cotsen Institute of Archaeology Press,
Monumenta Archaeologica 37, Los Angeles, pp. 443–470.
Langdon, Susan., 2010. Art and Identity in Dark Age Greece, 1100–700 BC. Cambridge
University Press, Cambridge.
Lankford, George E., 2011. The raptor on the path. In: Lankford, G.E., Reilly IIIF.K.,
Garber, J.F. (Eds.), Visualizing the Sacred: Cosmic Visions, Regionalism, and the Art
of the Mississippian World. University of Texas Press, Austin, pp. 240–250.
Lankford, George E., 2007. The ‘Path of Souls’: Some death imagery in the Southeastern
Ceremonial Complex. In: Reilly IIIF.K., Garber, J.F. (Eds.), Ancient Objects and
Sacred Realms, Interpretations of Mississippian Iconography. University of Texas
Press, Austin, pp. 174–212.
Lebedeva, E.Y., 2005. Archaeobotany and study of the Bronze Age agriculture in Eastern
Europe. Opus: Mezhdistsiplinarnye Issledovaniya v Arkheologii (Moscow) 4, 50–68.
Lincoln, Bruce, 1991a. Homeric lyssa: “wolfish rage”. In: Lincoln, B. (Ed.), Death, War,
and Sacrifice, Studies in Ideology and Practice. University of Chicago Press, Chicago,
pp. 131–137.
Lincoln, Bruce, 1991b. The hellhound. In: Lincoln, B. (Ed.), Death, War, and Sacrifice,
Studies in Ideology and Practice. University of Chicago Press, Chicago, pp. 96–106.
Losey, R.J., Bazaliiskii, V.I., Lok, S.G., Germonpré, M., Leonard, J.A., Allen, A.L.,
Katzenberg, M.A., Sablin, M.V., 2011. Canids as persons: Early Neolithic dog and wolf
burials, Cis-Baikal, Siberia. J. Anthropol. Archaeol. 30 (2011), 174–189.
Mallory, J.P., 2008. Migrations in prehistoric Eurasia: problems in the correlation of ar-
chaeology and language. Aramazd (Armenian Journal of Near Eastern Studies) 3 (2),
7–37.
Mallory, J.P., 2007. Indo-European warfare. In: Pollard, Tony, Banks, Iain (Eds.), War and
Sacrifice: Studies in the Archaeology of Conflict. Brill, Leiden, pp. 77–98.
Mallory, J.P., Adams, D.Q., 2006. The Oxford Introduction to Proto-Indo-European and
the Proto-Indo-European World. Oxford University Press, Oxford.
Mazzorin, J. deG., Minniti, C., 2002. Dog sacrifice in the ancient world: a ritual passage?
In: Dogs and People in Social, Working, Economic or Symbolic Interaction, edited by
Lynn M. Snyder and Elizabeth A. Moore, pp. 62–66. Durham: 9th ICAZ Conference.
McCone, Kim, 1987. Hund, wolf, und Krieger bei den Indogermanischen. In: Meid, W.
(Ed.), Studien Zum Indogermanischen Wortschatz. Innsbruck, pp. 101–154.
Morales Muniz, Arturo, Antipina, Ekaterina, 2003. Srubnaya faunas and beyond: a critical
assessment of the archaeozoological information from the East European steppe. In:
148
View publication stats
Journal of Anthropological Archaeology 48 (2017) 134–148
Levine, M., Renfrew, C., Boyle, K. (Eds.), Prehistoric Steppe Adaptation and the
Horse. MacDonald Institute, Cambridge, pp. 329–351.
Murphy, Eileen, 2016. A Bioarchaeological Study of Prehistoric Populations from the
Volga Region. In: Anthony, D.W. (Ed.), A Bronze Age Landscape in the Russian
Steppes: The Samara Valley Project. Cotsen Institute of Archaeology Press,
Monumenta Archaeologica 37, Los Angeles, pp. 149–215.
Olsen, Sandra L., 2000. The secular and sacred roles of dogs at Botai, north Kazakhstan.
In: Dogs Through Time: An Archaeological Perspective, Crockford, S.J. ed., pp. ∗∗.
BAR International Series 889, Oxford.
Ornan, Tallay, 2004. The Goddess Gula and her dog. Israel Museum Studies in
Archaeology 3, 13–30.
Piggott, Stuart, 1962. Heads and hoofs. Antiquity 36 (142), 110–118.
Pike-Tay, Anne, Anthony, D.W., 2016. Dog days of winter: seasonal activities in a srub-
naya landscape. In: Anthony, D.W. (Ed.), A Bronze Age Landscape in the Russian
Steppes: The Samara Valley Project. Cotsen Institute of Archaeology Press,
Monumenta Archaeologica 37, Los Angeles, pp. 373–384.
Popova, Laura, 2016. Floral Data Analysis: Report on the Pollen and Macrobotanical
Remains from the Krasnosamarskoe Settlement. In: Anthony, D.W. (Ed.), A Bronze
Age Landscape in the Russian Steppes: The Samara Valley Project. Cotsen Institute of
Archaeology Press, Monumenta Archaeologica 37, Los Angeles, pp. 333–349.
Puhvel, Jan 1987. Comparative Mythology. Johns Hopkins, Baltimore.
Raghavan, M., Skoglund, P., Graf, K.E., Metspalu, M., Albrechtsen, A., Moltke, I.,
Rasmussen, S., Stafford Jr, T.W., Orlando, L., Metspalu, E., Karmin, M., Tambets, K.,
Rootsi, S., Mägi, R., Campos, P.F., Balanovska, E., Balanovsky, O., Khusnutdinova, E.,
Litvinov, S., Osipova, L.P., Fedorova, S.A., Voevoda, M.I., DeGiorgio, M., Sicheritz-
Ponten, T., Brunak, S., Demeshchenko, S., Kivisild, T., Villems, R., Nielsen, R.,
Jakobsson, M., Willerslev, Eske, 2014. Upper Palaeolithic Siberian genome reveals
dual ancestry of Native Americans. Nature 505, 87–91.
Rowan, Yorke M., 2012. Beyond belief: the archaeology of religion and ritual.
Archeological Pap. Am. Anthropol. Assoc. 21 (1), 1–10.
Ringe, Don, 2013. The Linguistic Diversity of Aboriginal Europe. In: Chen, Shu-Fen, Slade,
Benjamin (Eds.), Grammatica et Verba, Glamor and Verve. Beech Stave Press, Ann
Arbor, pp. 202–212.
Russell, Nerissa, 2012. Social Zooarchaeology: Humans and Animals in Prehistory.
Cambridge University Press, Cambridge.
Russell, Nerissa, Brown, Audrey, Brown, Emmett, 2016. Human-animal relationships at
Krasnosamarskoe. In: Anthony, D.W. (Ed.), A Bronze Age Landscape in the Russian
Steppes: The Samara Valley Project. Cotsen Institute of Archaeology Press,
Monumenta Archaeologica 37, Los Angeles, pp. 421–441.
Schlerath, Bernfried, 1954. Der Hund bei den Indogermanen. Paideuma 6, 25–40.
Spiedel, Michael P., 2002. Berserks: a history of Indo-European “mad-warriors”. J. World
Hist. 13 (2), 253–290.
Stutley, M., 2003. Shamanism: An Introduction. Routlegde, London.
Tambiah, Stanley J., 1969. Animals are good to think and good to prohibit. Ethnology 8,
423–459.
Vander Linden, Marc, 2015. An impossible dialogue? On the interface between archae-
ology, historical linguistics, and comparative philology. In: Mailhammer, R.,
Venneman, T., Olsen, B. (Eds.), The Linguistic Roots of Europe: Origin and
Development of European Languages. Museum Tusculanum Press, Copenhagen, pp.
1–28.
Van Gennep, Arnold, 1960 [1909]. The Rites of Passage. trans. M. B. Vizedom and G. L.
Caffee. Chicago: University of Chicago Press.
van Hoof, Leon, Dally, Ortwin, Schlöffel, Marlen, 2012. Staying home or staying with
your cattle? Different reactions to environmental changes in the Late Bronze Age of
the lower Don area (Southern Russia). Etopoi Journal for Ancient Studies Special 3,
71–75.
Watkins, Calvert, 1995. How to Kill a Dragon: Aspects of Indo-European Poetics. Oxford
University Press, Oxford.
West, M.L., 2007. Indo-European Poetry and Myth. Oxford University Press, Oxford.
White, David Gordon, 1991. Myths of the Dog-Man. University of Chicago Press, Chicago.
Whitehouse, Harvey, Mazzucato, C., Hodder, I., Atkinson, Q.D., 2014. Modes of re-
ligiosity and the evolution of social complexity at Ҫatalhöyük. In: Hodder, Ian (Ed.),
Religion at Work in a Neolithic Society. Cambridge University Press, Cambridge, pp.
134–155.
Wiseman, T.P., 1995. The god of the Lupercal. J. Roman Stud. 85, 1–22.
Witzel, Michael, 1995. Rgvedic History: poets, chieftans, and polities. In: The Indo-Aryans
of Ancient South Asia: Language, Material Culture and Ethnicity, edited by George
Erdösy, pp. 307–352. Walter de Gruyter, Indian Philology and South Asian Studies v.
1, Berlin.
Zimmer, Stefan, Review of Das, R.P., Meiser, G., 2004. Geregeltes Ungestüm.
Brüderschaften und Jugendbünde bei altindogermanischen Völkern, Bremen. J. Indo-
Eur. Stud. 32, 207–216.