The cultural evolution of shamanism

Abstract

Shamans, including medicine-men, mediums, and the prophets of religious movements, recur across human societies. Shamanism also existed among nearly all documented hunter-gatherers, likely characterized the religious lives of many ancestral humans, and is often proposed by anthropologists to be the “first profession”, representing the first institutionalized division of labor beyond age and sex. This paper proposes a cultural evolutionary theory to explain why shamanism consistently develops, and in particular, (1) why shamanic traditions exhibit recurrent features around the world, (2) why shamanism professionalizes early, often in the absence of other specialization, and (3) how shifting social conditions affect the form or existence of shamanism. According to this theory, shamanism is a set of traditions developed through cultural evolution that adapts to people's intuitions to convince observers that a practitioner can influence otherwise unpredictable, significant events. The shaman does this by ostensibly transforming during initiation and trance, violating folk-intuitions of humanness to assure group-members that he or she can interact with the invisible forces that control uncertain outcomes. Entry requirements for becoming a shaman persist because the practitioner's credibility depends on them “transforming”. This contrasts with dealing with problems that have identifiable solutions (like building a canoe), where credibility hinges on showing results and outsiders can invade the jurisdiction by producing the outcome. Shamanism is an ancient human institution that recurs because of the capacity of cultural evolution to produce practices adapted to innate psychological tendencies.

Full text

The cultural evolution of shamanism
Manvir Singh
Department of Human Evolutionary Biology, Harvard University, Cambridge,
MA 02138
manvirsingh@fas.harvard.edu
http://www.manvir.org
Abstract: Shamans, including medicine men, mediums, and the prophets of religious movements, recur across human societies.
Shamanism also existed among nearly all documented hunter-gatherers, likely characterized the religious lives of many ancestral
humans, and is often proposed by anthropologists to be the “first profession,”representing the first institutionalized division of labor
beyond age and sex. In this article, I propose a cultural evolutionary theory to explain why shamanism consistently develops and, in
particular, (1) why shamanic traditions exhibit recurrent features around the world; (2) why shamanism professionalizes early, often in
the absence of other specialization; and (3) how shifting social conditions affect the form or existence of shamanism. According to this
theory, shamanism is a set of traditions developed through cultural evolution that adapts to people’s intuitions to convince observers
that a practitioner can influence otherwise unpredictable, significant events. The shaman does this by ostensibly transforming during
initiation and trance, violating folk intuitions of humanness to assure group members that he or she can interact with the invisible
forces that control uncertain outcomes. Entry requirements for becoming a shaman persist because the practitioner’s credibility
depends on his or her “transforming.”This contrasts with dealing with problems that have identifiable solutions (such as building a
canoe), in which credibility hinges on showing results and outsiders can invade the jurisdiction by producing the outcome.
Shamanism is an ancient human institution that recurs because of the capacity of cultural evolution to produce practices adapted to
innate psychological tendencies.
Keywords: anthropology; culture; division of labor; evolution; magic; professions; religion; shamanism; trance
A patient, debilitated by the bulbous, blood-bloated ghosts
swarming through her body, sits in an unlit igloo.
1
She is
joined by an angakok –a magician and medicine man –
who crouches in the corner, his body draped in a caribou
skin.
The angakok is unlike himself: One of his helper spirits
has possessed him, and it speaks in a rapid, strange lan-
guage through his mouth. The illness-causing ghosts
respond with fright, abandoning the sick person’s body
before hiding outside of the igloo. The angakok sends his
spirit helpers in pursuit, while members of the onlooking
audience coax the evil ghosts back with half-lies: “Come
in, come in,”they say, “somebody here is waiting for you.”
The evil ghosts return and are slaughtered. The angakok
attacks them with his snow knife, slaying as many as he can.
When he’sfinished, blood covers his hands, unadorned
proof of the killing.
In the days afterward, the patient slowly recovers.
Or, in the days afterward, the patient dies. The shaman
expresses a fatalistic regret: In the end, the ghosts were
too numerous. One man can kill only so many ghosts
when a person has broken so many taboos.
1. Introduction
The Inuit angakok, like the Mentawai sikerei (Loeb 1929),
the Korean mu (Kendall 1985), the Azande boro ngua
(Evans-Pritchard 1937), the !Kung n/um k”ausi (Katz
1982), and the Quaker founder George Fox (Thomas
1971a), is a shaman. I here define shamans as practitioners
who enter trance to provide services. Because my objective
is to identify the social and cognitive foundations of a more
general, cross-cultural suite of practices and beliefs, I
follow authors who prefer a broad definition of shamanism
(e.g., Peters & Price-Williams 1980; Samuel 1990; Wallace
1966; Wright 2009). This usage contrasts with more specific
definitions, such as reserving shamanism for the practices
of Siberian peoples (see discussion in Price 2001)or
using restrictive criteria such as death-and-rebirth initia-
tions, soul journey trances, and animal helper spirits (Win-
kelman 1990). Although many of these traits appeared in
societies outside of Siberia and Central Asia (Eliade
1964), they lack generality and exclude the trance-healing
practices of many societies, including ones commonly
referred to as shamanic (e.g., Kendall 1985; Loeb 1929;
Nadel 1946). The ensuing discussion therefore includes
not only the trancing witch doctors of hunter-gatherer soci-
eties, but also the ecstatic prophets of religious movements,
the mediums of chiefdoms, and the marginal cultists of
contemporary states.
Cross-cultural analyses of shamanism aim to recognize
the particular cultural traits that are universally associated
MANVIR SINGH is a Ph.D. candidate in Human Evolu-
tionary Biology at Harvard University. He studies the
social, psychological, and cultural evolutionary founda-
tions of complex cultural practices that recur across
human societies, including shamanism, music, witch-
craft, and law.
BEHAVIORAL AND BRAIN SCIENCES (2018), Page 1 of 62
doi:10.1017/S0140525X17001893, e66
© Cambridge University Press 2018 0140-525X/18 1
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0140525X17001893
Downloaded from https://www.cambridge.org/core. IP address: 65.112.10.66, on 06 Apr 2018 at 17:52:44, subject to the Cambridge Core terms of use, available at

with this institution. This research has converged on a set of
practices and beliefs that nearly always characterizes sha-
manic traditions (Charles 1953; Eliade 1964; Harner
1990; Peters & Price-Williams 1980; Vitebsky 1995a; Win-
kelman 1986a;1986b;1990):
1. The practitioner enjoys jurisdiction over the treat-
ment and diagnosis of a select set of problems, most fre-
quently serving as a healer and diviner.
2. The practitioner is believed to have special powers
that normal individuals either possess to a less developed
degree or lack completely. These always include some
means of seeing or interacting with invisible forces. But
they also can include flight, invisibility, immunity to fire,
and control of weather and animals.
3. The practitioner engages in a temporary trance state
for at least some of his or her interventions. Definitions
of trance vary considerably in the anthropological litera-
ture, from those that claim universal neurological states
(Harner 1990; Winkelman 2000) to those that emphasize
social conceptions of special powers (Rouget 1985). Never-
theless, most usages concur that trance represents a tempo-
rary state that appears psychologically and behaviorally
distinct from normal human functioning. The behavioral
manifestations of trance differ within and across popula-
tions but include “trembling, shuddering, horripilation,
swooning, falling to the ground, yawning, lethargy, convul-
sions, foaming at the mouth, protruding eyes, large extru-
sions of the tongue, paralysis of a limb, [etc.]”(Rouget
1985, p. 13). The cultural interpretation of this trance
state is also variable and can include spirit possession
(reviewed in Lewis 2003) and soul journeying (reviewed
in Eliade 1964), as well as special sight (e.g., Azande:
Evans-Pritchard 1937), boiling healing energy (e.g.,
!Kung: Katz 1982), or several of these changes simultane-
ously (e.g., Akawaio: Wavell et al. 1988).
4. Entrance into the practitioner class is restricted, typ-
ically by ritualistic initiations (e.g., death and rebirth, ritual
surgery, magical treatment of body parts) or dramatic expe-
riences (e.g., violent illness, epileptic fits, asceticism). Com-
munities often regard individuals with some innate
peculiarity, such as perennial illness (Lebra 1966; Schefold
1988), an extra finger (Bernstein 2008), epilepsy (Nadel
1946), or ambiguous sexual identity (Coleman et al. 1992;
Peletz 2006), as more capable of becoming shamans.
Afinal important characteristic is that shamans represent
a profession, often the only such group in many small-scale
societies (La Barre 1970; Rogers 1982). By profession,I
mean a class of individuals with entry requirements
whose unique expertise or abilities provide them jurisdic-
tion over the treatment or diagnosis of some problems.
For shamans, entry requirements can include prolonged
training from other shamans, special initiations, or sponta-
neous events, such as serious illness (Eliade 1964). This
usage of profession is based on definitions in the sociologi-
cal literature, such as those by Abbott (1988, p. 8) and Mac-
Donald (1995, p. 1), who respectively described professions
as “exclusive occupational groups applying somewhat
abstract knowledge to particular cases”and “occupations
based on advanced, or complex, or esoteric, or arcane
knowledge.”
Shamanism has existed in most documented human soci-
eties, including the majority of hunter-gatherers. Eliade
(1964) famously reviewed ethnographic descriptions of
shamans around the world, documenting similarities in
practice and mythology between Siberian shamanism, on
the one hand, and practitioners in Asia, the Americas,
and Oceania, on the other. Winkelman (1986a;1986b;
1990) coded a modified subsample of the Standard
Cross-Cultural Sample and found trance practitioners in
43 of 47 societies surveyed. A recent review of hunter-gath-
erer religion found shamans in 29 of 33 hunter-gatherer
societies examined (Peoples et al. 2016)
2
; of the remaining
4 societies, the recent ancestors of one (the Siriono) likely
had shamans (Walker et al. 2012; see also sect. 6.2),
whereas an ethnographer noted that members of another
(the Mbuti) visited “the local witch doctor”of nearby
farmers (Putnam 1948, p. 340). Importantly, shamanism
is not restricted to small-scale societies; see, for example,
the benandati cult of medieval Italy (Eliade 1975); Roma-
nian folkloric traditions (Eliade 1975); Neolithic China
(Chang 1999); Tibetan Buddhism (Samuel 1993); contem-
porary Korea (Kendall 1985); founders of religious sects in
20th-century Japan (Blacker 1975); trance channeling in
the contemporary United States (Hughes 1991); Hebrew
prophets (Newsom 1984); early religious leaders of the
Camisards (Knox 1950), Quakers (Thomas 1971a), and
American spiritualists (Albanese 1992); and neo-shamans
in Sweden (Lindquist 1997) and the United States (Braun
2010).
That shamanism appears so regularly in human societies,
especially among hunter-gatherers, suggests that it char-
acterized the lives of many ancestral humans as well.
Inferring the existence of shamanistic practice from archae-
ological findings is notoriously tenuous (Dubois 2009), but
researchers nevertheless have argued prehistoric shaman-
ism from burial sites (Grosman et al. 2008) and art
(Dowson & Porr 2001; Lewis-Williams & Dawson 1988).
The recurrence and similarity of shamanic traditions
highlight several puzzles: (1) Why do these particular
beliefs and practices so frequently develop in concert? (2)
Why do shamans constitute the only professional class in
many societies? (3) Which conditions determine the exis-
tence or collapse of shamanism? These questions have
remained largely absent from the evolutionary and psy-
chological literatures, despite considerable progress in the
study of religion (for exceptions, see Rossano 2007; Winkel-
man 2002; Wright 2009). Meanwhile, the puzzle of
shamanism has attracted the attention of major anthropo-
logical theorists since the discipline’s inception (e.g.,
Evans-Pritchard 1937; Frazer 1922; Lévi-Strauss 1963a;
1963b; Malinowski 1948; Mauss 1902/2001; Tylor 1883),
but as Narby and Francis (2001, p. 8) concluded in their
collection of writings on the topic from the last half-millen-
nium, “even after five hundred years of reports on shaman-
ism, its core remains a mystery.”
This article presents a novel theory of shamanism based
in universal cognitive dispositions and cultural evolutionary
processes. The theory proposes that shamanism is a suite
of practices developed through cultural evolution that
adapts to people’s intuitions to convince observers that a
practitioner can influence otherwise uncontrollable
events. The shaman does this by transforming in the eyes
of the community, during both initiation and trance, assur-
ing group members that he or she can interact with the
invisible forces that control unpredictable, significant
outcomes.
Singh: The cultural evolution of shamanism
2BEHAVIORAL AND BRAIN SCIENCES, 41 (2018)
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0140525X17001893
Downloaded from https://www.cambridge.org/core. IP address: 65.112.10.66, on 06 Apr 2018 at 17:52:44, subject to the Cambridge Core terms of use, available at

The article is structured as follows. I begin by reviewing
alternate theories of shamanism in section 2. In section 3, I
elaborate on the logic of the proposed theory, providing
empirical and theoretical support from psychology and cul-
tural evolution. I use section 4 to explain three central fea-
tures of shamanic traditions: the jurisdiction, trance, and
what I call transformative practices. Section 5 concerns
why shamanism professionalizes in the absence of other
specialization, and section 6 concludes with predictions
for how shifting social conditions should mediate the trans-
formative practices and, in some instances, contribute to
shamanism’s collapse.
2. Alternate theories of shamanism
The most salient features of shamanism to early Western
observers were practitioners’use of sleight of hand and
their ostensible psychological pathology (Krippner 2002;
Narby & Huxley 2001). These led authors to frequently
describe the practice as either a form of charlatanism
(Diderot 1765/2001; Gmelin 1751/2001) or psychopathol-
ogy (Devereux 1961b; Novakovsky 1924; Silverman
1967). Although both perspectives explain some features
of shamanism, ethnographic observations challenge their
simplicity. That shamans seem to believe in their and
others’powers undermines a basic charlatan hypothesis
(Elkin 1977; Evans-Pritchard 1937; Métraux 1944). Mean-
while, shamans in many societies do not suffer from an
abnormal psychology, yet these traditions exhibit the
same patterns in practice and mythology (e.g., Australia:
Elkin 1977; Bhutan: van Ommeren et al. 2004; Akawaio:
Wavell et al. 1988). Furthermore, neither charlatanism
nor psychosis can explain the jurisdiction of shamans or
the use of trance for problem-solving. A comprehensive
theory of shamanism should explain these inconsistencies.
Many authors have proposed that shamanism and related
practices provide benefits to clients or the group (Achter-
berg 1985; Sax 2014), such as through ritually induced
social cohesion (Frecska & Kulcsar 1989), therapeutic
effects mediated by placebo or hypnosis (Kaptchuk 2002,
2011; McClenon 1997), and the psychological comfort
that comes from addressing uncertainty (Achterberg
1985; for a more general discussion of magic, see Malinow-
ski 1948). As with charlatan or psychosis hypotheses,
accounts emphasizing benefits do not explain many cross-
cultural patterns, including the early professionalization of
the practice or the reasons certain practices are considered
effective. Moreover, aside from mixed results from psy-
chology (Calin-Jageman & Caldwell 2014; Damisch et al.
2010), studies and ethnographies finding support for
ritual efficacy tend to rely on reported outcomes by
clients or other community members (e.g., Kleinman &
Sung 1979; Raguram et al. 2002; Sax 2009). Thus, it
appears that most evidence of ritual efficacy concerns the
community perception of outcomes. This does not invali-
date accounts emphasizing benefits, but it suggests that
shamanism may be sustained because of a perception of
results.
Many approaches examine how patterns in human social
and cultural life, such as incest taboos (Fessler & Navarrete
2004; Lieberman et al. 2003) and folk biology (Atran 1998),
reflect universal proclivities resulting from evolved
psychological mechanisms (Sperber 1985;1996a; Sperber
& Hirschfeld 2004). Winkelman’s(2000;2002) neurotheo-
logical theory of shamanism adopts this approach. The
theory attributes the recurrent emergence of shamanism
to (1) psychological effects of universal trance states and
(2) the benefits that the shaman provides. According to
this theory, trance states elicited by dancing, hallucinogens,
and other triggers have an “integrative”effect on cognition,
allowing crosstalk among modules evolved for theory of
mind, social intelligence, and natural history. This integra-
tion enhances practitioners’social abilities, allowing them
to provide useful services to the group and individual
clients. Invoking cultural traditions that leverage aspects
of our core psychology to produce group-level benefits,
the neurotheological theory resembles recent theorizing
on the cultural evolution of prosocial religion (Atran &
Henrich 2010; Norenzayan 2013; Norenzayan et al. 2016;
Purzycki et al. 2016).
The neurotheological theory, although ambitious, suffers
from important shortcomings. Especially problematic is the
central argument that shamanism involves a cross-culturally
consistent trance state that integrates various aspects of
cognition. First, it is unclear what an “integrated”mode
of consciousness is, considering that normal human cogni-
tion involves communication and cooperation among func-
tionally differentiated regions (Hagmann et al. 2008;
Sporns 2011). Second, research on altered states of con-
sciousness suggests that different trance states involve non-
analogous changes in physiology and cognition (Farthing
1992; Vaitl et al. 2005). Altered states induced by sensory
deprivation, for example, disengage the individual from
her surroundings, broaden attention, promote cognitive
flexibility, and stimulate sensory dynamics (richness, vivid-
ness, synesthesia, hallucinations) (Barabasz & Barabasz
1993; Suedfeld 1980; Vaitl et al. 2005). By contrast,
research conducted with pathologically starving patients
suggests that extreme dieting produces the opposite
effects (Ben-Tovim & Walker 1991; Grunwald et al.
2001; Roberts et al. 2007; Vaitl et al. 2005). The contrasting
effects of different trance states challenge the neuro-
theological theory, because, not only do trance states vary
considerably around the world, but also there even exists
within-culture variation or variation among otherwise
similar traditions. For example, the trance state of the Jap-
anese miko, which involves “violent shaking of the clasped
hands”and “stertorous breathing or roaring,”differs in all
of its manifestations from that of the ascetic, who enters