7 BaYaka Pygmy multi-modal and mimetic communication
traditions as suggestive of the social origins of language
Anthropological analyses of Central African hunter-gatherer ethnography reveals a set of
similarities shared by Pygmy groups across the Congo Basin that together with genetic
evidence, suggest that they are remnants of an ancient hunter-gatherer culture. These
hunter-gatherers share a particular mimetic language style, polyphonic music, forest
hunting and gathering, an egalitarian social and economic organization, and genetic
distinctiveness from non-Pygmy local populations.
A focus on the BaYaka Pygmies’ communicative practice – composed of signing,
whistles, animal sounds, plant signs, words from other peoples’ languages, percussion,
singing and dancing, and re-enactments – serves as a reminder of the range of modalities
that humans use to communicate. By placing these gendered practices in the social
context of a hunter-gatherer society we gain insight into the social conditions in which
language evolved. BaYaka have developed this range of communicative practices in order
to address appropriately different audiences: from human groups to animal species. They
explain that in order to be able to speak (pfofa) in the ‘language’ (djoki) or speech
(pfofedi) of others, whether neighbouring farmers, monkeys, or other camp members, one
must mimic their sounds back to them. These ‘languages’ range from animal calls to local
languages, and complex polyphonic musical performances that are intended to
communicate with forest spirits or game animals, and to the forest as a sentient multi-
8.2 The comparative ethnography of hunter-gatherer societies
The ethnography of hunter-gatherer sociality and language-use throws light on some of
the potential conditions facing our hominid ancestors, notably in social and economic
organisation, but also on the situation of being both predator and prey. The responses of
contemporary hunter-gatherers to these conditions are suggestive of how some human
communicative abilities may have emerged and influenced the diversity of languages and
musical styles evident today.
In the late 1970s James Woodburn developed his comparative analysis of the
ethnography of hunter-gatherers to show that they could be divided into ‘immediate-
return’ or ‘delayed-return’ societies (1982). Although taking economic activity as the
starting point, the implications of the difference between immediate- and delayed-return
societies go well beyond economics to determine key aspects of social structure and
In summary, members of immediate-return societies consume most of their food
on the day that they produce it, do not depend on specific others for access to land,
resources or tools, and have an economy based on ‘demand sharing’. They do not invest
in long-term production strategies. By contrast, people in delayed-return societies invest
labour over long periods before a yield is obtained. Typical examples include farming,
herding, or capitalist systems, but also certain hunter-gatherer societies that invest labour
over time or store yields (such as the Kwakiutl and Inuit, and most Amazonian farmer-
foragers). The requirement to manage labour during the period in which the yield is being
produced result in relations of dependence and authority developing between people to
assure that labour is put in at the right times and that those who contribute are
recompensed when the yield is obtained, and so willingly provide their labour again.
Control over the distribution of vital resources promotes political inequality and
hierarchy through the emergence of elites. Whereas delayed-return societies are by
necessity hierarchically organised with inequalities between peers, seniors and juniors
and gender groups, immediate-return societies are politically and economically
egalitarian. While both delayed and immediate return societies exist among hunter-
gatherers, only delayed return societies exist among non-hunter-gatherers.
In immediate–return systems relations are economically egalitarian through
procedures that impose sharing on anyone with more than they can immediately
consume, and so prevent saving and accumulation. A range of mechanisms, such as
demand sharing, gambling, ritual or gifting, ensure that valued goods circulate without
making people dependent on one another. People are systematically disengaged from
property and therefore from the potential for property to be used to create dependency.
Each member of such a society has direct individual access to the resources on which
they depend for survival, to the means of coercion, and to freely move where they want.
Such societies are politically egalitarian, since no one can coerce others to do their will.
People who brag or try to assert their wishes or views on others are mercilessly teased,
avoided and if they persist even exiled. Such societies are indeed rare today, but include
some Pygmy groups in Central Africa (Aka, Baka, Bayaka, Biaka, Efe, Mbendjele,
Mbuti), Hadza in Tanzania, some San groups in Namibia and Botswana, several groups in
India such as the Andaman Islanders, Hill Pandaram and Nayaka, and in south-east Asia,
the Agta, Batek, Maniq, Penan and others.
Though numerically insignificant, these societies are hugely significant for
anthropology since their immediate-return orientation represents such a radically different
mode of social organisation to the numerous hierarchically organised delayed-return
systems that currently dominate human societies. The distribution of these immediate
return traits across the world suggests that these are such successful human adaptations to
life as a hunter-gatherer that they are likely to have great antiquity.
In reviewing how his typology had stood up to the evidence from 30 years of new
ethnography Woodburn noted that immediate-return societies have shown remarkable
resilience over time. Despite the combined forces of government sedentarization and
assimilationist policies, agricultural expansion, industrial exploitation and fortress
conservation all putting huge pressures on these societies, they tenaciously cling to their
immediate-return lifestyle. Immediate-return societies have shown themselves to be
stable, enduring and resilient systems, internally coherent and meaningful to those who
live in them, and resistant to change even when under intense pressure, such as
experienced by Rwanda’s Twa Pygmies (Lewis and Knight1995, and Lewis 2000).
While societies organised around delayed-return economies dominate today’s
world, it is likely that during the period in which the language capacity evolved there
would have been a far greater variety of immediate-return societies, and possibly a
dominance of such societies since,
‘immediate-return systems though not simple in form, are intrinsically simpler than
delayed-return systems and it seems plausible to argue that there will have been a time
when all societies had immediate return systems’ (Woodburn 2005: 20).
Given the commonalities between immediate-return hunter-gatherers distributed
across the globe, these modern forms of more ancient structures and practices provide
important clues to guide our understanding of past human societies and key aspects of the
social relations and conditions during language evolution.
Just as archaeologists that seek to re-learn old skills such as flint-knapping (e.g.
Stout and Chaminade 2007) so as to understand the physical constraints and cognitive
needs for performing such tasks skilfully, anthropologists may observe the constraints
and affordances facing people living by hunting and gathering for indications of those
that faced our evolving ancestors. In previous work (Lewis 2009) on language origins I
focussed on some of the constraints and practical issues posed if spear hunting large,
clever, and dangerous animals such as elephants or herd animals. The solutions that
contemporary spear-hunters have developed to safely kill such dangerous animals, and
their implications on other areas of social life, provide us with insight into the key
constraints and possible solutions that could have been employed by ancestral
populations. In this chapter I will explore these insights to consider whether language has
its origins in certain key contexts of immediate-return hunter-gatherer social life.
8.3 African Hunter-Gatherers and the Pygmies of Central Africa
Africa has the greatest range and variety of hunter-gatherers in the world. Within the
borders of almost every sub-Saharan state live hunter-gatherers and former hunter-
gatherers ranging from extremely egalitarian, immediate-return groups such as the Hadza
of Tanzania to more hierarchically organised delayed-return groups such as the Okiek of
Kenya. The greatest number of contemporary hunter-gatherers in the world live in the
forests of the Congo Basin and estimates of their overall numbers range from 100,000 to
500,000. While there is great diversity among the many Pygmies groups of the Congo
Basin there are also some remarkable similarities.
The BaYaka groups I will be focussing on here occupy forest west of the Ubangi
River, in Central African Republic (CAR), Congo-Brazzaville (Congo), Cameroon and
Gabon. They are made up of Mbendjele (15-20,000), Baka (45-60,000), Aka (15-20,000)
and several smaller groups such as the Mikaya, Luma, Kola, Gyeli, Bongo and others
(maybe around 10-15,000). Many still largely depend on hunting and gathering in an
immediate-return society, though others, such as the Bongo, Kola, Gyeli, Luma and
increasingly Baka too, are engaged in increasingly diversified economies. The term
‘BaYaka’ is contracted to different extents and used by Aka, Baka, Luma, Mbendjele and
Mikaya, typically as bayaka, baaka, or baka . I shall use BaYaka to encompass all these
western groups, but use their individual ethnonyms when providing specific examples.
In addition to a past or actual close association with a forest hunter-gatherer way
of life, ethnomusicologists and ethnographers of Pygmy groups across the Congo Basin
remark on certain cultural aspects of Pygmy societies that have similar structural impacts
on their social, economic and political organisation which serve to reproduce a
particularly effective forest hunter-gatherer adaptation. In particular, these researchers
have pointed out the remarkable similarities between groups living very far apart in their
unusual non-hierarchic, highly integrated choral yodelling (alternating between chest and
head voice) and polyphonic (multiple overlapping melodies) singing style: Arom 1978,
1981, 1985 (on western Pygmies); Cooke 1980; Demolin 1993 (on the eastern Pygmies);
Fürniss 1993, 1999, 2006, 2007 (western Pygmies); Fürniss and Bahuchet 1995 (western
Pygmies); Fürniss and Olivier 1997 (on the differences between Bushman and Pygmy
polyphony); Grauer 2009 (on the similarities between Bushman and Pygmy polyphony);
Kazadi 1981 (on similarities across Tua or Twa Pygmy groups in DRC); Lomax 1962 (on
an overall Pygmy-Bushman musical style); Merriman 1980 (on similarities in DRC);
Oloa Biloa 2011 (on polyphony among Bagyeli Pygmies in Cameroon); Rouget 2004 (on
the Pygmy musical style); Hewlett 1996 on ritual similarities, and others. Becker et al
recently formulated it thus: ‘In Central Africa, cultural criteria, such as the way of life,
Since the Baka, and probably the Mikaya, are Oubangian language group speakers, whereas the Aka and
Mbendjele are Bantu language group speakers, I write the ethnonym as BaYaka to emphasise this dual
identity with forest, language, music, or social interactions, are often used to distinguish
Pygmies from non-Pygmies’ (2010: 19).
Bahuchet tabulates his observations of cultural similarities and differences
between Kola, Bongo, Baka, Aka, Twa, Asua, Mbuti and Efe Pygmies stretching from
west to east across the Congo Basin (Table 5.1, 1993: 109). Across the region yodel and
polyphony together are consistently associated with forest mobility, camps made of leaf
and liana huts, woven-handled axes, and an egalitarian political and economic social
order. Before exploring the causal links behind these associations more explicitly, it is
useful to note that the greater the degree of acculturation to farmer and village lifestyles
the less frequent is yodelled polyphonic music. Those groups Bahuchet identifies as no
longer singing polyphonies (Kola and Bongo) are those that Verdu et al 2009 show to be
the most influenced by outsiders’ genes (Verdu et al 2009 figure S3). Since Bahuchet
compiled this table new research has described polyphony among the Gyeli/Kola (Oloa-
Biloa 2011) and I have heard Bongo in Congo performing polyphony in 2005. Therefore
the common characteristic of all the groups listed is that they all have a recent forest
dependent hunter-gatherer lifestyle and they still participate in polyphonic music, despite
significant differences in the types, amount and varieties performed. These differences
are loosely proportional to the differences in their degree of acculturation to a sedentary
By combining linguistic, material culture and genetic evidence with paleoecology
of the Congo Basin, Bahuchet hypothesises that the region was populated by the
ancestors of today’s Pygmies 40-30,000 years ago during a period of forest expansion
(1993:112). As the forests regressed again between 30-12,000 years ago these groups
became isolated in forest refuges, developing the degree of genetic and linguistic
diversity observed between them today (Cavalli-Sforza 1986). As the forest grew again
from 12,000 years ago to the present, the isolated Pygmy groups moved out from the
forest refuges to populate the regions we find them in today. Verdu et al (2009) further
argue that the arrival of agricultural peoples from 5,000 years ago, but particularly from
2800 years ago, created a second period of isolation between Pygmy groups that led to
the genetic diversity of Western Pygmy groups in Cameroon and Gabon.
Those still engaged in forest-oriented lifestyles and dependant on forest resources
share similar cultural traits, whose traces can still be found in partial ways among those
‘Pygmy’ groups no longer living in the forest, or who have become dependent on
agriculture. The different Pygmy groups have been isolated from one another for long
enough to have developed different languages, genes, technologies and techniques for
exploiting forest resources. But there are underlying structural similarities in music, a
predatory and incorporative mimetic language style, ritual styles, forest hunter-gatherer
lifestyle, a gendered division of labour based on the symbolism of blood (Ichikawa 1987;
Lewis 2008), economies based on demand-sharing, non-hierarchical political
organisation and their widely acknowledged status as the ‘first people’ of the region.
These similarities suggest that their shared traits are likely to be remnants of a more
ancient Pygmy culture dating back to an ancestral population whose diversification I
Verdu et al (2009) suggest that the population ancestral to both Pygmy groups and
Bantu-speaking agriculturalists separated sometime between 54,000 and 90,000 years
ago, though the date is presumed to fall towards the later period since the Bongo genetic
sample was exceptionally mixed due to more recent interactions. Genetic research by
Chen et al (2000) further shows that today’s Pygmy populations are related to today’s San
Bushman populations and once formed a single ancestral population. Victor Grauer
compares this work to ethnomusical analyses to make the strong claim that
‘Biaka Pygmies … could represent one of the oldest human populations … (and that) the
Kung exhibited a set of related haplotypes that were positioned closest to the root of the
human mtDNA phylogeny, suggesting that they too, represent one of the most ancient
African populations … (therefore) the almost indistinguishable musical practices of the
two groups may well date to at least the time of their divergence from the same
population – a period that could … date to at least 76,000 years ago, but possibly as much
as 102,000 years ago (Chen et al 2000: 1371).’ Grauer (2007:6).
Grauer uses this genetic evidence to support an earlier analysis by Alan Lomax
(1962) based on the metrics he had selected to compare folk music from around the world
(the cantometric data base) to show that San and Pygmy music represent a distinctive and
unique polyphonic musical style in their world sample. By combining Colin Turnbull’s
early recordings and ethnography (e.g. 1966) of the Mbuti with cantometric analysis,
Lomax sought to illuminate the relationship between music, performance and social
‘There is a difference in kind between the main performance structure of Western
European folk song, where a lone voice dominates a group of passive listeners … and the
situation in which every member of a group participates, not only in the rhythm and the
counter-point of a performance, but in recreating the melody, as in the Pygmy hocketing
style ... A comparison of the structure of inter-personal relationships and of role-taking in
the two societies shows the same order of contrast, strongly hinting that musical structure
mirrors social structure or that, perhaps, both structures are a reflection of deeper
patterning motives of which we are only dimly aware.’ Lomax 1962: 435.
Additionally he considered the act of making sound, in particular the way that the
distinctive Pygmy yodelling created by alternating between chest and head voice, is
produced by the voice in its most relaxed state when the chords and resonating chambers
in the chest are at their widest and largest.
‘This extra-ordinary degree of vocal relaxation, which occurs rarely in the world as an
over-all vocal style, seems to be a psycho-physiological set, which symbolizes openness,
nonrepressiveness, and an unconstricted approach to the communication of emotion. …
the Pygmy-Bushman profile represents the most extreme case of total focus on choral
integration in our world sample, and in this sense it is unique among folk cultures. The
vocal empathy of the Pygmies seems to be matched by the cooperative style of their
culture.’ (1962: 437).
This insight may go some way to explaining why the more sedentarised and
agricultural the Pygmy group today, the less they perform polyphony and the less they
use yodeling in singing the polyphony.
The evidence from these studies suggests that yodelled polyphonic singing is
directly proportional to adherence to a particular immediate-return hunter-gatherer
lifestyle, since as this lifestyle is abandoned so too are the songs and forms of musical
participation that support it. Applying this logic in the opposite direction suggests that
those groups still practising polyphonic yodelling and a forest hunter-gatherer lifestyle
represent modern vestiges of an extraordinarily resilient cultural system.
This seems self-evident to my Mbendjele friends in northern Congo-Brazzaville
when, in 2010, they heard recordings of Mbuti music made by Colin Turnbull in the late
1950s over a thousand miles to the east. Almost immediately they exclaimed that ‘They
must be Bayaka to sing like this!’ Bahuchet (1993: 109) and Cavali-Sforza (1986) agree
with them, but point out that the two groups demonstrate such genetic difference that they
have been isolated from one another for a very long time.
In sum, while location, language, ecology, technologies and genes have changed,
musical performance and a forest hunter-gatherer economy have remained remarkably
consistently associated. Even if the specific dates or scenarios suggested by these
different authors are inaccurate there appears to be a convergence of evidence for the
association of a forest hunter-gatherer lifestyle with a particular egalitarian or immediate-
return form of social and economic organisation, a gendered division of labour based on
blood, particularly mimetic language styles and a distinctive musical style. These
elements, whilst too specific to emerge from convergent evolution and with genetic
evidence proving a shared past, appear to be key components of a highly resilient and
effective adaptation to forest hunting and gathering. Using the Mbendjele BaYaka as an
example of these shared traits as they relate to the range of activities BaYaka describe as
communicative, I seek to provide insight into the social origins of language.
8.4 Forest Hunter-Gatherers and Gendered Language Styles
The equatorial forests of Central Africa are home to great floral and faunal diversity.
Among the many mammals are large dangerous ones such as forest elephant, buffalo,
leopard, chimpanzee and gorilla. BaYaka women are especially fearful of being attacked
by these dangerous animals. This fear is culturally elaborated into a complex of
associations and taboos collectively referred to as ekila (Lewis 2008) that organize and
naturalize the gendered division of labour based around keeping different types of blood
apart: that of killing animals from that of human fertility, as exemplified by menstruation.
One result of this is that men and women behave differently in the forest. Women
move in large noisy groups, accompanied by children and often yodel loudly as they walk
to frighten game. The daily communalism of women’s lives results in strong solidarity
between them that they readily use to resist men and assert their influence in conflicts and
decision-making. Women’s speech is more song-like than men’s, and they accompany
each other’s utterances with sung expletives that contribute to increasing the volume and
distinctive melodiousness of their conversations.
Men also accompany speech but in a subdued manner by comparison with
women. When out hunting men explicitly value quietness in speech and movement. If
passing through dense noisy undergrowth men sit down to be quiet and scout with their
ears. They discuss in signs and whispers the sounds around. Listening so intently to the
forest, one learns the subtle particularities distinguishing pig grunts or the direction of
bees’ flight from the forest soundscape, or whether a ‘crack’ is an old branch falling from
a tree or buffalo treading on one nearby. Recognizing these distinctions accurately can
determine whether there is meat for dinner, honey for dessert, or that you must run for
The importance of correctly interpreting sound leads BaYaka to pay careful
attention to the sounds of key events. When people tell stories they draw on a standard
stock of these conventionalized accompanying sounds I refer to as ‘sound
signatures’ (Lewis 2009). When recounting lived experience narrators also pay careful
attention to precise sound mimicry of the event described, sometimes without offering
any additional verbal explanation. Indeed they take pride in perfectly mimicking such key
sounds when recounting events, often dropping lexical descriptions altogether. These
sounds tell the forest educated listener all they need to know as well as reminding or
educating others about what different sounds signify and what one should do in response.
Lewis provides more detail (2009; 2013).
Mixing lexical description with sonic mimesis results in a distinctive style of
story-telling common to all western forest-orientated Pygmies and eastern Ituri Pygmies.
Among BaYaka it is cultivated most elaborately by men in secret locations called njanga
to which only the initiated are allowed. Here, men re-enact great hunting moments,
mimicking the sounds and typical postures and movements of themselves and the prey
animal perfectly. This is an important apprenticeship for young men in disguised
communication, hunting techniques and animal behavior. Group hunting techniques, for
instance, rely on sign language and hunters’ disguising their speech as bird calls to
coordinate and manage the group as they encircle the prey and coordinate the attack.
A striking use of such disguised modes of communication by men is to fake
animal calls to lure in prey. These techniques work so well because people can fake
sounds that the animals concerned do not fake, making them successful time and again.
This is a common hunting technique among other human groups too; European and North
American hunters use a wide range of commercially available lures for different prey –
from ducks and wildfowl to deer and moose; many Amazonian Indians call monkeys out
of high trees. Many Central African hunters, like BaYaka men, fake duiker, monkey, pig,
crocodile or buffalo calls to draw them out of undergrowth or high trees and into view so
as to spear or shoot them. In dense forest this is an important contributor to hunting
success. A strategy related to that of mimicking animals’ ‘speech’ to better hunt them is
also applied to extract desirable goods from their villager neighbours. Like the Mbuti
Pygmies described by Turnbull in ‘Wayward Servants’, BaYaka contrive to extract with
minimal effort and danger what they want from farmers by speaking in the villagers’
tongue, claiming pity, and by playing up to the villagers’ claims to ‘own’ BaYaka (Köhler
and Lewis 2002).
BaYaka use their voices so as to make their speech an open, expansive medium
for communication that imitates any other languages or meaningful sounds and actions
that work to communicate with the significant others with whom the wish to maintain
social relations. These include duikers, pigs, crocodiles, and monkeys, as well as other
BaYaka and villagers from many different language groups. To achieve this range of
communicative possibility BaYaka actively cultivate their skills as mimics. But men and
women do so in different ways.
Women’s chat (besime ya baito) often uses mimicry. A typical situation might be
when a group of women sit-down by the sides of a path to rest from their heavy baskets
to chat. In more focused moments, called moadjo (re-enactment), one or two may rise
and re-enact a recent event of note. Poor sharing or hoarding, or abusive, violent, stupid
or outrageous behaviour inspire comical mimicry of those involved. Never mentioning
names, the audience works it out for themselves. The critical commentary this elicits
shares key values through their negation during the re-enactment. Moadjo often focus on
men’s behavior. Only widows and elder women are tolerated to shame others publicly,
and have power in society as a consequence.
Egalitarian gender relations do not mean that everyone is the same, but
rather that each gender group has strengths or qualities that are different from the other,
but socially they are equally valued. Women’s social value is rooted in their ability to
bear life, in the communalism of their daily activities and the solidarity this entails, and in
their use of mimicry to mock behavior that goes against the norms of society. This
contrasts with men’s social value which derives from taking life and providing meat to
grow and sustain human bodies, from their physical strength and toughness, and from
using mimicry of animal and non-BaYaka sounds and actions to ensure safe access to
dangerous but desired products such as wild animals, forest spirits or farmers’ goods.
Mbendjele and other Pygmy groups’ multi-modal communicative strategies
remind us of the more environmentally embedded context of language use likely to have
dominated in the past. By contrast, most language users today think of languages as
conceptually fixed to a particular human group, speech style, vocabulary and grammar,
and as being political by being selective, exclusive and oppositional, as well as
communicative. By contrast, BaYaka seek to speak as many ‘languages’ (djoki) as they
can. Their speech is incorporative, open, encompassing and inclusive. It is a multi-modal
skillful deployment of a range of capacities inherent to human bodies that serve to
establish relationships with as many creatures as possible.
8.5 Music and Language as Alternative Communication Modes Adapted to
Different Audiences and Purposes
The apparent lack of concern to separate different human and animal spheres of
communication is, as are the large corpus of gano fables, evidence for the implicit way
that BaYaka see themselves not as subjects in a society outside nature, but rather as a
society of nature. Just as a society of people implies communication and transaction
between them, so a society of nature implies communication and transaction between its
members. This is typical of an ‘animist’ cosmology (Bird-David 1999; Ingold 2000),
where sentience can be a property of many natural kinds, from other animals to trees,
rocks and even landscapes such as valleys, mountains or forests. Concerned with
maintaining relationships, animist people ‘communicate’ and transact with natural kinds
in many ways – from hunting, gathering and sharing them, to more abstract interaction
such as Aboriginal painting or Inuit animal carving, masked spirit dancers or shamanic
While all people are exquisitely attuned to their environment BaYaka, like other
animist people, have explicitly developed techniques to communicate with many non-
human sentient beings around them. When you understand the noises being made by
other sentient beings, the forest’s sound can be experienced as an inter-species
conversation. To Mbendjele the forest is ‘talking’ to them all the time: elephant is over
there; monkeys have seen pigs; bees are going home (i.e. go home too); frogs invite you
to drink, etc. Interspecies communication is more common and important than many
realize. So duikers drawn to fruiting trees by colobus calls, eat what the monkeys let fall
from their mouths as they greedily shove in the fruit, whilst knowing that they will be
warned of an approaching leopard by the monkey’s alarm calls.
By sharing such meaningful sounds relationships are efficiently established with a
wide range of sentient beings in the forest from whom Mbendjele want things. To ensure
the sounds are appropriate to the sentient being concerned, and therefore meaningful,
Mbendjele mimic the sounds used by these sentient beings back to them. Mimic duikers,
crocodiles, pigs and monkeys if you want to catch them more easily, or mimic farmer’s
languages and ape their stereotypes of you if you want to get things safely from them.
And so it is with the forest as a whole.
Mbendjele say that the forest likes to hear ‘good’ sounds coming from people;
polyphonic song, story-telling, laughter, happy conversations, and the calls of children
playing. Such sounds please the forest so that it ‘opens the camp for food’ and all the
things people need for a good life are easily available. In the same way that Mbendjele
listen to the forest to know about it, so they say the forest is listening to them in order to
know about them.
Since the forest’s ‘song’ is a polyphony of animal, insect and bird sound-making,
it is unsurprising that people so adept at environmental mimicry will ‘speak’ back to the
forest by singing polyphonically . Such communication with forest has been elaborated
into a wide range of rituals that call forest spirits (who are said to ‘eat’ song) into the
camp using hocketed polyphonic singing. These rituals are generically called ‘spirit
plays’ (mokondi massana) by Mbendjele (Lewis 2002; 2013), and called me by Baka
(Tsuru 1998). Spirit plays are communicative technologies explicitly developed to
‘soften’ (charm) the forest and those that hear it. So Malobe, Bula and Yeli demand
specific game animals, while Ejengi celebrates abundance and so on.
Spirit plays are the major social arena for learning key forest skills, cooperation
and group co-ordination skills that are crucial to the success of hunting and gathering
(Lewis 2002). Musical performances involve a huge range of potential meanings and
functions – from the sound and structure of the music itself to the social and political
relationships established between performers in order to produce it, or the way it signifies
culture-specific concepts or identity, and organizes time. In other work (Lewis 2013) I
analyze this style of singing as a ‘foundational cultural schema’ (Shore 1996). Through
the performance of spirit plays key cultural models that are non-linguistically organized
and cross-cut diverse cultural realms, such as economics, politics, history and cosmology,
are re-experienced and learnt by each generation.
Like Fela Kuti’s sonic mimicry of the sounds of Lagos traffic in his music, environmental mimicry has
often inspired music.
During spirit play performances the whole camp assembles in the central space
sitting closely together, resting limbs on each other and touching. As their bodies
intertwine, so too do their voices singing out different melodic lines that overlay each
other to constitute the polyphonic song. It is easy to lose oneself in this physical and
acoustic mass and experience profound communitas. Singers seek to coordinate
excellently just because it is beautiful, and the more beautiful it becomes the more you
lose your sense of self and enter the isεngo (joy). The arrival of the forest spirits among
the singers symbolizes the achievement of communitas with the forest.
To achieve this Mbendjele explicitly work to establish a certain quality of
relations between participants: no arguing, shouting or chatting, and all should share what
they have by contributing as best they can. In conjunction with a musical education there
is a social and political one. There is no hierarchy during musical performances; although
one may begin a song anyone else can stop it and start a new one; everyone is free to join
whichever part of the polyphony they wish. To contribute appropriately listening is as
important as singing: so that you do not drown out your neighbours, or sing the same
melody as they do. If too many sing in unison people immediately and instinctively
diverge by choosing alternative melodic modules to maintain the polyphony.
Regularly singing like this instils certain ways of coordinating and structuring
group activities that are applied outside spirit play contexts. For instance, the instinctive
way that singers avoid unison has economic implications. In an egalitarian society daily
hunting and gathering activities are intuitively co-ordinated without the need for anyone
to tell others what to do. If too many do the same thing there may be nothing to eat, so
being musically primed to do something different but complementary to others improves
the chances that the camp will eat well. Similarly, knowing a sufficient range of melodic
modules and when to insert them into the song structurally resembles the way
environmental knowledge is employed to identify and extract resources from the forest
efficiently. Musical participation in spirit plays is the main avenue through which BaYaka
learn these unspoken grammars of daily interaction (Lewis 2013). Musical performance
resembles what Nick Enfield calls a ‘diagram’, in the technical sense. Here, relations
between participants in musical performance are iconic for their relations in the social
organization (Enfield 2005; 2009).
When considering the range of communicative objectives that Mbendjele sound-
use achieves Ian Cross’ (2005) suggestion to think of music and language as part of a
human communicative continuum is useful. Mbendjele have adapted them to different
purposes: language to express individual intentions, needs and to organize and negotiate
interpersonal relationships and activities; whereas music structures groups and enables
them to ‘speak’ to other groups as collectives rather than as individuals.
When Mbendjele group together in ritualized ways to sing and dance they speak
as one. If only one spoke for them all, it would imply leadership, if each talked at once
nothing would be understood. But when all sing the message is reinforced and repetition
strengthens the point rather than tiring listeners. Crucially a singing group can say things
that no individual in the group could say with out fearing repercussions. Strong,
provocative, insulting or political statements can be made without giving the intended
recipients any space to respond or interrupt. This enables the full statement to be made
and allows tensions to be expressed and acknowledged even if they cannot be resolved.
Widely distributed spirit plays such as Ejengi and Yel i can be found among
BaYaka Pygmy groups speaking different languages and living in different countries
(Aka - Central African Republic; Baka – Gabon, Congo, Cameroon; Mbendjele, Mikaya
and Luma in Congo) suggesting that they are ancient. Ejengi, in particular, establishes a
special arena in which living people connect with the ancestors by re-enacting mythical
narratives concerning the creation of society (described in chapter 24). In the Mbendjele
case, the initiation ceremony re-forges the ancient pact between men’s and women’s
groups that established society as we find it today (Lewis 2002: 173-195).
The analysis of BaYaka communicative practices shows the importance of
mimetic practices to drive their creative spoken and sung engagement with humans,
animals and the forest. Learning to sing polyphonically and participating appropriately
when it is performed inculcate particular cultural dispositions and patterns of behaviour
central to reproducing BaYaka hunter-gatherer culture and society. The association noted
earlier between forest hunting and gathering and yodelled polyphonic singing explains
why Mbendjele and other BaYaka groups hold up musical form, ritual practice and forest
skill rather than language as the key indicators when judging the extent to which other
people are forest hunter-gatherers ‘like themselves’ (Lewis 2002: 54-70). The wide
distribution across Central Africa of this unusual musical style suggests that it is of some
considerable antiquity, and therefore that the cultural dispositions it primes participants
towards are probably refractions of a much more ancient culture.
Genetic and socio-cultural evidence supports this emic perception and my
interpretation, but in reverse. Comparative work such as that of Bahuchet (1993) and
Verdu et al (2009) cited earlier, demonstrate that the more sedentarised and genetically
mixed the Pygmy population is, the less egalitarian they are, the less they participate in
the yodelled polyphonic singing characteristic of spirit plays, and the less they depend on
forest hunting and gathering.
8.6 Some considerations for the social origins of language
From a BaYaka perspective following in the ancestors’ path produces a distinct socio-
cultural aesthetic that includes particular speech, singing and performance styles, a
particular oral tradition, a taste for forest foods above all other food, and a love for the
cool, shady forest over the hot open spaces of rivers, fields and villages.
Here, language in a formal sense is manifestly not synonymous with culture.
Many BaYaka groups have adopted grammatical structures and extensive vocabulary,
even a new language in the Baka case, from non-BaYaka groups without losing their
distinctive cultural identity. Mbendjele and Baka, for instance, see each other as sharing
the same origins and culture, despite Mbendjele speaking a Bantu language and Baka an
Ubangian one. They contrast their BaYaka lifestyles and values to those of their villager
neighbours, even when they speak the same language as the villagers.
They identify performance styles rather than vocabulary; speech protocols (see
Lewis 2009) such as the mosambo (public speaking) rather than grammatical form; the
habit of dropping consonants and otherwise disguising speech; of perfectly mimicking
animal sounds and people’s languages; of excellence at singing and dancing, and the art
of calling forest spirits into camp; of teasing, joking and clowning. It is not what words
people are singing but the polyphonic yodelling singing style they use, not which dances
they dance or which spirits they call but the ritual structures they follow when doing so,
not the language they speak but how it is spoken. The perception of what it means to be
BaYaka is based on an aesthetic quality in which structure or ‘style’ matters more than
This chapter has described a range of communicative techniques, including
language, used by present-day hunter-gatherers to establish fruitful social relations
between people and with their environment. These techniques are such an effective
adaptation, ensuring the efficiency and safety of production and reproduction, and for
maintaining an egalitarian polity, that they persist with striking similarities among groups
of hunter-gatherers who have been separated for many thousands of years.
A striking characteristic of their communicative practices is the impressive role of
mimesis; ranging from literal to stylized, costly to deceptive, physical to acoustic, aimed
at corporate communication and sung, or aimed at individual interaction and spoken. The
extent and range of these mimetic practices offer important clues to help us understand
some of the likely pathways by which arbitrary sounds became accepted as
conventionalized signifiers in the evolving context of hunter-gatherer societies. In chapter
24, Knight and Lewis explore these clues to propose a general theory of how language
The BaYaka ethnography shows how gendered mimicry can drive both lexicon
and normativity; from men that mimic animals or farmer’s languages, to women who use
mimicry to humble antagonists and enforce social norms. This suggests that the gendered
use of mimicry by early hominids could have first developed as a means to deceive
animals, and only later became a means to communicate between people – this seems to
accord with the two-stage hypotheses presented by Whitehead (14), Power (17) and Watts
(18) and in their respective contributions to this volume. For arbitrary signs to
communicate meaning, people must agree to adopt linguistic conventions and categories,
and play the language game honestly (Knight 2009; Lamm, 22 this volume). Mimicry
BaYaka-style facilitates both these processes.
The central role of mimicry in the evolution of language is further hinted at by
Brent Berlin’s demonstration of the role of onomatopoeia and sound synesthesia
(phonoaesthesia) in determining suitable names for things (especially 2005; 2006). This
work suggests that ‘non-arbitrary sound-symbolic, phono-mimetic reference must have
had enormous adaptive significance for our hominid ancestors… that the intuitively
plausible and metaphorically motivated principles of phonoaesthesia served to drive
lexicon in general’ (2006: 49). Ramachandran and Hubbard show that phonoaesthesia
appears to be based on cross-sensual mimicry. In the context of lexicon this occurs from
the inputs received by the senses to movements of the tongue on the palate (2001:19).
Mimicry pervades the human language faculty.
Just as each sex employs different reproductive and productive strategies, so too
do they differ in their use of similar propensities for mimicry. Based on insight from
Mbendjele, women’s mimicry is aimed outwards to ward off dangerous animals, and
inwards against individuals who don’t respect the moral order. Women’s mimicry
depends on their solidarity for its success: in the case of keeping dangerous animals away
they group together and mimic the forest and each other to produce overlapping sounds
that deceive animals about the size of the group. In the second case, they use mimicry to
collectively shame those who have behaved in socially unacceptable ways and so impose
a normative order on society. The first is an example of what Power (17) and Knight (20)
discuss in this volume as ‘counter-dominance’ – here the singers’ alliance works to resist
domination by a potential predator; the second is an example of ‘reverse dominance’ –
the collective domination of an individual not respecting social norms (Boehm 2001).
Both these uses of mimicry enable high levels of trust to be generated and
maintained between members of the social group. This leads to a context in which it
becomes conceivable for deceptive signals to be shared within the group, so enabling
something extraordinary to happen: for the men returning from the hunt to be able to use
acoustic and gestural animal mimicry to share their experiences with the group – for
instance in describing a daring hunt or explaining an accident to non-participants back at
camp. Thus early lexicon and language-like behaviour initially evolved not for in-group
communication but for deceiving other species. But when this animal and sound mimicry
was redeployed for social reasons within a trusting group, say, in the context of early
story-telling, they cease being hunting technique and become language.
1. Since the Baka, and probably the Mikaya, are Oubangian language group speakers,
whereas the Aka and Mbendjele are Bantu language group speakers, I write the ethnonym
as BaYaka to emphasise this dual classification.
2. Like Fela Kuti’s sonic mimicry of the sounds of Lagos traffic in his music,
environmental mimicry has often inspired music.
Arom, S. 1978. ‘Puisque personne ne sait à l’avance ce que tout autre que lui-même va
chanter dans la seconde qui suit…’. Entretien avec Vincent Dehoux, Musique en jeu n
°32, pp. 67-71.
Arom, S. 1981. ‘Un ethnomusicologue chez les Pygmées’. Entretien avec A. Jaubert, La
Recherche 123, p. 768-773.
Arom, S.1985. Polyphonies et polyrythmies d’Afrique centrale, structure et
méthodologie. Paris: SELAF, 2 vol., 906 p.
Bahuchet, S. 1992. Dans la forêt d’Afrique Centrale: Les Pygmées Aka et Baka. Paris:
Bahuchet, S. 1996. ‘Fragments pour une histoire de la Forêt Africaine et de son
peuplement: les données linguistiques et culturelles.’ In: C.M. Hladik, A. Hladik, H.
Pagezy, O.F. Linares, G.J.A. Koppert et A. Froment (Eds). L’alimentation en forêt
tropicale : interactions bioculturelles et perspectives de développement. Paris: Éditions
UNESCO, pp. 97–119.
Becker, N., P. Verdu, B. Hewlett, and S. Pavard. 2010. ‘Can Life History Trade-Offs
Explain the Evolution of Short Stature in Human Pygmies? A Response to Migliano et al.
2007’. Human Biology, February 2010, 82(1): 17–27.!
Berlin, B. 2005. ‘Just another fish story? Size-symbolic properties of fish names’, in A.
Minelli, G. Ortalli & G. Sanga (eds.), Animal names. Venice: Instituto Veneto di Scienza,
Lettre ed Arti, 9-21.
Berlin, B. 2006. ‘The First Congress of Ethnozoological Nomenclature’, in R. Ellen (ed.),
‘Ethnobiology and the Science of Humankind’, Journal of the Royal Anthropological
Institute, Special Issue No 1, 29-54.
Bird-David, N. 1999. ‘Animism Revisited.’ Current Anthropology 40, Supplement pp.
S67 – S91.
Boehm, C. 2001. Hierarchy in the Forest. The evolution of egalitarian behavior.
Cambridge, MA: Harvard University Press.
Cavalli-Sforza, L. L., ed. 1986. African Pygmies. London: Academic Press.
Chen, Y-S., A. Olckers, T. G. Schurr, A. M. Kogelnik, K. Huoponen, and D. C. Wallace,
2000. ‘Mitochondrial DNA Variation in the South African Kung and Khwe—and their
Genetic Relationships to other African Populations’. American Journal of Human
Cooke, P. 1980, ‘Pygmy Music’, in The New Grove Dictionary of Music and Musicians.
15: p. 482-483.
Cross, I. 2005. ‘Music and meaning, ambiguity and evolution’, in D. Miell, R.!
MacDonald & D. Hargreaves (eds.), Musical Communication. Oxford: OUP, pp. 27-43.
Demolin, D. 1993. ‘Les rêveurs de la forêt. Polyphonies des Pygmées Efe de l’Itouri
(Zaïre)’. Cahiers de Musiques Traditionnelles n°6, ‘Polyphonies’, pp. 139-151
Enfield, N. J. 2005. The body as a cognitive artifact in kinship representations: Hand
gesture diagrams by speakers of Lao. Current Anthropology 46(1): 51-81.
Enfield, N. J. 2009. Everyday ritual in the residential world. In G. Senft, & E. B. Basso
(Eds.), Ritual communication. Oxford: Berg, pp. 51-80.
Fürniss, S. 1993. ‘Rigueur et liberté: la polyphonie vocale des pygmées Aka
(Centrafrique)’, in C. Meyer (ed.), Polyphonies de tradition orale. Histoire et
traditions vivantes. Paris: Créaphis, 101- 131.
Fürniss, S. 1999, ‘La conception de la musique vocale chez les Aka : Terminologie et
combinatoires de paramètres’, Journal des africanistes 69(2), pp. 147-162.
Fürniss, S. 2006, ‘Aka Polyphony: Music, Theory, Back and Forth’, in Michael Tenzer
(ed.), Analytical Studies in World Music (chapter 5), Oxford University Press, pp.
Fürniss, S. 2007. Approche interdisciplinaire des musiques pygmées, HDR, dir. Serge
Bahuchet, Université Paris X – Nanterre.
Fürniss, S. and S. Bahuchet. 1995. ‘Existe-t-il des instruments de musique Pygmées?’ In
Dehoux, V., S. Fürniss, S. Le Bomin, E. Olivier, H. Rivere et F. Voisin (éds), Ndroje
balendro. Musiques, terrains et disciplines. Textes offerts à Simha Arom. Louvain-Paris:
PEETERS (Selaf 359), pp. 67-86.
Fürniss, S., and Emmanuelle O. 1997. ‘Systématique musicale pygmée et bochiman :
deux conceptions africaines du contrepoint’. Musurgia. Analyse et Pratiques Musicales n
Grauer, V. 2007. ‘New perspectives on the Kalahari debate: a tale of two “genomes”’.
Before Farming. 2:4 online edition.
Grauer, V. 2009. “Concept, Style and Structure in the Music of the African Pygmies and
Bushmen: a Study in Cross-Cultural Analysis”. Ethnomusicology 53(3): 396-424.
Hewlett, B. 1996. ‘Cultural Diversity among African Pygmies’. In Cultural Diversity
Among Twentieth-Century Foragers. An African Perspective. Susan Kent (ed.)
Cambridge: Cambridge University Press, pp. 215-244.
Ichikawa, M. 1987. ‘Food Restrictions of the Mbuti Pygmies, Eastern Zaire.’ African
Study Monographs, Supplementary Issue 6: 97-121.
Ingold, T. 2000. ‘Totemism, animism and the depiction of animals’ In ‘The Perception of
the Environment’, London: Routledge, pp. 111 - 131.
Kazadi, N. 1981. ‘Méprises et Admires: L'ambivalance des Relations entre les Bacwa
(Pygmées) et les Bahemba (Bantu)’. Africa 51 (4): 836-847.
Köhler, A., and J. Lewis. 2002. ‘Putting Hunter-Gatherer and Farmer Relations in
Perspective. A Commentary from Central Africa’. In Ethnicity, Hunter-Gatherers, and the
‘Other’: Association or Assimilation in Southern Africa? Edited by Susan Kent.
Washington: Smithsonian Institute, pp. 276-305.
Knight, C., 2009. ‘Language, ochre, and the rule of law’. In Rudolf Botha and Chris
Knight (eds) The Cradle of Language, Volume 2: African Perspectives, Oxford: OUP, pp.
Lewis, J., and J. Knight. 1995. The Twa of Rwanda. Assessment of the Situation of the
Twa and Promotion of Twa Rights in Post-War Rwanda. IWGIA Document 78. London:
World Rainforest Movement/Copenhagen: IWGIA.
Lewis, J. 2000. The Batwa of the Great Lakes Region, Minority Rights Group Report,
London: Minority Rights Group.
Lewis, J. 2002. Forest Hunter-Gatherers and their World: A Study of the Mbendjele Yaka
Pygmies and their Secular and Religious Activities and Representations. PhD Thesis,
University of London.
Lewis, J. 2008. ‘Ekila: Blood, Bodies and Egalitarian Societies’. Journal of the Royal
Anthropological Institute. 14: 297-315.
Lewis, J. 2009. ‘As Well as Words: Congo Pygmy Hunting, Mimicry and Play’. In R.
Botha and C. Knight (eds) The Cradle of Language, Volume 2: African Perspectives,
Oxford: OUP, pp. 232 – 252.
Lewis, J. 2013. A Cross-Cultural Perspective on the Significance of Music and Dance on
Culture and Society, with Insight from BaYaka Pygmies. In Michael Arbib (ed)
Language, Music and the Brain: A mysterious relationship. MIT Press.
Lomax, A. 1962. Song Structure and Social Structure. Ethnology 1: (4): 425-451
Merriam, A. P. (1980), ‘Zaïre. 3. Pygmy Music’, in Stanley Sadie (coord.): The New
Grove Dictionary of Music and Musicians. 20: 623. London: MacMillan.
Mithen, S. 2007. ‘Music and the Origin of Modern Humans’, in P. Mellars, K. Boyle, O.
Bar-Yosef and C. Stringer (eds.), Rethinking the Human Revolution. Oxford: Oxbow
Olivier, E. 1999. ‘Seuls les humains chantent. Ce que nous dissent les Ju’hoan sur leur
pratique musicale’. Journal des Africanistes 69(2): 169-182.
Oloa Biloa, C. 2011. La musique des Pygmées Bagyéli (Sud-ouest Cameroun). Etude des
conséquences du changement de mode de vie sur le patrimoine musical", mémoire de
Master 2, Université Paris Ouest-Nanterre.
Ramachandran, V.S. and E.M. Hubbard. 2001. ‘Synesthesia: A window into perception,
thought and language’, Journal of Consciousness Studies, 8: 3-34.
Rouget, G. 2004. “L’efficacité musicale: musiquer pour survivre. Le cas des Pygmées »,
L’Homme n°171-172, ‘Musique et anthropologie’, pp. 27-52
Shore, B. 1996. Culture in mind: Cognition, culture and the problem of meaning. New
York: Oxford University Press.
Stout, D. and T. Chaminade. 2007. ‘The evolutionary neuroscience of tool making’.
Neuropsychologia. 45: 1091-1100.
Tomasello, M., M. Carpenter, J. Call, T. Behne, & H. Moll. 2005. ‘Understanding and
Sharing Intentions: The Origins of Cultural Cognition’, Behavioral and Brain Sciences,
Tsuru, D. 1998. ‘Diversity of spirit ritual performances among the Baka Pygmies in
south-eastern Cameroon’. African Study Monographs, Supplementary Issue 25: 47-84.
Turnbull, C. 1966. Wayward Servants. The Two Worlds of the African Pygmies. London:
Eyre and Spottiswoode.
Verdu, P., F. Austerlitz, A. Estoup, R. Vitalis, M. Georges, S. Théry, A. Froment, S. Le
Bomin, A. Gessain, J.-M. Hombert, L. Van der Veen, L. Quintana-Murci, S. Bahuchet,
and E. Heyer. 2009. ‘Origins and Genetic Diversity of Pygmy Hunter-Gatherers from
Western Central Africa.’ Current Biology 19: 312–318.
Widdess, R. 2006. ‘Musical structure, performance and meaning: the case of a stick-
dance from Nepal’. Ethnomusicology Forum 15(2): 179-213.
Widdess, R. 2012. ‘Meaning in Music.’ Empirical Musicology Review. Vol. 7, No. 1-2:
Woodburn, J. 1982. ‘Egalitarian Societies.’ Man, the Journal of the Royal
Anthropological Institute, 17, no. 3: 431-51.
Woodburn, J. 2005. ‘Egalitarian Societies Revisited’. In Thomas Widlok and Wolde
Gosse Tadesse (eds) Property and Equality. Volume 1: Ritualisation, Sharing,
Egalitarianism. Oxford and New York: Berghahn, pp. 18-31.