Report

Knowledge-Sharing Networks in Hunter-Gatherers

and the Evolution of Cumulative Culture
Highlights Authors
d BaYaka Pygmies use plants for medicine, foraging, and social Gul Deniz Salali, Nikhil Chaudhary,
beliefs James Thompson, ..., Ruth Mace,
Lucio Vinicius,
d Knowledge of medicinal plant use is shared between spouses Andrea Bamberg Migliano
and family members

d Plant knowledge related to co-foraging and social beliefs is

Correspondence
shared among campmates guldeniz.salali.12@ucl.ac.uk

d Use of some medicinal plants is positively correlated with In Brief

children’s health Salali et al. provide empirical evidence on
how hunter-gatherer social structure
plays a role in the diversification of plant
knowledge and cumulative culture. While
medicinal knowledge, which has positive
effects on offspring health, is shared
within families, knowledge about
cooperative activities and social norms is
shared by campmates.

Salali et al., 2016, Current Biology 26, 2516–2521

September 26, 2016 ª 2016 Elsevier Ltd.
http://dx.doi.org/10.1016/j.cub.2016.07.015
 Current Biology

Report

Knowledge-Sharing Networks in Hunter-Gatherers

and the Evolution of Cumulative Culture
Gul Deniz Salali,1,* Nikhil Chaudhary,1 James Thompson,1 Olwen Megan Grace,2 Xander M. van der Burgt,2 Mark Dyble,1
Abigail E. Page,1 Daniel Smith,1 Jerome Lewis,1 Ruth Mace,1 Lucio Vinicius,1 and Andrea Bamberg Migliano1
1Department of Anthropology, University College London, London WC1H 0BW, UK
2Royal Botanic Gardens, Kew, Surrey TW9 3AB, UK
*Correspondence: guldeniz.salali.12@ucl.ac.uk
http://dx.doi.org/10.1016/j.cub.2016.07.015

SUMMARY tion [15] and rarely investigate how the content and function
of cultural information affects knowledge-sharing mechanisms
Humans possess the unique ability for cumulative [16, 17]. Because human cumulative culture is diversified into
culture [1, 2]. It has been argued that hunter-gath- functional domains [16–19], it may also require corresponding
erer’s complex social structure [3–9] has facilitated differentiation of knowledge-sharing mechanisms and underly-
the evolution of cumulative culture by allowing infor- ing social structure [20]. Here we analyze the reported uses
mation exchange among large pools of individuals of 33 plants among the Mbendjele BaYaka pygmies from the
Republic of Congo. We explored the effects of family and
[10–13]. However, empirical evidence for the interac-
camp ties on the reported co-occurrence of plant use in
tion between social structure and cultural transmis-
dyads, which we define as the ‘‘shared knowledge’’ between
sion is scant [14]. Here we examine the reported two individuals.
co-occurrence of plant uses between individuals in
dyads (which we define as their ‘‘shared knowledge’’ Uses of Plants by BaYaka Pygmies
of plant uses) in BaYaka Pygmies from Congo. We Our interviews showed that some plants were used mostly for
studied reported uses of 33 plants of 219 individuals medicinal purposes, and others for foraging or social beliefs
from four camps. We show that (1) plant uses by and norms. Most reported medicinal uses were for treating
BaYaka fall into three main domains: medicinal, digestive (35%) and respiratory (25%; Table 1) disorders. The
foraging, and social norms/beliefs; (2) most medici- BaYaka use some plants for collecting caterpillars or honey
nal plants have known bioactive properties, and and as a poison for killing monkeys or fish; these were classified
as foraging uses. Others were used to regulate social life
some are positively associated with children’s BMI,
and were classified as social norms and beliefs. For example,
suggesting that their use is adaptive; (3) knowl-
some plants are believed to be selectively poisonous to liars,
edge of medicinal plants is mainly shared between while others are involved in sexual taboos (Supplemental Exper-
spouses and biological and affinal kin; and (4) knowl- imental Procedures, section S1; Table S1).
edge of plant uses associated with foraging and
social norms is shared more widely among camp- Medicinal Properties of Plants
mates, regardless of relatedness, and is important Use of similar medicinal plants across cultures would suggest
for camp-wide activities that require cooperation. that they have adaptive benefits and real medicinal properties
Our results show the interdependence between so- [21–23]. Medicinal plants have already been shown to improve
cial structure and knowledge sharing. We propose health in other traditional populations with limited access to
that long-term pair bonds, affinal kin recognition, modern medicine [24]. Out of 33 plants cited by the BaYaka in
exogamy, and multi-locality create ties between our interviews, we successfully identified 31 species. Of these,
15 are also used by Baka Pygmies from Cameroon and Gabon
unrelated families, facilitating the transmission of
[25, 26]. We found a positive correlation between the number of
medicinal knowledge and its fitness implications.
times each of the 15 plants was reported to be used for medic-
Additionally, multi-family camps with low inter-relat- inal purposes by the BaYaka and Baka (Figure 1, n = 15, b =
edness between camp members provide a frame- 0.81, p < 0.01). Moreover, 26 species in our sample are also
work for the exchange of functional information used as medicine by at least one other Central African Pygmy
related to cooperative activities beyond the family population, including the Mbuti and Efe from Democratic Re-
unit, such as foraging and regulation of social life. public of Congo, the Aka from Central African Republic, and
the Baka from Cameroon and Gabon [25–27]. Eight species
are known to be used as medicine by gorillas [28–30], and six
RESULTS by chimpanzees [29, 31, 32] (Table S1). Finally, 24 plants
(77%) have known bioactive properties (Table S1). These find-
Studies of cultural evolution have mainly focused on mecha- ings indicate that medicinal plant knowledge by the BaYaka is
nisms such as fidelity, combination, innovation, and modifica- likely to be adaptive.

2516 Current Biology 26, 2516–2521, September 26, 2016 ª 2016 Elsevier Ltd.
Table 1. Uses of Plants by Mbendjele BaYaka Pygmies
Percentage in all
answers (219
Category Subcategory individuals 3 33 plants)
Medicinal digestive 16.60
respiratory 11.86
pain and injuries 7.07
infections 5.77
wounds 3.27
genitourinary 1.34
pregnancy 0.75
ill-defined 0.32
skin 0.12
poisonings 0.06
circulatory 0.04 Figure 1. Cross-Population Use of Medicinal Plants
Percentage of Mbendjele BaYaka (n = 219) that used a particular plant as a
subtotal medicinal 47.20
medicine and the number of times the same plant was used as a treatment by
Beliefs social norms concerning 1.65 the Baka Pygmies from Cameroon (n = 37, data from [25]) and Gabon (n = 6,
liars data from [26]). Each dot refers to a plant species (n = 15). The shaded area
social norms concerning 1.29 corresponds to 95% confidence interval. See also Table S1.
sexual taboos
luck in finding a partner 0.35 Maternal Knowledge of Medicinal Plants Affects
luck in hunting 0.32 Children’s Body Mass Index
luck in fishing 0.32 To examine potential health effects of medicinal plant uses, we
better sing 0.06 investigated the 14 most frequently used medicinal species
among BaYaka mothers of children aged 0–5 years. Seven
better fight 0.03
plants were used primarily for treating respiratory diseases,
for rain 0.03
and the other seven for digestive system disorders. We found
better share 0.03 that mothers with higher plant-use scores (calculated as the
better work 0.03 number of plants used for medicinal purposes by each mother
subtotal beliefs 4.08 out of the seven possible plants) for treating respiratory system
Foraging bee plants 0.73 disorders had children with significantly higher body mass index
fish poison 0.62 (BMI) (Table 2). However, there was no effect of plant scores for
caterpillar tree 0.50 digestive disorders on children’s BMI (Table 2). These results
indicate that certain medicinal plant uses may provide fitness
monkey poison 0.48
benefits.
subtotal foraging 2.34
Other uses food 3.45 Medicinal Plant Knowledge Is Shared within Families
mat 1.29 Mixed-effect models revealed that dyads represented by biolog-
food additives 0.35 ical or affinal kin ties had increased odds of reporting the same
pirogue 0.32 medicinal plant use (Figures 2A and S1A). A 0.25 increase in
firewood 0.14 the coefficient of relatedness within a dyad increased the odds
uncategorized 0.12 of reported co-occurrence of medicinal plant use by 22%
(odds ratio [OR] = 1.22, 95% confidence interval [CI] = 1.17,
animal food 0.11
1.27; risk ratio [RR] = 1.19; risk difference [RD] = 3%; Table
axe 0.11
S2). Breaking down the effects of kinship, dyads including
hut 0.10 mother and offspring had an increase of 57% in the odds of
avoid animal attacks 0.07 co-occurrence of medicinal plant use (OR = 1.57, 95% CI =
drum 0.07 1.33, 1.84; RR = 1.46; RD = 6%; Figure S1A; Table S2). The effect
basket 0.03 was much smaller (28%) but still significant for father and
subtotal other uses 6.14 offspring (OR = 1.28, 95% CI = 1.04, 1.56; RR = 1.23, RD =
Not available 0.10 3%). Being siblings increased the odds by 40% (OR = 1.40,
95% CI = 1.18, 1.65; RR = 1.33; RD = 5%).
Plant not used 40.14
Affinal ties were also important in explaining co-occurrence
total 100.00
of medicinal plant uses (Figure 2A; Table S2). The odds of co-oc-
currence of medicinal plant use increased by 61% between
spouses (OR = 1.61, 95% CI = 1.32, 1.96; RR = 1.49; RD =
7%). Even distant affinal kin were more likely to report similar

Current Biology 26, 2516–2521, September 26, 2016 2517

Table 2. Mixed-Effects Linear Regression Models
Respiratory Uses
Model 1-1 Model 1-2 Model 1-3 Model 1-4
Coefficient (SE) p Value Coefficient (SE) p Value Coefficient (SE) p Value Coefficient (SE) p Value
(Intercept) 0.84 (0.67) 0.22 1.24 (0.54) 0.03 0.3 (0.31) 0.35 0.37 (0.32) 0.26
Use score 0.21 (0.1) 0.04 0.2 (0.09) 0.05 0.1 (0.08) 0.21
Age 25–35 0.28 (0.48) 0.56
Age 35–45 0.34 (0.51) 0.51
Age 45–55 0.26 (0.79) 0.74
Forest camp 2 0.61 (0.5) 0.23 0.55 (0.47) 0.25 0.73 (0.49) 0.15
Forest camp 3 1.51 (0.54) 0.01 1.47 (0.54) 0.01 0.88 (0.49) 0.08
Town camp 0.4 (0.4) 0.32 0.44 (0.38) 0.26 0.13 (0.37) 0.73
Sex: male 0.46 (0.3) 0.14
AIC 119.83 117.64 120.36 120.17
N observations 42 42 42 42
N groups 33 33 33 33
Digestive Uses
Model 2-1 Model 2-2 Model 2-3
Coefficient (SE) p Value Coefficient (SE) p Value Coefficient (SE) p Value
(Intercept) 0.35 (0.65) 0.59 0.3 (0.31) 0.35 0.04 (0.34) 0.91
Use score 0.03 (0.13) 0.85
Age 25–35 0.51 (0.5) 0.32
Age 35–45 0.54 (0.54) 0.33
Age 45–55 0.09 (0.89) 0.92
Forest camp 2 0.9 (0.54) 0.11 0.73 (0.49) 0.15 0.7 (0.48) 0.15
Forest camp 3 0.92 (0.52) 0.09 0.88 (0.49) 0.08 0.86 (0.48) 0.08
Town camp 0.15 (0.41) 0.72 0.13 (0.37) 0.73 0.05 (0.36) 0.89
Sex: male 0.36 (0.34) 0.31 0.46 (0.28) 0.15
AIC 125.41 120.36 119.52
N observations 42 42 42
N groups 33 33 33
Models 1-1 to 1-4: mothers’ use score of seven plants for respiratory-system disorders on children’s (aged 0 to 5) z-BMI. Models 2-1 to 2-3: mothers’
use score of seven plants for digestive-system disorders on children’s z-BMI. Control variables: mother’s age group, camp residence, and children’s
sex. The models were fit by maximum likelihood. Models 1-1 and 2-1 were the full models.
For respiratory use score, the optimum model was Model 1-2, which included mother’s use score for respiratory problems and her camp residence.
Dropping the variable Use score from Model 1-2 significantly decreased the model fit (for Models 1-2 and 1-3: P[c2(1) > 4.72] < 0.05). For digestive use
score, the optimum model was Model 2-2, which included mother’s camp residence. Dropping the variable Sex from Model 2-2 did not affect the
model fit (for Models 2-2 and 2-3: P[c2(1) > 2.84] = 0.09). Coefficient is the regression coefficient obtained from the model, and SE is its standard error.

medicinal uses of plants (Figure 2A). The odds of co-occurrence co-occurrence of medicinal plant use compared to female-male
of medicinal plant use increased by 41% (OR = 1.41, 95% CI = dyads, but the effect size (7%) was small (Figure 2A; OR = 1.07,
1.26, 1.58; RR = 1.34; RD = 6%) between an individual and their 95% CI = 1.05, 1.09; RR = 1.06; RD = 1%; Table S2).
spouse’s primary kin and increased by 24% (OR = 1.24, 95%
CI = 1.17, 1.31; RR = 1.20; RD = 3%) between an individual Plant Knowledge Related to Cooperative Foraging and
and their spouse’s distant kin. We also observed spouses col- Social Beliefs Is Shared among Campmates
lecting medicinal plants and preparing medicines together Unlike medicinal plant knowledge, plant uses related to foraging
(Movie S1). and social norms and beliefs co-occurred more frequently
Unlike the large effect of family ties, camp ties (when the two among camp members, regardless of family ties (Figures 2B
individuals in the dyad reside in the same camp) increased and S1B). Kin and non-kin effects on odds of reporting similar
odds of co-occurrence of medicinal plant use by only 6% (Fig- plant uses were similar in the two categories (Tables S3
ure 2A; OR = 1.06, 95% CI = 1.04, 1.08; RR = 1.05; RD = 1%; and S4), which were therefore merged. Being from the same
Table S2). Dyads in which individuals belonged to the same age camp increased the odds of co-occurrence of plant use in
group had increased odds of reporting the same medicinal plant foraging and social norms and beliefs by 84% (Figure 2B;
use by 36% (Figure 2A; OR = 1.36, 95% CI = 1.33, 1.39; RR = 1.30; OR = 1.84, 95% CI = 1.72, 1.97; RR = 1.83; RD = 0.3%; Table
RD = 4%; Table S2). Female-female dyads had increased odds of S3). In contrast, neither relatedness (Figure 2B; OR = 0.91,

2518 Current Biology 26, 2516–2521, September 26, 2016

 A B Figure 2. Odds Ratios for the Predictor
Variables
Odds ratios are calculated based on mixed-effects
logistic regression models (Tables S2 and S3,
full models). Response variable is reported co-
occurrence of plant use for (A) medicinal purposes
or (B) purposes related to foraging and social be-
liefs. The dots show the odds of co-occurrence of
plant use when individuals in a dyad belong to the
same camp; are genetically related (odds ratio
calculated for a 0.25 increase in coefficient of
relatedness); have one of the following affinal kin
ties: spouse, spouse’s primary kin, or spouse’s
distant kin; are females; are males; or belong to the
same age group. Error bars show 95% confidence
intervals. ***p < 0.001, *p < 0.05. See also Fig-
ure S1, Tables S2–S4, and Movies S1 and S2.

(where camps one and two were located)

are known as ‘‘children of the flowers’’
because they are known to be forest ori-
ented and good honey collectors [33]. So-
cial norms and beliefs, on the other hand,
95% CI = 0.78, 1.06; RR = 0.92; RD = 0%; Table S3) nor spousal help to regulate camp-wide processes, such as social conflict
ties (Figure 2B; OR, RR = 0.82; RD = 0.1%; Table S3) had an ef- resolution, punishment of cheaters, and coordination of cooper-
fect on odds. The odds of co-occurrence of plant use decreased ation through rituals (Table 1). Camp dependence on social
by 26% between a person and his or her spouse’s distant kin norms and beliefs regardless of family ties (Figure 2B) may favor
(OR, RR = 0.78, 95% CI = 0.64, 0.96; RD = 0.1%). Similarity in cultural drift in plant knowledge, exemplified by the distinct ritu-
age group (OR, RR = 1.51; RD = 0.02%) and sex (for male- alistic ‘‘forest spirit’’ dances across Pygmy groups [35]. A second
male dyads: OR = 1.13; RR = 1.12, RD = 0.01%; for female- example is that only people from the Ibamba camp are known as
female dyads: OR = 0.91; RR = 0.92; RD = 0%) had significant ‘‘people who can fly,’’ due to their particularly rich rituals [33].
effects, but the effect sizes for sex were small (Figure 2B and Overall, our results suggest that variation in plant knowledge
Table S3). Patterns of co-occurrence of plant use are similar across families and camps cannot be explained purely by
for foraging and social norms and beliefs, as they both refer to ecological variation. If similar plant uses were a result of local
camp-wide activities. As an example, we observed multi-family variation in plant availability, camp co-residence would have
groups fishing with plant poison (Movie S2). an equal effect on the distribution of all types of plant knowledge.
However, residing at the same camp had a very small effect on
DISCUSSION similarities in medicinal plant use. We propose instead that a
multi-layered social structure provides underlying channels for
Our results showed that family ties have a significant effect on cultural transmission and diversification of plant knowledge
variation in medicinal plant uses among BaYaka hunter-gath- among the BaYaka. This is suggested by the correlations we
erers, while camp co-residence has the strongest effect on vari- found between social structure (family ties and camp ties) and
ation in plant knowledge related to foraging and social norms plant uses. Attempts to detect patterns and direction of cultural
and beliefs. We suggest that this pattern is a consequence of transmission by asking people from whom they learned partic-
two unique aspects of human social structure: pair-bonding ular information (the ‘‘retrospective method’’) are known to be
with affinal kin recognition, and co-residence with unrelated indi- problematic, as they are affected by memory biases and social
viduals in camps. Joint production of medicine by parents (Movie norms [19, 36]. For this reason, assessing similarity of cultural
S1) is consistent with the frequent co-occurrence of medicinal knowledge among individuals is seen as a better way of mapping
plant uses between spouses. Information exchange between pathways of cultural transmission [18, 37]. By mapping dyadic
families is also likely to be valuable because mothers with correlations (or co-occurrence) in plant uses between individuals
higher medicinal plant-use scores had healthier children. We onto the underlying social structure, we could reveal the roles of
also observed grandmothers (maternal and paternal) preparing biological kin, marriage, and camp ties on the diversification of
medicines, which creates additional opportunity for transmission plant knowledge.
of medicinal knowledge to grandchildren exposed to treatment. Social interactions create the conditions for cultural transmis-
We also show that co-residence of unrelated families in camps sion through various modes of social learning [38]. Among
is associated with camp-specific plant uses in the domains of the BaYaka, social learning predominantly happens through
foraging and social norms and beliefs. Differences in foraging observation and imitation (a young woman observing her mother
uses may reflect distinct levels of foraging activities in each preparing a medicine), through being a recipient of actions
camp [33, 34]. For example, people from the Minganga region relying on cultural knowledge (a child being treated with a

Current Biology 26, 2516–2521, September 26, 2016 2519

particular medicine by parents), or through sharing experiences did not report any uses), the dyadic response was coded as 0. When multiple
(co-participation in rituals). Active teaching is also present, uses were reported by the same individual, we only included the first use
(which occurred in only 2% of the responses).
although learning through observation, participation, and prac-
tice is more common among African Pygmies [38, 39]. In this
Statistical Analysis
context, it must be noted that social learning and cultural trans- Because we had 33 responses for each dyad, we used mixed-effects logistic
mission are not exclusively human traits. Some African apes also regression to predict the reported co-occurrence of plant use in a dyad. Our
use medicinal plants for similar diseases as humans and may ac- fixed predictors were biological kin ties (measured first as coefficient of relat-
quire plant knowledge through observation and imitation of other edness, and subsequently as presence of a specific biological kin tie, e.g.,
individuals [28, 29], as well as through asocial learning. The fact mother-offspring; we analyzed and described models based on each measure
separately), affinal kin ties, camp ties (residing in the same camp), age group,
that eight plants are medicinally used by gorillas [28–30] and six
and sex. We used dyad id as a random effect. We performed separate ana-
by chimpanzees [29, 31, 32] makes it unlikely that learning hap- lyses for three categories of plant use (medicinal, foraging, and social norms
pens solely through trial and error in those species. However, and beliefs).
their medicinal plant uses are not comparable to the vast diver-
sity of plants used by the BaYaka and other human populations. Maternal Medicinal Plant Use and Child BMI
The close match between hunter-gatherer multi-level social We calculated z scores of BMI (body mass index) using 1-year intervals for
structure and diversification of medicinal plant knowledge indi- children aged 0 to 5 to compare their health status. Plant-use scores were
calculated by summing the number of plants used by a mother out of the seven
cates that the complex structure of pair bonding, affinal kin
most commonly used plants for treating respiratory or digestive system disor-
recognition, and co-residence of multiple nuclear families ders. We used linear mixed-effects models for testing the effect of plant-use
created an environment for cultural transmission, as well as score of each mother on offspring BMI (response variable), controlling for
knowledge specialization and innovation, exclusive to humans. mother’s age, camp residence, id (as there were 33 mothers and 42 children,
In addition, co-residence of multiple families allows for the trans- random effect), and sex of the child (Table 2).
mission and accumulation of plant knowledge related to group-
wide activities such as foraging and rituals, which enhance group SUPPLEMENTAL INFORMATION
coordination. All of these factors may have contributed to the
Supplemental Information includes Supplemental Experimental Procedures,
adaptive differentiation of cultural domains and the diversity of
one figure, four tables, and two movies and can be found with this article online
human cumulative culture.
at http://dx.doi.org/10.1016/j.cub.2016.07.015.

EXPERIMENTAL PROCEDURES
AUTHOR CONTRIBUTIONS

All experiments and procedures were approved by the UCL Ethics Committee
A.B.M. conceived the project; G.D.S designed the research and collected the
(UCL Ethics code 3086/003).
data; G.D.S., A.B.M., R.M., N.C., J.T., M.D., A.E.P., and D.S. helped with pro-
tocol design and data collection; O.M.G. and X.M.vdB. identified the plant spe-
Study Population cies and helped with the plant literature research; G.D.S. analyzed the data
Mbendjele BaYaka hunter-gatherers are a subgroup of the BaYaka Pygmies with several contributions of A.B.M. and L.V.; J.L. assisted in fieldwork;
whose residence spans across the rainforests of the Republic of Congo and
G.D.S., A.B.M., and L.V. wrote the manuscript with the help of all other authors.
Central African Republic. The BaYaka live in multi-family camps consisting
of a number of huts in which nuclear families reside. Social ties among
ACKNOWLEDGMENTS
camp members affect food-sharing patterns [40] and individuals’ fitness
[41, 42].
Many thanks to all the Mbendjele BaYaka participants for their hospitality and
We visited four BaYaka camps in the Republic of Congo: three in the forest
sharing their knowledge of the forest; Bakima Arnaud, Ngulawe, and Fongola
(Longa: n = 59, Masia: n = 22, Ibamba: n = 31), and one in a logging town (Sem-
for showing us trees in the forest; Aude Rey and our Mbendjele translators Nic-
bola: n = 107; Supplemental Experimental Procedures, section S2).
olas, Gifhanou, and Paul for their help in data collection. This project was
funded by the Leverhulme Trust grant RP2011-R-045 to A.B.M. and R.M.
Measuring Plant Knowledge and Use
R.M. also received funding from European Research Council Advanced Grant
Fifteen adult informants (10 men, 5 women) were asked to list the names of
AdG 249347, and G.D.S. from Funds for Women Graduates, Foundation Main
plants they used for any purpose. We then chose a subset of 33 plants that
Grant 156240.
are used by the population and asked another 219 individuals (101 men, 118
women) across four campsites whether they knew each of the 33 species,
Received: February 4, 2016
and if so, whether they used it for any purpose (Supplemental Experimental
Revised: June 29, 2016
Procedures, section S1). Later, we classified plant uses into four categories:
Accepted: July 8, 2016
medicinal, social norms and beliefs, foraging, and other. Each category had
Published: September 8, 2016
sub-categories (Table 1). We used the Economic Botany Data Standard for
sub-categories of medicinal uses [43].
REFERENCES
Dyadic Sample
1. Tomasello, M., Kruger, A.C., and Ratner, H.H. (1993). Cultural learning.
From 219 individuals we obtained 23,871 dyads. Each dyad had responses for
Behav. Brain Sci. 16, 495–552.
uses of 33 plants, resulting in a possible 787,743 data points. In 151,038 data
points (19%), no individual used a given plant, and these points were omitted, 2. Dean, L.G., Vale, G.L., Laland, K.N., Flynn, E., and Kendal, R.L. (2014).
resulting in a sample of 636,705 data points and 23,868 dyads. Human cumulative culture: a comparative perspective. Biol. Rev. Camb.
Philos. Soc. 89, 284–301.
Measuring Co-occurrence of Plant Use, or ‘‘Shared Knowledge’’ , M., Eder, J., Headland, T., Hewlett, B.,
3. Hill, K.R., Walker, R.S., Bozicevic
For each dyad, if individual A and individual B reported the same use for a given Hurtado, A.M., Marlowe, F., Wiessner, P., and Wood, B. (2011). Co-resi-
plant, their dyadic response was coded as 1 (‘‘shared knowledge’’). For all dence patterns in hunter-gatherer societies show unique human social
other cases (when individuals reported different uses, or when one of them structure. Science 331, 1286–1289.

2520 Current Biology 26, 2516–2521, September 26, 2016

 4. Dyble, M., Salali, G.D., Chaudhary, N., Page, A., Smith, D., Thompson, J., indices of child health in the Bolivian Amazon. Proc. Natl. Acad. Sci.
Vinicius, L., Mace, R., and Migliano, A.B. (2015). Human behavior. Sex USA 104, 6134–6139.
equality can explain the unique social structure of hunter-gatherer bands. 25. Betti, J.L. (2004). An ethnobotanical study of medicinal plants among the
Science 348, 796–798. Baka pygmies in the Dja biosphere reserve, Cameroon. Afr. Study
5. Chapais, B. (2013). Monogamy, strongly bonded groups, and the evolu- Monogr. 25, 1–27.
tion of human social structure. Evol. Anthropol. 22, 52–65. 26. Betti, J.L., Yongo, O.D., Mbomio, D.O., Iponga, D.M., and Ngoye, A.
6. Chapais, B. (2008). Primeval Kinship: How Pair-Bonding Gave Birth to (2013). An Ethnobotanical and Floristical Study of Medicinal Plants
Human Society (Cambridge, MA: Harvard University Press). Among the Baka Pygmies in the Periphery of the Ipassa- Biosphere
7. Foley, R., and Gamble, C. (2009). The ecology of social transitions in hu- Reserve, Gabon. European J. Med. Plants 3, 174–205.
man evolution. Philos. Trans. R. Soc. Lond. B Biol. Sci. 364, 3267–3279. 27. AFlora Committee (2013). AFlora: The Database of Plant Utilization in
8. Hamilton, M.J., Milne, B.T., Walker, R.S., Burger, O., and Brown, J.H. Africa. Cent. African Area Stud. Kyoto Univ. Available at: http://aflora.
(2007). The complex structure of hunter-gatherer social networks. Proc. africa.kyoto-u.ac.jp/records/top.
Biol. Sci. 274, 2195–2202. 28. Cousins, D., and Huffman, M.A. (2002). Medicinal Properties in the Diet of
9. Marlowe, F. (2004). Marital residence among foragers. Curr. Anthropol. 45, Gorillas : an Ethno-Pharmacological Evaluation. Afr. Study Monogr. 23,
277–284. 65–89.
10. Hill, K.R., Wood, B.M., Baggio, J., Hurtado, A.M., and Boyd, R.T. (2014). 29. Masi, S., Gustafsson, E., Saint Jalme, M., Narat, V., Todd, A., Bomsel,
Hunter-gatherer inter-band interaction rates: implications for cumulative M.C., and Krief, S. (2012). Unusual feeding behavior in wild great apes, a
culture. PLoS ONE 9, e102806. window to understand origins of self-medication in humans: role of social-
11. Powell, A., Shennan, S., and Thomas, M.G. (2009). Late Pleistocene ity and physiology on learning process. Physiol. Behav. 105, 337–349.
demography and the appearance of modern human behavior. Science 30. Huffman, M.A. (1997). Current evidence for self-medication in primates:
324, 1298–1301. A multidisciplinary perspective. Am. J. Phys. Anthropol. 104, 171–200.
12. Henrich, J. (2004). Demography and cultural evolution: how adaptive cul- 31. Krief, S., Hladik, C.M., and Haxaire, C. (2005). Ethnomedicinal and bioac-
tural processes can produce maladaptive losses: the Tasmanian case. tive properties of plants ingested by wild chimpanzees in Uganda.
Am. Antiq. 69, 197–214. J. Ethnopharmacol. 101, 1–15.
13. Derex, M., and Boyd, R. (2015). The foundations of the human cultural 32. Pebsworth, P., Krief, S., and Huffman, M.A. (2006). The role of diet in self-
niche. Nat. Commun. 6, 8398. medication among chimpanzees in the Sonso and Kanyawara
14. Derex, M., and Boyd, R. (2016). Partial connectivity increases cultural Communities, Uganda. In Primates of Western Uganda, N.E. Newton-
accumulation within groups. Proc. Natl. Acad. Sci. USA 113, 2982–2987. Fisher, H. Notman, J.D. Paterson, and V. Reynolds, eds. (New York:
Springer), pp. 105–133.
15. Lewis, H.M., and Laland, K.N. (2012). Transmission fidelity is the key to the
build-up of cumulative culture. Philos. Trans. R. Soc. Lond. B Biol. Sci. 33. Lewis, J. (2002). PhD Thesis. Forest Hunter-Gatherers and Their World: A
367, 2171–2180. Study of the Mbendjele Yaka Pygmies of Congo-Brazzaville and Their
Secular and Religious Activities and Representations.
16. Mesoudi, A., Whiten, A., and Dunbar, R. (2006). A bias for social informa-
tion in human cultural transmission. Br. J. Psychol. 97, 405–423. 34. Salali, G.D., and Migliano, A.B. (2015). Future Discounting in Congo Basin
Hunter-Gatherers Declines with Socio-Economic Transitions. PLoS ONE
17. Reyes-Garcı́a, V., Luz, A.C., Gueze, M., Paneque-Gálvez, J., Macı́a, M.J.,
10, e0137806.
Orta-Martı́nez, M., and Pino, J.; TAPS Bolivian Study Team (2013). Secular
trends on traditional ecological knowledge: An analysis of different domains 35. Lewis, J. (2015). Where goods are free but knowledge costs: the eco-
of knowledge among Tsimane’ men. Learn. Individ. Differ. 27, 206–212. nomics of ritual among forest hunter-gatherers in Western Central
Africa. Hunt. Gatherer Res. 1, 1–27.
18. Reyes-Garcı́a, V., Broesch, J., Calvet-Mir, L., Fuentes-Peláez, N.,
McDade, T.W., Parsa, S., Tanner, S., Huanca, T., Leonard, W.R., and 36. Mesoudi, A. (2011). Cultural Evolution (Chicago: University of Chicago
Martı́nez-Rodrı́guez, M.R. (2009). Cultural transmission of ethnobotanical Press).
knowledge and skills: an empirical analysis from an Amerindian society. 37. Aunger, R. (2000). The Life History of Culture Learning in a Face-to-Face
Evol. Hum. Behav. 30, 274–285. Society. Ethos 28, 445–481.
19. Henrich, J., and Broesch, J. (2011). On the nature of cultural transmission 38. Hewlett, B.S., Fouts, H.N., Boyette, A.H., and Hewlett, B.L. (2011). Social
networks: evidence from Fijian villages for adaptive learning biases. learning among Congo Basin hunter-gatherers. Philos. Trans. R. Soc.
Philos. Trans. R. Soc. Lond. B Biol. Sci. 366, 1139–1148. Lond. B Biol. Sci. 366, 1168–1178.
20. Dı́az-Reviriego, I., González-Segura, L., Fernández-Llamazares, Á., 39. Hewlett, B.S., and Roulette, C.J. (2016). Teaching in hunter-gatherer in-
Howard, P.L., Molina, J.L., and Reyes-Garcı́a, V. (2016). Social organiza- fancy. R Soc Open Sci 3, 150403.
tion influences the exchange and species richness of medicinal plants in 40. Dyble, M., Thompson, J., Smith, D., Salali, G.D., Chaudhary, N., Page,
Amazonian homegardens. Ecol. Soc. 21, 1. A.E., Vinicuis, L., Mace, R., and Migliano, A.B. (2016). Networks of Food
21. Henrich, J., and Henrich, N. (2010). The evolution of cultural adaptations: Sharing Reveal the Functional Significance of Multilevel Sociality in Two
Fijian food taboos protect against dangerous marine toxins. Proc. Biol. Hunter-Gatherer Groups. Curr. Biol. 26, 2017–2021.
Sci. 277, 3715–3724. 41. Chaudhary, N., Salali, G.D., Thompson, J., Rey, A., Gerbault, P.,
22. Saslis-Lagoudakis, C.H., Klitgaard, B.B., Forest, F., Francis, L., Stevenson, E.G.J., Dyble, M., E Page, A., Smith, D., Mace, R., et al.
Savolainen, V., Williamson, E.M., and Hawkins, J.A. (2011). The use of (2016). Competition for Cooperation: variability, benefits and heritability
phylogeny to interpret cross-cultural patterns in plant use and guide me- of relational wealth in hunter-gatherers. Sci. Rep. 6, 29120.
dicinal plant discovery: an example from Pterocarpus (Leguminosae). 42. Chaudhary, N., Salali, G.D., Thompson, J., Dyble, M., Page, A., Smith, D.,
PLoS ONE 6, e22275. Mace, R., and Migliano, A.B. (2015). Polygyny without wealth: popularity in
23. Saslis-Lagoudakis, C.H., Williamson, E.M., Savolainen, V., and Hawkins, gift games predicts polygyny in BaYaka Pygmies. R. Soc. Open. Sci. 2,
J.A. (2011). Cross-cultural comparison of three medicinal floras and impli- 150054.
cations for bioprospecting strategies. J. Ethnopharmacol. 135, 476–487. 43. Cook, F.E.M. (1995). Economic Botany Data Collection Standard.
24. McDade, T.W., Reyes-Garcı́a, V., Blackinton, P., Tanner, S., Huanca, T., Prepared for the International Working Group on Taxonomic Databases
and Leonard, W.R. (2007). Ethnobotanical knowledge is associated with for Plant Sciences (TDWG) (Kew: Royal Botanic Gardens, Kew).

Current Biology 26, 2516–2521, September 26, 2016 2521