1

Ethnic Boundaries and Cultural Change in an Amazonian Population

P.I. John Bunce, University of California, Davis
INTRODUCTION
Overview
This 26-month investigation examines the relationship between inter-ethnic interactions,
perceptions, and the cultural content of ethnic groups. Evidence from ethnography, psychological
experiments, and evolutionary models supports the intuition that inter-ethnic interactions among
individuals play an important role in the changing “cultural content” of an ethnic group, i.e., the
norms (beliefs about what is acceptable behavior) and the symbolic markers (dress, language,
etc.) perceived to be shared by co-ethnics. However, the mechanisms underlying this relationship
remain poorly understood. This study asks the question: How do patterns of interaction among
people of different ethnicities relate to the actual and perceived distributions of ethnic-
typical norms of behavior and symbolic markers across an ethnic boundary? This question
is addressed in two ways. First, ethnographic and experimental methods are used to explore
correlations among these variables in the field. Second, agent-based models are developed to
explore causal mechanisms underlying the observed patterns. The study population consists of
indigenous Matsigenka and majority Mestizo ethnic groups in the Manu region of southeastern
Peru. Although limited, Matsigenka-Mestizo interaction is increasing in both scope and
frequency in the region, spurring intense debate within the Matsigenka communities as well as
among regional authorities about how such interaction leads to change in traditional Matsigenka
behavioral norms (e.g., Shepard et al. 2010; Terborgh 1999). This study contributes to the
discussion by producing new insight and clarifying uncertainty about the roles of inter-ethnic
interaction, markers, and out-group perceptions in cultural change across an ethnic boundary.
This proposal is a resubmission of a 2011 proposal with the same name, for which the
NSF review panel recognized “a significant strength”, namely that “the hypotheses are well
developed, with clear links to theory”, the testing of which “is likely to generate interesting
knowledge.” However, the panel expressed concern that: 1) the PI might not have enough
ethnographic experience, despite a 2010 pilot study; 2) it is difficult to identify norms and
markers; 3) it is difficult to learn the Matsigenka language; and 4) an indigenous federation
should be consulted. All of the previous NSF panel’s concerns are addressed in the present
proposal. In 2011, Bunce gained additional experience working as an ethnographer for an
illness-perception study directed by Dr. N. Ross (Anth., Vanderbilt) and Dr. J. Maupin (Anth.,
Arizona State) designing, conducting, and analyzing 90 hour-long ethnographic interviews in
Spanish with Mexican immigrants in Nashville, TN. He then conducted a second five-month
pilot study in the Matsigenka study community of Tayakome, where he identified important
norms and markers (below). Bunce is now conversational in Matsigenka (although he can find
no institution with authority to certify his skills) and has conducted over 100 formal interviews in
the language. Tayakome is affiliated with the Native Federation of the Madre de Dios River and
its Tributaries (FENAMAD). In a 2011 meeting with Bunce, the secretary of FENAMAD
expressed support for this project and an interest in the final results, which Bunce will provide.

Theoretical Background
Definitions and Relationships among Terms
An ethnic group is here defined as a social group to which individuals ascribe
themselves and to which they are ascribed by others, based on a perception of shared ancestry
 2

and, usually, a range of subjectively important shared norms (beliefs about what is appropriate
behavior) and symbolic markers (language, dress, etc.) (adapted from: Barth 1998[1969]; Gil-
White 1999; Weber 1968). A boundary between two ethnicities is usually characterized by a
sharp division of perceived ancestry, norms, and markers between individuals identifying with
different ethnic groups. People who can credibly claim membership in more than one ethnic
group tend to choose to identify with only one at a time (Alba 1990; Nagel 1996; Waters 1990),
or to create a new separate ethnicity (Nagata 1981), thereby preserving boundaries.
Boundaries between ethnicities can be maintained despite the fact that the norms and
markers of ethnic groups may change over time. In particular, norms held by individuals of two
frequently-interacting ethnic groups may evolve to become similar in the contexts of interaction,
as shared norms tend to make coordinated interactions more mutually beneficial (Barth
1998[1969]; McElreath et al. 2003). For example, shared beliefs about what constitutes
“fairness” make exchanges more efficient. Norms are broadly defined as beliefs about what is
appropriate behavior, and encompass the personal, descriptive, and social norms of Bicchieri
(2006). Norms can have important behavioral consequences (Schultz et al. 2007), and much
inter-ethnic behavioral variation in a given context is here assumed to result from the application
of different norms. Such norm differences may often result from inter-ethnic differences in
knowledge (Atran et al. 2002), though they need not. A coordination interaction (“interaction”
below) is an interaction in which all actors receive a greater (though not necessarily equal)
benefit if they act in concordant rather than discordant manners, and, by extension, hold
concordant rather than discordant norms in the domain of interaction (McElreath et al. 2003).
Examples include interactions between employers/employees, buyers/venders, and spouses.
Choosing interaction partners with suitable norms is often difficult, as many norms are
unobservable until they are put into practice (e.g., norms of spousal and parenting behavior:
Nave 2000). One potential mechanism allowing individuals to bias interactions toward others
who share similar, though unobservable, norms is the use of arbitrary markers (e.g., dress,
language, etc.) that covary with, and can thereby signal, the norms of interest (McElreath et al.
2003). For example, in lowland Peru, most Mestizos habitually speak Spanish (a marker) and, as
Christians, are averse to polygynous marriages (a norm). In contrast, most indigenous
Matsigenka habitually speak Matsigenka and are not necessarily averse to polygyny (Rosengren
2004). In a mixed Matsigenka-Mestizo population, high norm-marker covariance would mean
that an unmarried person’s language is a good indicator of her/his as yet unobservable marriage
norm. As it is used here, a marker is simply the manifestation of a norm (usually a descriptive
norm: Bicchieri 2006) that is readily observable without any significant interaction needing to
occur between individuals. Norm-marker covariance refers to covariance between a marker and a
different, less easily observable norm that has important consequences for a particular type of
interaction (e.g., a language marker covarying with a marriage norm).

Evolutionary Models: Inter-Ethnic Interactions and Markers

McElreath et al. (2003) developed an agent-based cultural evolutionary model to
illustrate a mechanism by which groups of people with distinctive norms and markers can evolve
from an initial population where such group distinctions do not exist. Under the assumptions,
among others, that individuals can modify their markers and that people tend to bias coordination
interactions toward others with similar markers, the model demonstrates that such group
structure can develop provided there is some mixing of interaction partners (termed
 3

“migration”). The basic predictions of this model have been upheld in controlled experimental
games with human participants (Efferson et al. 2008).
To better approximate the study population in the current investigation, Bunce conducted
simulations of the McElreath et al. (2003) model using initial conditions modified to represent
interacting minority and majority groups with initially distinct distributions of norms and
markers. Under the assumptions of the original model, these simulations revealed that when
interactions across the inter-group boundary are present but low, distinct norm distributions can
be maintained in both minority and majority groups, and norm-marker covariance is high among
individuals engaging in inter-group interactions. Both results are due to frequency-dependent
selection. However, with even moderate increases in the level of inter-group interaction,
distinctive norms and markers of the minority group are quickly lost from the population.
This model generates two important predictions in need of testing in the field: 1) change
in the norm and marker distributions of an ethnic group can be mediated through peaceful inter-
ethnic coordination interactions (also predicted by Barth 1998[1969]); and 2) ethnic markers play
an important role in structuring coordination interactions, particularly for those individuals at an
ethnic boundary (i.e., people with high levels of exposure to both co-ethnics and foreigners) .

Social Psychology: Importance of Out-Group Perceptions

When people are divided into social groups, they tend to perceive individuals in out-
groups as: 1) more similar to each other than are members of the in-group (the out-group
homogeneity effect: Ostrom & Sedikides 1992); 2) more different from members of the in-group
than individuals within the in-group are from each other (social accentuation: Tajfel 1982); and
3) inferior to and less worthy of resources than members of the in-group (inter-group bias and
discrimination: Tajfel 1982). Thus, where an ethnic boundary exists, one might predict that
individuals in each ethnic group would perceive the norms and markers of the out-group as
homogenously distributed, different from, and inferior to those of the in-group. Even if not
objectively accurate, these perceptions may influence people’s inter-ethnic interaction behavior.
However, several other factors also influence out-group perceptions. Wilder (1978)
demonstrated that discrimination against the out-group decreases when out-group members are
perceived as individuals rather than as members of a group (“individuation”), and Tajfel (1982)
suggested that such individuation might be achieved through increased personal inter-group
interactions. The nature of inter-group interactions may also influence perception. Riketta and
Sacramento (2008) show that people more readily project characteristics of themselves onto
members of the out-group when the out-group is perceived as cooperative rather than
competitive. This is true even when the projected characteristics are unrelated to the domain of
cooperation or competition. Thus, where an ethnic boundary exists, one might predict that as
inter-ethnic coordination interactions (a form of cooperation) increase, people perceive out-group
individuals as more similar to themselves, even in domains where no inter-ethnic interaction has
yet taken place. Such perceptions may lead individuals to expand inter-ethnic interaction into
these new domains, thereby exposing new norms to the possibility of cultural change.
The relative status of social groups may also affect perceptions of the out-group. Brauer
(2001), interpreting the results of several studies, proposed that, all else being equal, high status
groups may be perceived by both in-group and out-group members as more heterogeneous than
low status groups. This may be because everyone pays more attention to (i.e., more readily
individuates) successful high status people compared to low status people (e.g., indirect bias:
 4

Boyd & Richerson 1985). Status may be determined in the context of coordination interactions
with unequal payoff structures, such that the individual perceived to be receiving the most
benefit from an interaction is the lower-status party. Power (resulting in high status) may rest
with the individual receiving the least benefit, as refusing the interaction would cost her/him less
than it would cost the lower-status individual. For instance, from his experiences in 2011, Bunce
learned that beer bought in Mestizo stores is a rare and valued commodity for Matsigenka, who
can get it nowhere else. However, for a Mestizo, selling beer to a particular Matsigenka (versus
selling it to anyone else) may yield only a small benefit (in terms of time), especially if there are
many other customers waiting to buy beer. Refusing to sell beer to a Matsigenka would cost the
Mestizo little, but would cost the Matsigenka a valued commodity. These unequal payoffs result
in greater power for the Mestizo in this interaction, which, in general, may translate into higher
Mestizo status. Thus, both Matsigenka and Mestizos may more readily individuate Mestizos,
resulting in less discrimination against Mestizos by Matsigenka, but not vice versa. This may, in
turn, affect future inter-ethnic interaction, and thus the degree to which cultural change is likely.

Emic Perspectives of Boundaries and Norms

Anthropological and sociological work on ethnicity emphasizes the importance of emic
perspectives of ethnic boundaries. All individuals have multiple (often nested) social identities
(e.g., Amazonian Indian, Peruvian, Matsigenka, Manu resident) that can be invoked in different
contexts for different purposes (de la Cadena 1992; Marggraff 2004; Nagata 1981; Tilly 2005;
Wimmer 2008). For any two domains of coordination interaction (e.g., marriage versus trade),
the relevant in-group versus out-group distinction, and the norms and markers of importance,
may be different. Thus, a researcher must take care that the ethnic boundary of interest is in fact
the social boundary perceived to be most relevant to the subjects in the context under study
(Degregori 1998; García 2000), and that the subjects and the researcher perceive the boundary in
the same way. This is particularly important for indigenous perceptions of indigenous-Mestizo
and other ethnic boundaries (e.g., de la Cadena 1992; 2005; Gow 1993; Rosengren 2003).
The effect of inter-ethnic interaction on cultural (norm and marker) change may depend
strongly on the meanings people attribute to their own norms and markers. An individual may be
much less willing to change his/her norm or marker to match that of a frequent interaction
partner, despite substantial benefits for doing so, if that particular norm or marker is well
articulated into his/her worldview (e.g., Varese 2006). For instance, bracelets made of lizard
(manke) skin were identified for Bunce by several Matsigenka interviewees as a Matsigenka
ethnic marker. In 2011, Bunce was told by an elder that the bracelets also serve to ward off snake
bites, as manke is the friend of snake (maranke). People who believe in the bracelet’s protective
power may be unwilling to remove it, even in situations where it would jeopardize a potentially
beneficial interaction, e.g., a potential Mestizo employer who perceives the bracelet as a marker
covarying with a norm of “laziness” (a common Mestizo criticism of the Matsigenka).

Ethnographic Complexities: Inter-Ethnic Interaction and Cultural Change

In general, the ethnographic record supports the hypothesis that, where ethnic boundaries
exist, inter-ethnic interactions can have an important influence on cultural change within ethnic
groups. However, as shown below, this influence is often complex and unpredictable.
Gans (1979) argued that, for descendants of European-American immigrants to the U.S.,
behavioral (i.e., norm) differences among ethnic groups tend to be attenuated as interaction
 5

across ethnic boundaries increases each generation. However, the unique identity of these
minority ethnic groups can be maintained by emphasizing a few minimally-costly symbolic
markers perceived as ethnic-typical (e.g., Irish-Americans wearing green on St. Patrick’s Day).
In support of this theory of “symbolic ethnicity”, Waters (1990) and Alba (1990) present
evidence that, in contrast to the extensive ethnic community-oriented norms of European
immigrants to the U.S., their European-American descendants tend to express more personal and
symbolic ethnic commitments which have little impact on everyday behavior. Thus, increasing
inter-ethnic interaction has, for many European-Americans, resulted in the loss of distinctive
norms governing important interactions, and the decoupling of ethnic markers from those norms.
The effect of inter-ethnic contact on the norms and markers of ethnic groups is not
always unidirectional or predictable, however. Nagel (1996) argues that attempts by the U.S.
government to eliminate distinctive Native American cultural behavior by forcing their
assimilation into the “mainstream” of U.S. society, actually laid the groundwork for the Native
American ethnic and cultural resurgence of the 1960s and 1970s. This movement consisted of a
reemphasis, rediscovery, and, in some cases, reinvention of the distinctive symbolic markers and
behavioral norms characterizing tribal (ethnic) groups. Similarly, when inter-ethnic interaction is
present but limited to certain domains, changes in norm and marker distributions can be difficult
to predict. For instance, Huntington and Hostetler (2002) show how Hutterite ethno-religious
groups have been remarkably successful at preserving their distinctive set of norms for over 400
years by restricting inter-ethnic interactions in some domains (e.g., friendships) while
encouraging them in others (community economic exchange). Thus, inter-ethnic contact need not
relate in a straightforward way to changes in the norms and markers of ethnic groups.
For indigenous Amazonian groups, cultural change is often a strategic choice. For
over four and half centuries, missionaries (Rosengren 2004; Varese 2006), governments (Varese
2006), rubber (Varese 2006) and oil companies (Rosengren 2004; Valdivia 2005), populist
movements (Degregori 1998; Delgado 1996), and development NGOs (Rosengren 2004) have
attempted to force native Amazonians to modify their culture by adopting “progressive” norms
of monotheistic spirituality, hierarchical political organization (Espinosa 2009; Rosengren 2004;
Smith 1996; Urteaga Crovetto 2007), capitalism, sedentary settlement patterns, and intensive
agricultural techniques (Rosengren 2004; Varese 2006) in return for technological, educational,
medical, and economic benefits. Many have strategically adopted majority-culture norms in
order to secure benefits (or simply survive) in a world dominated by such norms. For instance, in
a representative example, Harakmbut indigenous groups in lowland Peru have strategically
adopted destructive gold mining practices and entrepreneurial norms as a way to secure access to
resources on their land before government-backed Mestizo miners do (Urteaga Crovetto 2007).
Recently, an alternative strategy to secure benefits has also become a viable option. By
emphasizing how different they are from the majority culture, indigenous groups can often enlist
the support and protection of governments and cultural and conservation NGOs. However, this
strategy usually entails conforming to majority-culture stereotypes regarding what it means to be
“indigenous” (e.g., Balslev Clausen 2008; Jackson 1989; 1995). For instance, Manu National
Park was created in lowland Peru in 1973 in an area occupied by several hundred Matsigenka.
Some Matsigenka opted to remain in the park, where, in exchange for the right to live in an area
protected from Mestizo colonization, they were forced to readopt or maintain subsistence
practices considered by park authorities to be “traditional” (Marggraff 2004; Shepard et al. 2010;
Terborgh 1999, Bunce, pers. obs.). There is considerable debate and uncertainty among
 6

Matsigenka both inside (Bunce, pers. obs.) and outside the park (Baer 2004; Rosengren 2003;
2004), about how to balance the adoption of new norms with the maintenance of “traditional”
norms, and what benefits or sanctions these strategic choices would elicit from majority-culture
power structures, such as governments and NGOs (e.g., Holt 2005; Valdivia 2005).

Motivation for the Proposed Study

As shown above, the relationship between inter-ethnic interaction and cultural change is
complex. Few field studies, however, have yet attempted to explicitly measure several variables
that are believed, on the theoretical grounds presented above, to play important roles in the
complex relationship between inter-ethnic interaction and cultural change. These variables
include: 1) individual variation in the frequency and nature of inter-ethnic interactions; 2)
covariance of markers and norms important in inter-ethnic interactions; and 3) individual
variation in perceptions of out-group norm and marker distributions. The results of several
pioneering studies of populations with ethnic boundaries suggest that important relationships
exist between cultural norm, knowledge, and behavior distributions and several of these
variables, e.g., market integration (inter-ethnic interaction) (Godoy et al. 2009; Henrich et al.
2010), cross-ethnic social networks (inter-ethnic interaction) (Atran et al. 2002), and perceived
out-group norm distributions (Ross et al. 2007). This study, however, will be the first to measure
all three variables in the same population, and can thus evaluate the relative importance of the
roles played by each in cultural change.

OBJECTIVES AND HYPOTHESES

The overall aim of this investigation is to determine the effects of inter-ethnic interactions
on the actual and perceived norm and marker complements of individuals on either side of a
Matsigenka-Mestizo ethnic boundary. Specific objectives, with corresponding hypotheses, are:
Objective 1: Identify and ethnographically describe a set of behavioral norms and ethnic
markers in each ethnic group, and a corresponding set of intra- and inter-ethnic coordination
interactions.
Hypothesis 1A: There exists a set of norms in each ethnic group that govern interactions
between individuals, and, in some contexts, these norms are signaled by markers.
Hypothesis 1B: Norms and markers differ in the degree to which they are integrated
into the worldviews of individuals.
Hypothesis 1C: The Matsigenka-Mestizo ethnic boundary is perceived by individuals of
both groups to be a salient boundary structuring a set of inter-ethnic interactions.
Objective 2: Measure the distributions of identified norms, markers, and interactions among
individuals of each ethnic group.
Hypothesis 2A: The distributions of norms and markers differ across the ethnic
boundary, i.e., Matsigenka and Mestizo cultural content differs at the population level.
Hypothesis 2B: Within each ethnic group there is inter-individual variation in norm and
marker complements, as well as in the frequency and nature of intra- and inter-ethnic
interactions involving those norms and markers.
Hypothesis 2C: The norms of individuals will match those of their most frequent or
most important interaction partners, regardless of whether those partners are co-ethnics
or belong to the out-group.
 7

Corollary 2C: Individuals perceived to receive the most benefit from a particular
interaction (low-status individuals) will change their norms to match those of
individuals perceived to receive less benefit from the particular interaction (high-
status individuals). When inter-ethnic interactions are frequent, the Matsigenka
(lower-status) tend to adopt Mestizo (higher-status) norms, and not vice versa.
Hypothesis 2D: When exposed to a range of potential interaction partners with
coinciding and conflicting norms, individuals will signal their norm with a marker. This
results in high norm-marker covariance.
Corollary 2D: The highest norm-marker covariance will be found in the subset of
Matsigenka and Mestizos with intermediate levels of inter-ethnic interaction, as
these individuals run the highest risk of choosing an interaction partner with
conflicting norms. The norms of individuals with high levels of inter-ethnic contact
will match those of the out-group (Hypothesis 2C) so the risk of mis-coordination,
and the importance of markers, for such individuals is reduced.
Hypothesis 2E: As the importance of particular norms and markers in the worldviews of
individuals increases, the less likely people are to modify them in response to changing
interaction partners.
Objective 3: Compare the actual and perceived distributions of norms and markers, in relation to
inter-ethnic interactions.
Hypothesis 3A: The greater the frequency or importance of an individual’s inter-ethnic
interactions, the more she/he “individuates” the out-group, and thus the more accurate
(less stereotyped) is her/his perception of the out-group norm and marker distributions
and norm-marker covariance, in domains both related and unrelated to the interactions.
Hypothesis 3B: Members of the lower-status ethnic group (Matsigenka) will, in general,
have more accurate perceptions of norm and marker distributions in the out-group than
will members of the higher-status ethnic group (Mestizos).
Objective 4: Use the patterns encountered in this population to develop general agent-based
evolutionary models illustrating causal mechanisms of cultural change among ethnic groups.

METHODS
Study Population
This study takes place in the indigenous Matsigenka community of Tayakome (October
2011 population: 75 adults in 35 households) located inside Manu National Park (MNP), and the
nearby Mestizo town of Boca Manu (approximately 30 households) located just outside the park
boundary in lowland tropical forests in the Department of Madre de Dios, Peru.
Within MNP, approximately 400 Matsigenka live in settled communities, of which
Tayakome is the oldest and second-largest (Ohl-Schacherer et al. 2007; Shepard et al. 2010). The
Matsigenka and other groups in the region had only sporadic contact (e.g., trade) with outsiders
until the rubber boom of 1895-1917, when many fled into the forest away from rivers to avoid
slave raids. Missionaries of the Summer Institute of Linguistics (SIL) founded Tayakome in the
1960s and attracted dispersed Matsigenka families with the promise of education, healthcare,
technology, and economic opportunities. SIL was forcibly removed when MNP was founded in
1973 (Shepard et al. 2010). The Matsigenka were allowed to continue living in MNP on the
condition that they maintain or readopt “traditional” subsistence practices of swidden
 8

horticulture, gathering, hunting, and fishing (Johnson 2003), which are thought to be less
detrimental to the park’s biodiversity (Levi et al. 2009; Ohl-Schacherer et al. 2007; Ohl et al.
2008; Shepard et al. 2010). Cash-cropping, livestock (other than chickens), firearms, and the
selling of forest products are prohibited (Shepard et al. 2010). Some Tayakome residents
habitually travel outside the community and the park to participate in ecotourism (Ohl-
Schacherer et al. 2008), wage labor, secondary education, and limited purchasing. The vast
majority of inter-ethnic interactions occurs outside of the community, as non-Matsigenka cannot
enter Tayakome, or the park, without a government permit. Most residents speak some Spanish.
Boca Manu is a 2-3 day boat trip from Tayakome and the first community encountered
upon leaving MNP’s one principal entrance/exit. Residents are mostly Spanish-speaking
Mestizos of Andean ancestry, with some indigenous Matsigenka and Yine. The economy of
Boca Manu depends on the regional ecotourism and boat-building industries (Shepard et al.
2010). Nearly all Tayakome residents have visited Boca Manu at least once in their lives, and
many of the younger men have worked there for several months. Boca Manu also contains a
boarding secondary school for Matsigenka, where four Tayakome boys are studying.
Tayakome and Boca Manu constitute an excellent study population for investigating
the relationship between inter-ethnic interactions and changing norm and marker distributions.
Because these communities are small and isolated, with little exposure to mass media (Tayakome
has one sporadically-functioning television in the president’s house), Matsigenka-Mestizo
interactions are, on the whole, relatively limited in scope, easy to monitor, and easy for people to
recall. Additionally, in Tayakome, there is much individual variation in the extent of inter-ethnic
interaction with Mestizos, and similar variation in inter-ethnic interaction experience is expected
in Boca Manu. Such variation is necessary to evaluate Hypotheses 2B-D and 3A.

Pilot Studies
Bunce conducted two pilot studies in Tayakome (Jun-Jul 2010 and Jun-Oct 2011). In
these seven months of fieldwork, he conducted semi-structured personal history interviews in the
Matsigenka language (back-translated for accuracy) with 59 of the 75 resident adults (35,
24), including at least one member from 33 of the 35 households occupied during the study
period. He also conducted 105 structured interviews with adults on focused topics relating to
ethnic identity and specific norms and markers. To examine the distribution of a social norm
specifying a man’s obligations to the community versus his obligations to his family,
interviewees (12, 21) were presented with a vignette in which a man decided to skip a
community work party (“obligatory” community function) so that he could stay at home to
construct arrows (necessary for family food acquisition). Interviewees indicated whether the
hypothetical man’s behavior was acceptable (kameti) or unacceptable (tera kameti). Results
(Table 1) suggest that, of those who believe the man in the vignette is wrong to shirk his
community obligations (Unacceptable), most have lived or worked outside of Tayakome in a
Mestizo town at some point in their lives (Outside). Of the two people who have never lived
among Mestizos but who also disapproved of the hypothetical man’s behavior (Inside,
Unacceptable), one was a woman whose parents, siblings, and husband had all spent
considerable portions of their lives among Mestizos. One explanation (among several) for this
pattern is that increased interaction with Mestizos is associated with a change from Matsigenka
family-centered norms (Johnson 2003) to more community-oriented norms typical of Mestizos.
 9

Table 1. Figure 1.
Acceptable Unacceptable
Outside 10 6
Inside 15 2

To examine the distribution of ideas regarding Matsigenka ethnicity, the same sample
was presented with a series of vignettes about a hypothetical adoption situation (e.g., a
Matsigenka baby adopted by Mestizos, a Mestizo baby adopted by Matsigenka, etc.) where
respondents indicated whether the adoptee, once grown, would be a Matsigenka or not. Using
non-linear PCA methods (Michailidis & de Leeuw 1998), Bunce reduced responses to a score on
a single component axis (x-axis in Figure 1) that could be interpreted as the degree to which a
respondent believed that ethnicity is inherited from biological parents versus acquired (e.g.,
consists of norms that are learned) from adoptive parents. Figure 1 shows considerable variation
in interviewee responses, both for those who have lived among Mestizos (Outside) and for those
who have not (Inside). One interpretation of these data is that most people perceive a salient
ethnic boundary in which at least some fundamental differences between Matsigenka and
Mestizos consist of cultural characteristics (e.g., norms and markers) that can be learned by
children (i.e., all people are not clustered at the “Ethnicity Inherited” pole).
Bunce has identified several Matsigenka norms and ethnic markers through extensive
participant observation in community life. For instance, Bunce participated in familial swidden
horticultural practices, including clearing (5 fields), burning (1 field), and planting (4 fields). In
one family, the wife did not participate in planting manioc and believed it to be her husband’s
job. In three other families, wives planted manioc to an equal or greater extent than their
husbands, and believed it to be the job of both spouses. Thus, variation in norms of spousal
obligation (not explainable by contact experience with Mestizos) is apparent in Tayakome.
These, as well as other norms (e.g., gender-segregated meals) and markers (e.g., women’s shaved
heads) identified in the pilot studies, constitute a considerable advance and a promising point of
departure for the proposed investigation. Additionally, Bunce spent 23 months between 2003 and
2006 conducting his dissertation fieldwork on primate ecology along the Manu River between
Tayakome and Boca Manu. His experience in the region greatly facilitates his current research.
In Tayakome, Bunce is joined by his wife Caissa Revilla-Minaya, a Peruvian and
advanced Cultural Anthropology Ph.D. student at Vanderbilt U, conducting her dissertation
research on Matsigenka environmental perceptions. Revilla-Minaya is applying for independent
funding for her research, and the current project includes no funding for her. During the 2010-
2011 pilot studies, both found that their marital status greatly facilitated ethnographic research in
this family-level society, especially Bunce’s interviews of women. Bunce and Revilla-Minaya
visit households together and conduct simultaneous independent interviews for their respective
studies. They engage in many participant observation opportunities (e.g., social visits, manioc
planting, barbasco fishing) as a married couple, following the example of Matsigenka couples.

Proposed Investigation: Data Collection

Phase 1: Identification of Mestizo Norms and Markers (15 Aug 2012 – 31 Dec 2012)
Bunce will live in Boca Manu for five months. During the first four months he will use
participant observation to identify domains (contexts) of coordination interaction, such as: 1)
employment (e.g., in Mestizo agricultural fields, in Mestizo logging outfits, or as boat crews); 2)
 10

commerce (e.g., among the small general stores in town); 3) education (e.g., in the town’s
boarding secondary school for Matsigenka boys); and 4) healthcare (in the town’s regional health
post), as well as the norms and markers that structure such interactions (Hypothesis 1A).
Participant observation is facilitated by Bunce’s fluency in Spanish (advanced oral and written
fluency certified by exam in the Dept. of Linguistics, Univ. Nacional Mayor de San Marcos,
Lima, Peru 2011), and his informal familiarity with these interaction domains gathered over
several years of work in the region. Bunce will select a purposive sample of four key Mestizo
informants (an older woman and older man, and a younger woman and younger man). An
additional key informant will be sought from among the small number of Matsigenka who live in
Boca Manu. Bunce has interacted frequently with several potential key informants in Boca Manu
since 2003. He will use semi-structured interviews (Bernard 2006) to explore how, and in what
domains, these informants distinguish Matsigenka from Mestizos, i.e., when/if this ethnic
boundary is meaningful for them (emic boundary perspectives - Hypothesis 1C). With these
informants, Bunce will also discuss the reasons behind the norms and markers he observes
during participant observation (Hypothesis 1B), e.g., are observed norms related to Christian or
Matsigenka beliefs? In addition, key informants will be asked to recount their life stories, with
emphasis on their personal history of interaction across the ethnic boundary (e.g., Oakdale 2008).
Particular attention will be paid to cross-cultural competence, where individuals temporarily
change their norms depending on the ethnicity of their interaction partner. All formal interviews
are confidential and will be recorded for transcription. Interviewees (in all project phases) are
read a statement of informed consent and assured of their right to stop an interview at any time.
During the final month of Phase 1, Bunce will identify norms and markers through
structured interviews with a random sample of 30 adults (names drawn at random from a pool
of willing participants). This pool of Boca Manu participants may include some resident
Matsigenka, and their responses will be analyzed separately from those of Mestizo residents. The
interview schedule is based on that of the pilot studies (above), and will result in sampling of
approximately half of the community adult population. Interviewees will free-list behaviors they
think are appropriate, i.e., their personal norms of behavior (Norm Free-list), and the observable
characteristics they would use to distinguish a Matsigenka from a Mestizo (Marker Free-list), in
a set of six domains of coordinated interaction identified during participant observation and prior
interviews. For example, for the marriage domain, interviewees may be asked:
What characterizes a good husband? (Possible Mestizo Norm Free-list responses:
provides for his family, drinks responsibly, etc.). What characterizes a good wife?
(Possible Mestizo Norm Free-list responses: cooks, raises children, etc.). Imagine your
son marries outside of Boca Manu, and then returns with his wife. When you first meet
her, how can you tell that she is a Mestiza and not a Matsigenka? (Possible Marker Free-
list responses: she speaks excellent Spanish, she wears shoes, metallic jewelry, etc.).
Bunce will choose six domains where the key informants perceived the Matsigenka-Mestizo
ethnic boundary to be salient. Three of the six domains will be chosen from among those
identified during pilot studies in Tayakome, where Matsigenka-Mestizo interactions tend to be
rare, e.g., marriage, meal sharing, community participation, etc. The remaining three domains
will be chosen from among those identified in Boca Manu, where inter-ethnic interactions are
more frequent. Correspondence between free-listed and practiced behavior of interviewees will
be observed opportunistically. The most frequently mentioned norm in all Mestizo interviewees’
Norm Free-lists and the most frequently mentioned marker in all Mestizo interviewees’ Marker
 11

Free-lists will be designated the Mestizo Popular Norm and the Mestizo Popular Marker,
respectively, for each of the six interaction domains. Thus, six Mestizo Popular Norms and six
Mestizo Popular Markers will be identified (norms and markers may be duplicated across
domains). The use of free-lists minimizes observer bias by requiring the norms and markers
under investigation to be those most salient to the subjects, rather than to the investigator.
Each of the 30 interviewees will be asked about the frequency and nature of their
interactions across the Matsigenka-Mestizo ethnic boundary, and their interactions within their
own ethnic group, for each of the six interaction domains. Particular attention will be paid to
whether the interviewee was satisfied or dissatisfied with each interaction, whether the
interaction was perceived as necessary or important, and whether an interaction with another
person could have accomplished the same thing according to the interviewee. Such responses
suggest the subjective payoff structure of the coordination interaction. In each domain, the
frequency of reported out-group interactions (discounted by the number of in-group interactions,
and weighted by the relative subjective importance of out-group versus in-group interactions, if
appropriate) will be used to calculate an out-group interaction index for each interviewee, for
each interaction domain. In addition, each interviewee will be asked to: 1) name any number of
people who she/he considers to be knowledgeable about each domain of interaction; 2) to rank
order this list in terms of knowledge; and 3) to order the list in terms of people with whom she/he
socializes the most (not necessarily in the domain of interaction). This ego-centered social
network data (Wasserman & Faust 1994), will be used to identify groups of socialization
partners, e.g., close friends, and also prestigious individuals with disproportionate influence on
community norms. This grouping variable will be included in subsequent analyses to control for
the possibility that an individual’s norms and perceptions are influenced by frequent socialization
partners outside of the interaction domains in which the norms of interest apply. Phase 1 will
result in a set of six Mestizo Popular Norms and six Mestizo Popular Markers for Boca Manu, as
well as six out-group interaction index values and six social network group designations (one for
each of the six identified interaction domains) for each of the 30 Boca Manu participants.

Phase 2: Identification of Matsigenka Norms and Markers (15 Jan 2013 – 31 Dec 2013)
Bunce will repeat the procedures of Phase 1 in Tayakome. Twelve months are required in
order for Bunce to assure (through extensive participant observation and discussion with
informants) the accuracy of his interpretations of the many unique Matsigenka norms and
markers. The four key informants in Tayakome will be representative individuals with whom
Bunce has already developed important relationships during the pilot studies. Interaction
domains for the Norm Free-lists and Marker Free-lists will match the six domains used in Phase
1. Phase 2 will result in a set of six Matsigenka Popular Norms and six Matsigenka Popular
Markers for Tayakome, as well as six out-group interaction index values and six social network
group designations for each of the 30 Matsigenka participants. During the final month of Phase
2, a Matsigenka assistant will be hired to back-translate the structured interview questions, as
well as to discuss possible variations of presentation for interviews, and (along with the key
informants) to be available to discuss any unexpected ethnographic observations.

Phase 3: Measurement of Norm and Marker Distributions (1 Jan 2014 – 28 Feb 2014)
A binary-choice vignette (Bernard 2006) will be designed for each of the six Matsigenka
Popular Norms and six Mestizo Popular Norms (12 total vignettes). Each vignette will describe a
 12

hypothetical interaction corresponding to one of the six interaction domains chosen in Phase 1.
There will be two vignettes for each domain, one illustrating violation of the Matsigenka Popular
Norm, and the other illustrating violation of the Mestizo Popular Norm for that domain. During
re-interviews with the 30 Tayakome and 30 Boca Manu participants, interviewees will indicate
either approval or disapproval of the action portrayed in each of the 12 vignettes, along with a
narrative explanation of how they arrived at their conclusion (i.e., how the illustrated norm
relates to their beliefs/worldview – for testing Hypothesis 2E). The set of 12 responses
(approval/disapproval) constitutes a participant’s Norm Complement. In Tayakome, the
Matsigenka assistant will again be hired for back-translation and presentation suggestions.
During each interview, Bunce will record whether the interviewee physically displays the
Matsigenka Popular Marker and/or the Mestizo Popular Marker identified for each of the six
interaction domains (12 total markers). If a particular marker is ephemeral (e.g., face paint rubs
off), the interviewee will be asked to recall whether he/she displayed the marker during his/her
last interaction in that domain. The set of markers displayed by the interviewee, from among the
set of 12 identified markers, constitutes a participant’s Marker Complement. Distributions of
particular norms and markers, as well as norm-marker covariance, both within ethnic groups and
across the ethnic boundary, are taken directly from the individual-level Norm and Marker
Complements and are used, with the out-group interaction index, to address Hypotheses 2A-D.

Phase 4: Perceptions of Norm and Marker Distributions (1 Mar 2014 – 30 Apr 2014)
The Norm and Marker Complements of each of the 60 study participants (Tayakome and
Boca Manu) will be recorded on index cards (without names). During re-interviews with the 30
Tayakome participants (aided by a local assistant for back-translation and presentation
suggestions) and the 30 Boca Manu participants, interviewees will be re-familiarized with the
vignettes from the previous series of interviews and: 1) Choose an index card at random, and,
without looking, guess the 12 recorded answers of the card’s Norm Complement (each of the
card’s answers corresponding to one of the 12 vignettes from the previous series of interviews),
knowing only to which community the card belongs. Guesses indicate the norms the interviewee
perceives to be held by most members of the card’s community. 2) Repeat 1, but now the
interviewee is told the card’s Marker Complement as well as the card’s community.
Modifications to guesses dependent on marker indicate that the interviewee perceives norm-
marker covariance among members of the community to which the card belongs.
To motivate deliberated responses, interviewees will receive a compensatory reward for
each correct guess (12 possible rewards for each of the two cards drawn per interviewee), similar
to the “guess game” of Gurven et al. (2008). In practice, the “strategy method” of Camerer and
Fehr (2004) will be used, so that interviewees make guesses about a card’s Norm Complement
under all possible combinations of the card’s community membership and Marker Complement,
before a card is actually drawn. The interviewee then draws a card at random and rewards are
calculated based only on those guesses that pertain to the card’s actual characteristics.
Interviewees will be asked to explain each of their guesses, e.g., Does your guess of
“disapprove” for norm 5 indicate that you believe that all people in Tayakome would disapprove
of this vignette, that most would disapprove, or that you don’t know and guessed at random?
Phase 4 will result in a set of perceptions (guesses), for each of the 60 participants, about
the most common norms in the in-group and the out-group under differing marker complements,
as well as a qualitative approximation of perceptions of variance in those norms (from the
 13

explanations of guesses). These perceived norm distributions and norm-maker covariance will be
compared to actual distributions and covariance in order to address Hypotheses 3A-B.

Phase 5: Development of Agent-Based Evolutionary Models (6 May 2014 – 30 Sep 2014)

Bunce will spend five months at UC Davis developing a formal model relating inter-
ethnic coordination interactions to changing norm and marker distributions in ethnic groups. He
will use the model of McElreath et al. (2003) as a starting point, but will incorporate new
insights gained through analysis of the field data. Particular attention will be paid to the
incorporation of unbalanced payoff structures in coordination interactions, i.e., coordinating
on norms yields different net benefits to the two interaction participants. This modification is
addressed only briefly by McElreath et al. (2003), and may be an important aspect of
Matsigenka-Mestizo interactions, relating to a difference in status between the two ethnic groups.
The relevance of an unbalanced payoff structure in the study population will be assessed by
using data from Phases 1, 2, and 3 to address Corollary 2C.
A second modification of the McElreath et al. (2003) model will be the incorporation of
agents’ perceptions of norm and marker distributions and norm-marker covariance. The new
model will explore the effects of such perceptions on the frequency of out-group interaction, and,
by extension, on changes in actual norm and marker distributions. The frequency of out-group
interactions will also be allowed to affect perceptions. These effects will be incorporated into the
model based on analyses using perception data from Phase 4 to address Hypotheses 3A-B.
The value of formal model development lies in the unique insights into causal
mechanisms that can be gained by making relationships explicit between variables that are
hypothesized to be important to the dynamics of the system (i.e., changing norm and marker
distributions). A long-term goal of this research program is to collect a dataset of cultural change
across the Matsigenka-Mestizo ethnic boundary that can be used to compare the original model
of McElreath et al. (2003) with the new model(s) that will be formulated in this investigation.
These comparisons will determine whether the proposed modifications to the model actually add
to our understanding of cultural change. Such a dataset requires a temporal dimension, and thus
the data collected in this investigation will constitute a baseline, to be augmented in the future.
This phase of the project will include the collaboration of an advanced Ph.D. student
in the Anthropology or Ecology graduate program at UC Davis who is trained in the
development and simulation of evolutionary models. This student collaborator will be appointed
to a half-time Research Assistant position, will work closely with Bunce, and will co-author
resulting publications as appropriate. This collaboration will enrich the research experience of a
graduate student, provide valuable assistance, and bring new perspectives to the project.

Data Analysis
Hypotheses 1A-C, that norms (signaled by markers) govern a set of coordination
interactions, are tied to people’s beliefs/worldviews, and define a salient Matsigenka-Mestizo
ethnic boundary that structures a set of inter-individual interactions, will be examined through
content analysis (Bernard 2006) of the transcribed and coded ethnographic data collected during
Phases 1 and 2 (e.g., field notes, informal interviews, and autobiographical narratives).
Inter-ethnic and inter-individual differences in norm and marker distributions will be
examined using principal components-based methods on data from Phase 3. For instance,
individual responses to the 12 norm vignettes (Norm Complement) can be reduced to component
 14

scores using non-linear PCA (Michailidis & de Leeuw 1998). Inter-ethnic norm differences
(Hypothesis 2A) would be indicated if Matsigenka and Mestizo individuals cluster in two
separate groups when their component scores are plotted. The extent of inter-individual variation
(Hypothesis 2B) would be indicated by the spread of the plotted scores within each group.
Multilevel logistic regression (Gelman & Hill 2007) will be used to examine the effect of
predictor variables such as out-group interaction index, ethnic group, worldview, age, and sex on
individuals’ binary responses to a particular norm vignette (an outcome variable). The
importance of out-group interaction index and ethnicity as predictors of vignette response might
suggest that people’s norms match those of their most frequent interaction partners (Hypothesis
2C). An (out-group interaction index)*(ethnicity) interaction might suggest that inter-ethnic
interaction has a larger effect on the norms of one ethnic group than on those of the other
(perhaps due to status differences: Corollary 2C). If holding a particular belief about a norm
(worldview) is a stronger predictor of vignette response than is the out-group interaction index,
such results might support Hypothesis 2E (norms strongly linked to beliefs are less likely to be
modified by interaction). In these regressions, the social network grouping variable is included as
a random effect to control for social influences on individual Norm Complements. The accuracy
of perceptions (guesses from Phase 4) about norm and marker distributions can be expressed as a
binary outcome variable [1 - |guessed response – mode(observed responses)|] and included in
similar multilevel models to examine the effect of worldview, out-group interaction index, and
ethnicity (and their interactions) on perceptions (Hypotheses 3A-B). The relationship between
norm-marker covariance and out-group interaction index (Hypothesis 2D) can be examined by
comparing the fit of quadratic versus linear equations to a plot of these two variables.

Work Schedule
Phase 1: Identification of Mestizo Norms and Markers 15 Aug 2012 – 31 Dec 2012
Location: Boca Manu
Phase 2: Identification of Matsigenka Norms and Markers 15 Jan 2013 – 31 Dec 2013
Location: Tayakome
Phase 3: Measurement of Norm and Marker Distributions 1 Jan 2014 – 28 Feb 2014
Location: Tayakome (January), Boca Manu (February)
Phase 4: Perceptions of Norm and Marker Distributions 1 Mar 2014 – 30 Apr 2014
Location: Boca Manu (March), Tayakome (April)
Phase 5: Development of Agent-Based Evolutionary Model s 6 May 2014 – 30 Sep 2014
Location: University of California, Davis
During Phase 5 and data analysis (Oct - Dec 2014), Dr. Mark Grote, Senior Statistician (Anth.,
UC Davis) will serve as a consultant, advising and trouble-shooting as needed while Bunce, who
has prior experience using all of the proposed analytical methods, designs the models and
conducts the analysis. Bunce has worked with Dr. Grote on several previous projects.

SIGNIFICANCE
Intellectual Merit
The major contribution of this study will be to forge connections between the broad
historically-based descriptive explanations of cultural change provided by Socio-Cultural
Anthropology (e.g., Varese 2006), the understanding of how perceptions influence inter-personal
and inter-group behavior provided by Social Psychology (e.g., Tajfel 1982), and formal models
 15

of cultural evolutionary theory (McElreath et al. 2003). This study is the first to formally test
hypotheses of cultural change by relating inter-individual interactions, norm and marker
distributions, and perceptions of these distributions across an ethnic boundary. Even more
importantly, by operationalizing and incorporating concepts such as status differences, and the
integration of norms and markers into individual worldviews, this study creates a conceptual link
between the historical processes that contributed to such inter-group differences in status and
worldview, and the effect of these differences on the individual decision-making processes that
result in cultural change at the population level. This study is transformative because it unites
disparate social science fields into a new approach for understanding general processes of
cultural change as they apply to an ethnically-stratified human population. The proposed study
constitutes the first step in a long-term research program, where replication and expansion of this
methodology to other Matsigenka and Mestizo populations will produce a longitudinal dataset
appropriate for comparing historical and evolutionary models of cultural change.

Broader Impacts
As field anthropologists, it is impossible to study cultural change without influencing the
process under study (Davies 1999). We have a duty to insure that this influence is beneficial
(e.g., García 2000; Said 1989), yet we have no authority to decide what constitutes beneficial
cultural change or maintenance (Jackson 1989). As an answer to this challenge, the goal of this
study is to understand the mechanisms of cultural change at work among both Matsigenka and
Mestizos, and use this knowledge to empower both communities to take control of the process to
the extent that they wish. Cultural change is important to the Matsigenka of Tayakome, but
Bunce witnessed uncertainty about the mechanisms by which it occurs when, in 2010, a debate
erupted in a communal meeting about whether to permit an adolescent girl to accompany a
teacher for a two-week vacation in the city of Cusco. At issue was whether the girl would lose
her Matsigenka customs. Thus, there is need for a clearer understanding of how inter-ethnic
interactions affect the loss of ethnic-typical norms. With such knowledge, the Matsigenka (and
other indigenous groups) might design effective cultural management plans to limit the loss of
norms and markers that they collectively want to preserve. Having such a plan might place the
Matsigenka in a stronger position to negotiate with park authorities over reform of paternalistic
policies to “protect indigenous culture” (Shepard et al. 2010). At the conclusion of fieldwork,
Bunce will discuss his findings and interpretations at communal meetings in Tayakome and Boca
Manu. In additional to scientific publications, technical reports (in Spanish) and focused
summaries (in Spanish and Matsigenka) of the results will be deposited with each community,
the park administration, FENAMAD, and the Pontificia Universidad Católica del Perú (Lima).
This study integrates research and teaching. At the request of Tayakome, Bunce
teaches English classes one day per week for Matsigenka adults and children. The idea for these
classes originated from within the community (not from Bunce), and places the children on a
more even footing with their Mestizo peers in Boca Manu who learn English as part of their
standard curriculum. The classes in 2011 were rewarding for both Bunce and the community,
and Bunce has been asked to expand them in 2013 to include math and reading for adults. In
Boca Manu, Bunce will offer instructional support and seminars to local middle- and high-school
students and adults on the region’s cultural and biological diversity. Phase 5 of this study
includes the collaboration of a graduate student, with an aim of advancing his or her research
expertise and professional development through training in methods, analysis, and publication.
 1

References Cited
Ethnic Boundaries and Cultural Change in an Amazonian Population
P.I. John Bunce

Alba RD. 1990. Ethnic Identity: The Transformation of White America. New Haven: Yale
University Press. 374 pp.

Atran S, Medin D, Ross N, Lynch E, Vapnarsky V, Ucan Ek’ E, Coley J, Timura C, Baran M.
2002. Folkecology, cultural epidemiology, and the spirit of the commons: a garden
experiment in the Maya lowlands, 1991–2001. Current Anthropology 43:421-50

Baer G. 2004. ¿Comes carne de armadillo? acerca del problema de la identidad de los
Matsiguenka del Bajo Urubamba en el Oriente Peruano. Boletín de Antropología
(Universidad de Antioquia) 18:317-36

Balslev Clausen H. 2008. La etnicidad de los Mazahuas en un municipio mexiquense: un estudio

de la violencia simbólica. In Revisitar la Etnicidad: Miradas Cruzadas en Torno a la
Diversidad, ed. D Gutiérrez Martínez, H Balslev Clausen, pp. 249-73. México: Siglo
XXI, El Colegio Mexiquense, El Colegio de Sonora

Barth F. 1998[1969]. Introduction. In Ethnic Groups and Boundaries: The Social Organization
of Culture Difference, ed. F Barth, pp. 9-38. Long Grove: Waveland Press, Inc.

Bernard HR. 2006. Research Methods in Anthropology: Qualitative and Quantitative

Approaches. Lanham: Altamira. 803 pp.

Bicchieri C. 2006. The Grammar of Society: The Nature and Dynamics of Social Norms.
Cambridge: Cambridge University Press. 260 pp.

Boyd R, Richerson PJ. 1985. Culture and the Evolutionary Process. Chicago: The University of
Chicago Press. 331 pp.

Brauer M. 2001. Intergroup perception in the social context: the effects of social status and group
membership on perceived out-group homogeneity and ethnocentrism. Journal of
Experimental Social Psychology 37:15-31

Camerer CF, Fehr E. 2004. Measuring social norms and preferences using experimental games: a
guide for social scientists. In Foundations of Human Sociality: Economic Experiments
and Ethnographic Evidence from Fifteen Small-Scale Societies, ed. J Henrich, R Boyd, S
Bowles, C Camerer, E Fehr, H Gintis, pp. 55-95. Oxford: Oxford University Press

Davies CA. 1999. Reflexive Ethnography: A Guide to Researching Selves and Others. London:
Routledge. 257 pp.

de la Cadena M. 1992. Las mujeres son más indias: etnicidad y género en una comunidad del
Cuzco. Isis Internacional, Ediciones de las Mujeres (Santiago de Chile) 16:25-71
 2

de la Cadena M. 2005. Are mestizos hybrids? the conceptual politics of andean identities.
Journal of Latin American Studies 37:259-84

Degregori CI. 1998. Pueblos indígenas y democracia en América Latina. Quehacer 114:36-41

Delgado G. 1996. Entre lo popular y lo étnico: notas de un debate para un debate. In Pueblos
Indígenas, Soberanía y Globalismo, ed. S Varese, pp. 31-79. Quito: Ediciones Abya-Yala

Efferson C, Lalive R, Fehr E. 2008. The coevolution of cultural groups and ingroup favoritism.
Science 321:1844-9

Espinosa Ó. 2009. Las organizaciones indígenas de la Amazonía y sus reivindicaciones.

Argumentos (Instituto de Estudios Peruanos) 3:
http://web.revistargumentos.org.pe/index.php?fp_cont=995

Gans HJ. 1979. Symbolic ethnicity: the future of ethnic groups and cultures in America. Ethnic
and Racial Studies 2:1-20

García ME. 2000. Ethnographic responsibility and the anthropological endeavor: beyond identity
discourse. Anthropological Quarterly 73:89-101

Gelman A, Hill J. 2007. Data Analysis Using Regression and Multilevel/Hierarchical Models.
Cambridge: Cambridge University Press. 625 pp.

Gil-White FJ. 1999. How thick is blood? The plot thickens . . .: if ethnic actors are
primordialists, what remains of the circumstantialist/primordialist controversy? Ethnic
and Racial Studies 22:789-820

Godoy R, Reyes-García V, Broesch J, Fitzpatrick IC, Giovannini P, Martínez Rodríguez MR,

Huanca T, Leonard WR, McDade TW, Tanner S, Bolivia Study Team TAPS. 2009.
Long-term (secular) change of ethnobotanical knowledge of useful plants: separating
cohort and age effects. Journal of Anthropological Research 65:51-67

Gow P. 1993. Gringos and wild indians: images of history in western Amazonian cultures.
L'Homme 33:327-47

Gurven M, Zanolini A, Schniter E. 2008. Culture sometimes matters: intra-cultural variation in

pro-social behavior among Tsimane Amerindians. Journal of Economic Behavior &
Organization 67:587-607

Henrich J, Ensminger J, McElreath R, Barr A, Barrett C, Bolyanatz A, Cardenas JC, Gurven M,

Gwako E, Henrich N, Lesorogol C, Marlowe F, Tracer D, Ziker J. 2010. Markets,
religion, community size, and the evolution of fairness and punishment. Science
327:1480-4

Holt FL. 2005. The Catch-22 of conservation: indigenous peoples, biologists, and cultural
change. Human Ecology 33:199-215
 3

Huntington GE, Hostetler JA. 2002. The Hutterites in North America. Belmont:
Wadsworth/Thomson Learning. 158 pp.

Jackson J. 1989. Is there a way to talk about making culture without making enemies?
Dialectical Anthropology 14:127-43

Jackson J. 1995. Culture, genuine and spurious: the politics of indianness in the Vaupés,
Colombia. American Ethnologist 22:3-27

Johnson A. 2003. Families of the Forest: The Matsigenka Indians of the Peruvian Amazon.
Berkeley: University of California Press. 258 pp.

Levi T, Shepard GH, Jr., Ohl-Schacherer J, Peres CA, Yu DW. 2009. Modelling the long-term
sustainability of indigenous hunting in Manu National Park, Peru: landscape-scale
management implications for Amazonia. Journal of Applied Ecology 46:804–14

Marggraff K. 2004. Aproximación a la identidad matsiguenka a partir de las influencias externas:

primeros acercamientos. Boletín de Antropología (Universidad de Antioquia) 18:337-49

McElreath R, Boyd R, Richerson PJ. 2003. Shared norms and the evolution of ethnic markers.
Current Anthropology 44:122-9

Michailidis G, de Leeuw J. 1998. The Gifi system of descriptive multivariate analysis. Statistical
Science 13:307-36

Nagata J. 1981. In defense of ethnic boundaries: the changing myths and charters of Malay
identity. In Ethnic Change, ed. CF Keyes, pp. 88-116. Seattle: University of Washington
Press

Nagel J. 1996. American Indian Ethnic Renewal: Red Power and the Resurgence of Identity and
Culture. New York: Oxford University Press. 298 pp.

Nave A. 2000. Marriage and the maintenance of ethnic group boundaries: the case of Mauritius.
Ethnic and Racial Studies 23:329-52

Oakdale S. 2008. The commensality of ‘contact’, ‘pacification’, and inter-ethnic relations in the
Amazon: Kayabi autobiographical perspectives. Journal of the Royal Anthropological
Institute (N.S.) 14:791-807

Ohl-Schacherer J, Mannigel E, Kirkby C, Shepard GH, Jr., Yu DW. 2008. Indigenous

ecotourism in the Amazon: a case study of ‘Casa Matsiguenka’ in Manu National Park,
Peru. Environmental Conservation 35:14-25

Ohl-Schacherer J, Shepard GH, Jr., Kaplan H, Peres CA, Levi T, Yu DW. 2007. The
sustainability of subsistence hunting by Matsigenka native communities in Manu
National Park, Peru. Conservation Biology 21:1174-85
 4

Ohl J, Wezel A, Shepard GH, Jr., Yu DW. 2008. Swidden agriculture in a protected area: the
Matsigenka native communities of Manu National Park, Peru. Environment, Development
and Sustainability 10:827–43

Ostrom TM, Sedikides C. 1992. Out-group homogeneity effects in natural and minimal groups.
Psychological Bulletin 112:536-52

Riketta M, Sacramento CA. 2008. `They cooperate with us, so they are like me': perceived
intergroup relationship moderates projection from self to outgroups. Group Processes &
Intergroup Relations 11:115-31

Rosengren D. 2003. The collective self and the ethnopolitical movement: “rhizomes” and
“taproots” in the Amazon. Identities: Global Studies in Culture and Power 10:221-40

Rosengren D. 2004. Los Matsigenka. In Guía Etnográfica de la Alta Amazonía, ed. F Santos, F
Barclay, pp. 1-157. Balboa: Smithsonian Tropical Research Institute, Instituto Francés de
Estudios Andinos

Ross N, Medin D, Cox D. 2007. Epistemological models and culture conflict: Menominee and
Euro-American hunters in Wisconsin. Ethos 35:478–515

Said EW. 1989. Representing the colonized: Anthropology's interlocutors. Critical Inquiry
15:205-25

Schultz PW, Nolan JM, Cialdini RB, Goldstein NJ, Griskevicius V. 2007. The constructive,
destructive, and reconstructive power of social norms. Psychological Science 18:429-34

Shepard GH, Jr., Rummenhoeller K, Ohl J, Yu DW. 2010. Trouble in paradise: indigenous
populations, anthropological policies, and biodiversity conservation in Manu National
Park, Peru. Journal of Sustainable Forestry 29:252-301

Smith RC. 1996. Las políticas de la diversidad: COICA y las federaciones étnicas de la
Amazonía. In Pueblos Indígenas, Soberanía y Globalismo, ed. S Varese, pp. 81-125.
Quito: Ediciones Abya-Yala

Tajfel H. 1982. Social psychology of intergroup relations. Annual Review of Psychology 33:1-39

Terborgh J. 1999. Requiem for Nature. Washington, D.C.: Island Press. 234 pp.

Tilly C. 2005. Identities, Boundaries, and Social Ties. Boulder: Paradigm Publishers. 269 pp.

Urteaga Crovetto P. 2007. The social construction of ethnic groups and indigenous peoples in the
Southeastern Peruvian Amazonia. Anales (Institute of Iberoamerican Studies, University
of Gothenburg) 9-10:73-106

Valdivia G. 2005. On indigeneity, change, and representation in the northeastern Ecuadorian

Amazon. Environment and Planning A 37:285-303
 5

Varese S. 2006. La Sal de los Cerros: Resistencia y Utopía en la Amazonía Peruana. Lima:
Fondo Editorial del Congreso del Perú. 350 pp.

Wasserman S, Faust K. 1994. Social Network Analysis: Methods and Applications. Cambridge:
Cambridge University Press. 824 pp.

Waters MC. 1990. Ethnic Options: Choosing Identities in America. Berkeley: University of
California Press. 197 pp.

Weber M. 1968. Economy and Society: An Outline of Interpretive Sociology. New York:
Bedminster Press. 398 pp.

Wilder DA. 1978. Reduction of intergroup discrimination through individuation of the out-
group. Journal of Personality and Social Psychology 36:1361-74

Wimmer A. 2008. The making and unmaking of ethnic boundaries: a multilevel process theory.
American Journal of Sociology 113:970-1022