Neanderthals

R E V I E W A R T I C L E

Rethinking the evolution of property and possession: A review and methodological proposition

Lucy Tibble1 | Susana Carvalho1,2,3,4

1Primate Models for Behavioural Evolution

Lab, Institute of Cognitive and Evolutionary

Anthropology, Oxford University, Oxford,

United Kingdom

2Gorongosa National Park, Sofala,

Mozambique

3Interdisciplinary Center for Archaeology and

the Evolution of Human Behaviour (ICArEHB),

University of Algarve, Faro, Portugal

4Centre for Functional Ecology (CEF),

University of Coimbra, Coimbra, Portugal

Correspondence

Lucy Tibble, Primate Models for Behavioural

Evolution Lab, Institute of Cognitive and

Evolutionary Anthropology, Oxford University,

Oxford, United Kingdom.

Email: [email protected]

Funding information

Leverhulme Trust, Grant/Award Number:

PLP-2016-114

Abstract Property is a key feature of modern human society; however, identifying the origin of this multi-

faceted behavior poses a formidable challenge. Here, we explore the methodologies for

researching the origin of property. We discuss how an interdisciplinary approach can shed light

on how our human ancestors shifted behaviorally from possessing an object to having exclusive

property control over it. Possession occurs when social group members only respect an individ-

ual's claim to have exclusive access to an object when the individual has physical control over

the object. Property occurs when an individual can claim exclusive access to an object, without

challenge, regardless of whether the object is in their physical control or not. Researchers across

different disciplines have asked what, if anything, distinguishes human property behavior from

the behavior of other animals? Further, when and how did this behavior evolve in our lineage?

Due to the considerable methodological challenges posed by researching this topic, few studies

have been able to directly address these questions. In this review, we explore the challenges

involved in defining property and possession and suggest a two-step approach to interdisciplin-

ary definitions. Next, we evaluate four core approaches to the study of property behavior: evo-

lutionary game theory, ethology, comparative cognition, and developmental psychology. Finally,

we propose an empirical study, using an ethological approach to test the presence of property

and possessive behavior in a natural setting, using our closest living relative, the chimpanzee

(Pan troglodytes). Overall, we argue that this field of research is at a turning point, where the

novel integration of various methods may provide an explanation to the origin of property.

KEYWORDS

behavioral evolution, chimpanzees, modeling, possession, property

1 | INTRODUCTION

Property behavior defines much of modern human life, from the coop-

erative transfer of land to our laws around thievery. The modern man-

ifestation of property behavior appears distinct from the rest of the

animal kingdom. However, research has established that property

behavior is derived from possessive behavior, widely documented in

nonhuman species.1 Numerous methodologies have been formulated

to understand how possessive behavior could evolve into what we

see today in modern humans. Game theoretic modeling has been used

to determine the contextual factors in successive competitive

scenarios which would lead to property behavior becoming an evolu-

tionarily stable strategy.2–6 Ethology and developmental psychology

have been used to investigate and categorize real-world possessive,

and property, behavior in animal and human interactions.1,7–10 Finally,

comparative cognition has been used to investigate the cognitive abili-

ties and biases of different species and to consider what features are

found in species that understand more complex rules of possession,

and possibly, property.11–16 There is debate regarding how

researchers should define property due to the differing practical

requirements of the four key methodologies. In this article, we consol-

idate and critique these methodologies, and their definitions, with a

view to advance the future of property research. We hope to encour-

age novel, interdisciplinary research which will address the evolution

of property behavior with direction and purpose. Corrections made on 16-October-2018, after first online publication: deletion

of three quotes from pages 3 and 6 of the article.

Received: 21 January 2018 Revised: 5 June 2018 Accepted: 1 August 2018

DOI: 10.1002/evan.21748

Evol Anthropol. 2018;27:285–296. wileyonlinelibrary.com/journal/evan © 2018 Wiley Periodicals, Inc. 285

2 | DEFINING POSSESSION AND PROPERTY

We define possession as a context in which social group members only

respect an individual's claim to have exclusive access to an object

when the individual has physical control over the object.7,17 Depend-

ing on the rank or dominance of the possessor, they may be able to

extend their physical proximity and maintain possession of the

object.1,7,10 Next, we broadly define property by a context in which an

individual can claim exclusive access to an object, without challenge,

regardless of whether the object is in their physical control or not.17

Thus, an individual could show either possession or property toward a

physically controlled object. However, if an individual maintains exclu-

sive access without physical control, this behavioral outcome is prop-

erty and not possession. Therefore, property could include a wide

range of objects and individuals from territories and fruiting trees, to

mates. In line with the two behavioral outcomes above, we would use

the following terms to describe the individual and the action they per-

form: “possessors” performing “possessive behavior” or “property

holders” performing “property behavior.” We opt for the term “prop-

erty holder” rather than “owner” in an attempt to overcome past

inconsistencies in the use of the term “owner.”2 For other definitions

of key terms used in this article, we provide a Glossary.

We propose a two-step approach to interdisciplinary definitions

of property. In the first step, we recommend that researchers use a

broad, species-neutral definition like the one we have just described.

This first step is in line with recent calls for researchers to use more

neutral terms.3 Species-neutrality is important because we expect

there to be differences between the human, highly derived form of

property and the behavior shown by the rest of the animal kingdom.

There is a risk of type II error in which research could fail to detect

property in other animals if the definitions are too strict and human-

specific. For example, it would be overly strict to suggest that prop-

erty must be rooted in the human institution of language. This would

immediately exclude the possibility of many species displaying prop-

erty behavior.24 Further, there is no evidence to suggest that property

behavior requires formal institutionalization through language.1 In the

second step, researchers can begin to restrict the first-step definition

as necessary for their chosen methodology and species. Restrictive

definitions can be useful to the point of narrowing research focus and

generating ideas for experimental design. Restrictions may be neces-

sary because research requires operational and specific criteria to

meet, or not meet, to attain results valued by the field. However,

restrictions can also increase the risk of type II error. For example,

researchers could choose to define property as social group members

defending an absent individual's exclusive access to an object in the

case of thievery.17 Third party punishment of thieves would be a

strong case for claiming that property exists in a species but a nega-

tive result based on a restrictive definition could be due to type II

error. To determine if the third party punishment restriction could fea-

sibly lead to a positive result, researchers can use pilot data testing

and the collection of anecdotal evidence. This preliminary work can

help researchers to justify why they have applied certain restrictions

to their definition of property. Overall, we hope that this two-step

approach forms a useful way for researchers to structure the composi-

tion of their definitions with the minimal number of restrictions.

The definitions we have provided above relate to the first of the

two-step approach. Although we acknowledge that some authors define

property as respect for possession, we prefer to separate these terms for

both semantic and practical reasons.6 Defining property as a respect for

possession is too restrictive for our preference. This is because posses-

sion is semantically linked to physical control. By extension, physical con-

trol cannot adequately capture abstract aspects of property. Practically,

we prefer to use two terms rather than using property as a blanket term

for the spectrum of behaviors between simple possession and abstract

human property. This is because we want to create a useful distinction

between the wide ranges of behaviors. Finally, our definitions are broad

and species-neutral enough to mitigate the risk of the separated terms

polarizing behavior, and therefore, failing to capture the full range of

behavioral variation. For example, a first-step definition which requires a

third party punishment system would fail to capture the repertoire of

species whose behavior which goes beyond physical control but does

not include third party punishment. Thus, we hope to tread the fine line

between specific and species-neutral definitions, general enough for us

to detect the behaviors, but specific enough to facilitate objective

research.

3 | KEY APPROACHES TO THE EVOLUTION OF PROPERTY

We discuss four key approaches: evolutionary game theory, ethology,

comparative cognition, and developmental psychology (see Table 1).

These can be combined to answer Tinbergen's four questions for

explaining behavior: function, evolution, causation, and ontogeny.25

4 | EVOLUTIONARY GAME THEORY

Evolutionary game theorists can answer Tinbergen's “function” ques-

tion by modeling the socioecological contexts, which lead to posses-

sive or property behavior and identifying the differences between

them.25 In simplified terms, evolutionary game theory argues that a

population of players with a range of available strategies will “evolve”

through games, or strategic interactions, over time.26 This is because a

strategy may provide a higher payoff to players when it is used against

other strategies and itself, under specific conditions. If a population of

players adopts this strategy and they do not switch to available alter-

natives, then it is an “evolutionarily stable strategy.”2,26,27 Therefore,

evolutionary game theory models identify crucial assumptions neces-

sary for possession or property to plausibly emerge as evolutionarily

stable strategies.26,27 This balance of focus on both possession and

property means that some researchers have been able to step away

from human-centric definitions of property and call for species-neutral

definitions.3 Evolutionary game theorists tend to use less restrictive

definitions than ethologists because they begin with a theoretical

behavior and generate strategies and contexts based on this behavior.

Game theory definitions directly impact the model assumptions and

strategies, therefore a minimal number of restrictions are still neces-

sary for model validity and plausibility. On the other hand, ethologists

must justify their categorization of recorded behavior. This

286 TIBBLE AND CARVALHO

methodology naturally leads researchers toward more restrictive,

species-specific definitions. Therefore, evolutionary game theorists

tend toward minimally restricted species-neutral definitions, which we

suggest are the crucial bases for every researcher investigating the

evolution of property.

Evolutionary game theory has established that possession is often

respected because the risk of personal injury is too high. The “Bour-

geois” strategy suggests that under certain simple model assumptions,

the optimal strategy is to defend an object if you possess it, and not

to thieve objects possessed by other individuals.2,4 This applies when

the value of winning the conflict over the object is less than the cost

of losing. The value of winning access to the object relates to an

improvement in fitness and ultimately, reproductive advantage.28

These strategies show that respect for possession can plausibly

spread through a population as an evolutionarily stable strategy. This

model illustrates that possession, a step toward property, can be

solved by common sense explanations such as avoidance of costly

fights.27 However, “Bourgeois” may not be the evolutionarily stable

strategy if conditions change. A higher payoff strategy may become

available, such as the assessor strategy in which a player can use an

assessment of their opponent's behavior to inform whether they

should escalate a fight, flee or attempt to share the resource.2 In line

with this strategy, recent research has shown that signals which indi-

cate a willingness to defend property can be more effective at main-

taining property, than actual prevention.29 In this way, evolutionary

game theory models have successfully identified key factors which

impact the emergence of both possessive and property behavior.

Other models have identified: the costs of competition, the value or

fitness advantage of the challenged object, whether objects are plenti-

ful and whether objects are dispersed or clustered.5,30

Evolutionary game theory research has also suggested that prop-

erty or “privatization” may have been favored by natural selection, as a

solution to the tragedy of the commons.3,31 The tragedy of the com-

mons occurs when a depletable resource is reduced by individual com-

petition which will reduce the fitness of the group.31–34 This “group”

could also refer to population or species levels.32 The tragedy is that

resources run out quickly when individuals are selfish, whereas if

individuals use the resource at a sustainable rate and allow it to replen-

ish, the group as a whole will benefit. Evolutionary game theory

research has shown that the tragedy can be solved by privatizing the

resource or reducing competitor rivalry by restricting the rate at which

they can exploit the resource.31 When individuals “privatize,” or show

property behavior toward, resources, the pressure on the individual to

use up that resource before others do, is removed. Therefore, the trag-

edy of the commons can be solved by property behavior and this pro-

vides one possible answer to Tinbergen's “function” question for

property behavior.25 Further, game theorists have found that privatiza-

tion is favored by natural selection at low value of relatedness.31 This

means that the evolution of property behavior could have occurred in

large, highly dispersed populations as well as in smaller, clustered

populations.

Recent work has suggested a strategy for the emergence of prop-

erty, where competitors are of equal strength, with no difference in

resource holding potential. The “Restraint with Retaliation” strategy

involves retaliating only when someone challenges your access to an

object.3 The strategy is evolutionarily stable, under strict model condi-

tions, which incorporate repeated raids over time and winner and

loser effects (see Glossary).3,23,35,36 Full restraint from raiding, mean-

ing respecting property, would not spread in a population of “uncondi-

tional raiders,” who will never show restraint. However, “Restraint

with Retaliation” could plausibly spread because as population size

increases, the presence of additional individuals will increase the likeli-

hood of a raiding individual being raided in return which would

increase the costs associated with raiding.3 Thus, an individual in a

population of individuals showing “Restraint with Retaliation,” will

have a higher pay off than individuals in a population of “uncondi-

tional raiders.”3 Importantly, this model uses an assumption of com-

petitive equality, which suggests that property can emerge without

common sense explanations such as the costs of competing. Instead, a

more complex explanation might be impacting the players' choices,

such as the impact of fights on long-term relationships in terms of sta-

tus and group hierarchy. Further, this model invites exchange between

evolutionary game theory and ethology. By investigating multiple spe-

cies displaying different strategies, ethologists could test the

TABLE 1 Summary of key approaches

Approach How different approaches investigate the evolution of property

Evolutionary game theory

This approach looks at how the behavioral outcome of a competitive event can be predicted using theoretical modeling of the possible “decisions” available to players. Using this method, researchers can model the required conditions for property behavior to emerge. Evolutionary game theory helps us to understand how the adaptive value of different behavioral decisions, which ultimately impact fitness, has led to the variety of possession and property behavior we see in the animal kingdom today. This method targets the “function” part of Tinbergen's four questions.25

Ethology Using natural behavioral evidence from the animal kingdom, researchers can gather data on social interactions involving access to objects. For example, the analysis of behavioral responses to scent-markings around territories. This data can be used to determine the extent to which outcomes can be explained by property or possession rules. By comparing behavior across species, ethology can help to answer Tinbergen's “evolution” question.25

Comparative cognition

Researchers compare the cognitive performances of different species when tested in a controlled environment. For example, researchers can test individuals' memory to investigate their ability to create associations between objects and individuals. This is more often performed in laboratory settings but can also be applied to a field experimental setting. Comparative cognition can inform our understanding of possessive and property behavior on a phylogenetic level, contributing to the question of how the “evolution” of possession and property behavior took place. It can also improve our understanding of the way the brain proximately operates in situations of possession and property, addressing the question of “causation.”25

Developmental psychology

By studying the emergence of psychological biases linked with property in human children's development, researchers can suggest which features of property behavior are most likely innate to humans and what may have defined a rudimentary form of human property behavior before the rise of complex cultural institutions. In this way, developmental psychology can answer Tinbergen's “ontogeny” question.25

TIBBLE AND CARVALHO 287

hypothesis that raiding will decrease when population sizes increase.

Ultimately, the population size predictions of the “Restraint with

Retaliation” strategy can be used to suggest a plausible time frame for

the emergence of property in the human lineage.

5 | ETHOLOGY

Ethological research has demonstrated a wide range of animal behav-

ioral traits surrounding the possession of objects. To answer the Tin-

bergen question of the “evolution” of property behavior, it is

important to understand and compare different species' behaviors.25

This method allows us to piece together the evolutionary emergence

of property, utilizing both researches on animals displaying the sim-

plest possessive behavior all the way up to complex property behavior

in humans. In addition, ethology is a crucial way in which the field can

challenge whether property behavior is unique to humans.

As mentioned, ethologists favor restrictive definitions. This is

because the categorization of behavior becomes less subjective if you

use stricter criteria. We believe that researchers will benefit from start-

ing research with a broad, species-neutral definition such as those used

by game theorists. However, we acknowledge that researchers need to

operationalize their definitions in the second of the two-step approach,

such that they are relevant for the particular species and context they

wish to test. Paradoxically, ethologists need a fairly clear idea of what

property and possession would look-like in a particular species to design

the test correctly. Historically, researchers have tackled this issue by per-

forming pilot studies. Thus, in line with our preferred method of defining

property, ethologists can still restrict their definitions through systematic

pilot studies and in-depth research of the zoological literature.

Ethological studies have found that several nonhuman primates

show possessive behavior over a broad range of objects, including

food, territory, mates, toys, and even tokens representing food.1,37–41

In some species, such as long-tailed macaques, Macaca fascicularis and

hamadryas baboons, Papio hamadryas, the outcomes of possession

conflicts can be reliably predicted by the rules of proximity and rank

(see definitions of these principles in the Glossary).1,7,10 These rules

are closely linked to physical control and therefore we have defined

them as possessive behavior. These principles illustrate that individ-

uals' physical proximity to a target object, as well as individuals' rank,

relative to opponents, are important variables in predicting the out-

come of challenges. It seems the proximity rule can often be more

important than dominance, with events in which the closest individual

to the object will win possession, even if they are of lower rank.37,42,43

Similar evidence has been published for brown capuchins, Cebus

apella.44 There are common-sense explanations for why a dominant

individual would permit a subordinate to retain possession of an

object. As discussed by evolutionary game theorists above, the costs

of competition may outweigh the possible benefits.44,45 For example,

if the target object is of low value and is difficult to move, there are

additional transportation costs with potentially low fitness benefits if

you win the competition. Thus, ethological research has provided

empirical evidence for two key variables likely to affect possession

challenges: rank and proximity. This research has demonstrated varia-

tion within possessive behaviour in other animals, from simple to

complex rules, suggesting a possible evolutionary trajectory towards

the complex property behaviour we see in humans today.

Using the breadth of data on possessive behavior in wild species,

ethological research has suggested an evolutionary path between

from simple possessive interactions to scenarios guided by complex

possession rules. Research suggests that behavior progresses from

the precedence of power to the precedence of dominance status, and

finally to the precedence of prior possession.10 Hermit crab conflicts

demonstrate the precedence of power, in which individuals simply

assess relative fighting power and the larger crab aggresses and wins

generally.10,46 This is effectively a real world example of the assessor

strategy, described in the evolutionary game theory section. Next,

precedence of power becomes precedence of dominance when inter-

actions of power repeat and individuals learn to recognize the status

of other individuals without challenging. Precedence of dominance

reduces the chance of fighting along with the associated costs of com-

petition. Finally, precedence of prior possession begins when individ-

uals learn that possessors win over challengers.10 In precedence of

prior possession, physical power is effective only when the challenger

is much stronger than the possessor is. Future research can test this

phylogenetic hypothesis utilizing quantitative methods and a broad

range of species data. By revisiting this proposed model, we can

determine what contexts likely triggered key evolutionary changes

that led to more complex possession behavior in certain species.

Ethology has already provided important rules that guide many

animals' interactions around valued objects but there are still key chal-

lenges for those researching wild interactions today. On a grand scale,

it is not feasible to attempt to test every species and define how they

behave over objects. However, through interdisciplinary collaboration,

researchers can suggest which species may be most likely to show

possession and property. For example, researchers can achieve this by

assessing relevant cognitive abilities in a range of species. Another

key challenge is designing property tests for wild behavior so that

researchers can rule out any alternative explanations. In the final

section of this article, we present an example of how we can test for

property behavior in a wild species. In this way, we hope to encourage

more researchers to test likely species for property as well as posses-

sion. We believe that this targeted approach represents a promising

step toward understanding how and why property behavior emerged.

6 | COMPARATIVE COGNITION

With the benefit of laboratory controls, researchers have identified

cognitive abilities and biases of animals, including humans, with

regards to possession and property. Here we discuss two main

approaches to comparative cognition. First, researchers can test ani-

mals for cognitive skills considered necessary to facilitate property in

a social group. For example, X species cannot display property behav-

ior because they cannot remember associations between an object

and an individual beyond an hour. Second, researchers can test ani-

mals for cognitive biases that result from acquiring possessions or

property. For example, X species only shows biases toward foods that

physically exist long enough to require the restraint of group

288 TIBBLE AND CARVALHO

members. These approaches can help us suggest likely species for

researchers to test for property behavior.

Researchers can use data on various species' cognitive features

relating to property and possession to answer Tinbergen's question of

“evolution.”25 Researchers can also use this data set to determine the

“causation” explanation to property behavior, as distinct from that of

possession.25 In general, researchers design experiments so that the

observed behavior can only be explained by a particular cognitive bias

or skill. Therefore, comparative cognition favors restrictive definitions

of possession and property behavior. As with ethology, we believe

that the most useful comparative cognition research will begin with

species-neutral definitions and restrict definitions using support from

pilot studies.

Within this framework, one species has emerged as exhibiting

many of the cognitive skills and biases associated with property: Pan

troglodytes (see Figure 1). This suggests that chimpanzees could be a

likely candidate, worthy of further study. Itakura's research on lin-

guistically trained chimpanzees has been highly informative about

the extent to which the chimpanzee, our closest living relative, is

cognitively capable of property behavior. Ai, a 13-year-old linguisti-

cally trained chimpanzee, demonstrated the ability to make symbolic

associations between individuals and objects.12,13 Ai used bowl color

to determine associations between feeding bowls and individuals. In

the beginning, Ai showed food demanding behavior toward bowls of

all three colors, but over time, her food demanding behavior

decreased when the experimenter held a bowl that was not associ-

ated with her. Itakura argues that Ai was able to perceive a symbolic

representation of the relationships between objects and individuals.

It should be noted that Ai did not reject bowls that were associated

with other individuals. Therefore, she did not show full property

behavior. However, she was capable of creating and remembering

associations between individuals and particular bowls with a high

degree of accuracy. Future research could replicate this study with a

nonlinguistically trained subject to determine if this finding can be

applied to wild chimpanzees. This finding suggests that chimpanzees

are cognitively able to respond to abstract rules of property with no

social group present, based on the memory of associations. Further,

there is evidence that chimpanzees have the cognitive skills neces-

sary to understand trade, are able to barter with human experi-

menters showing rational maximization, and successfully comply

with exchange requests when linguistically trained.1,37,47,48 Further,

chimpanzees recognize working for a reward as establishing exclu-

sive property access.17 Therefore, comparative cognition has shown

that cognitive skills required to facilitate property are present in our

closest living relatives.

Cognitive biases associated with possession and property, are origi-

nally detected in humans and then experiments are designed to identify

if the bias exists in other species. For example, we know that the

endowment effect, a bias for valuing objects based on possession, has

been found in four great ape species-dependent on context.14–16 Fur-

ther, loss aversion, reference-dependence biases, and the endowment

effect have been observed in the highly encephalized and proficient

tool-user primates of the New World, the capuchins, Cebus sp. (see

Glossary).49,50 These findings suggest that similar cognitive biases

toward possessions exist in primate species, other than humans. It

seems that evolutionary pressure to secure access to valued posses-

sions may have driven the emergence of similar cognitive biases of both

humans and nonhuman primates. It would be useful to investigate what

these species share and how they differ from primates who do not

show these biases. One similarity which we discuss later, in a case

study, is the use of stone tools. Therefore, research into cognitive

biases relating to object access is useful in order to identify species with

similar behavioral repertoires to our own.

Overall, comparative cognition plays a key role in understanding

the evolution of possessive and property behavior. Species sharing

cognitive skills and biases comparable with humans invite further

behavioral investigation. Chimpanzees, capuchins, orangutans,

FIGURE 1 Top: Chimpanzee observing conspecifics using their

preferred toolsets to crack nuts. Middle: Chimpanzee being chased by the conspecifics after using their preferred toolsets. Bottom: Chimpanzee nut-cracking. Efficiency can be measured as the time taken to crack a nut or the number of blows required

TIBBLE AND CARVALHO 289

bonobos, and gorillas are all on this list.1,12–17,38,39,47–50 By under-

standing the cognitive features of other species, we can generate an

evolutionary tree of property behavior. Comparative cognition can

also contribute to our knowledge of the causation, or mechanism, of

possession and property. Future research could also explore if there

are human biases which appear during property, but not possessive

contexts.

7 | DEVELOPMENTAL PSYCHOLOGY

Developmental psychology provides us with information about the

extent to which innate psychological systems are able to explain

the modern manifestation of property behavior in humans. In real-

ity, numerous factors may impact the emergence of property in a

species, including: the objects targeted by property behavior, psy-

chological propensities toward property, the influences of culture

on individual characteristics of adults, and the social context.9

Psychological approaches can help us to distinguish which aspects

of property behavior are due to innate motivational systems and

which aspects are due to culture. Here, we focus on the evidence

of the development of property behavior in humans; however, we

hope that future research will explore the innate and socially

learnt aspects of possessive and property behavior in nonhuman

animals.

To answer Tinbergen's “ontogeny” question, developmental psy-

chologists have historically defined the developmental steps between

possession and property behavior.25 In our proposed definition, we

call for a species-neutral definition of property, as distinct from pos-

session. In line with this definition, current evidence suggests that

humans begin to demonstrate property behavior from the age of

24–30 months, as explored in further detail below. Before this age,

humans typically illustrate possessive behavior.51 Our method of

defining possession as distinct from property does not limit develop-

ment psychologists. Researchers can restrict their definitions of pos-

session into chronological phases before the age of 24–30 months, in

the same way that ethologists can restrict their definition to particular

species. It should also be noted that studying the increasing complex-

ity in possessive behaviors is just as important as studying the emer-

gence of property behavior. Studies which find the presence of

possession at particular ages, rather than property, constitute useful

contradictory evidence to proposed landmark ages in the develop-

ment of property behavior.

Early research attributed the widespread nature of property in

humans to a psychological tendency for accepted social

practices.52–60 Since then, developmental psychology research has

established supportive evidence for this tendency in the early and

effortless development of abstract property rules.61,62 For example,

there is an early and consistent use of the first possession rule in

human infants (see Glossary).8,26,62,63 Across cultures, children are sig-

nificantly more consistent and decisive in attributing property when

one of the potential holders has created the target object: the “creator

rule” (see Glossary).9,18–21 There also seems to be general psychologi-

cal propensities to support property behavior that develop in children,

although these propensities require cross-cultural validation. From

around 24–30 months, toddlers can: assert and acknowledge property

rights recognize and protect their own and others' property communi-

cate property status to others, and act to reestablish the rights of

property holders to access their property.51 Overall, human property

behavior is supported by consistently developing property rules and

psychological propensities to enforce norms with third party punish-

ment, to communicate with others to gain support for enforcing these

laws and finally to collaborate jointly and with shared

intentionality.17,64–66

At the age of three, in the cultures studied thus far, social factors

start to play a key role in shaping the development of possession and

property behavior. Children begin to experience possession as alienable

(i.e., tradeable and negotiable in exchanges) and they are forced to

socially evaluate if values in trade are “good,” and to take note of

accountability and reputation.67–69 Examples of cultural impact on pos-

session are cross-cultural differences in the development of the “first

possession heuristic”. For example, there is no evidence of any develop-

ment of the endowment effect over individually owned objects in the

case of the Hadza (see glossary).70,71 We suggest that property behav-

ior may exist in societies like the Hadza, but that cultural factors have

transferred individual behavior to the group level, over group-owned

property. Cultural factors may also influence individuals determining

which abstract possession or property rules are of most importance rel-

evant to the social context. For example, even though we may respect

“first possession” as a rule for possession rights, this may be discarded

in favor of rules by sex, the length of possession or prior pursuit.8 This

suggests that the demography, as well as the size, of the group present,

plays a key role in determining whether individuals show particularly

kinds of possessive or even property behavior.

Developmental psychology plays the key role in indicating the

ontogeny of property behavior. This research allows us to make

valuable expectations for when modern children typically display

this behavior. Further, with an ontogenetic understanding of prop-

erty, we can determine if property is an inherited part of human

behavior. This can be achieved by validating developmental results

across many cultures. This will have important implications for

designing future research, as we would gear the work of ethology

and comparative cognition toward investigating such innate

features.

8 | SIX STEPS FOR AN INTEGRATED APPROACH

All of the methodologies above are integral to answering Tinbergen's

questions, as applied to the evolution of property.25 This approach

involves a broad focus on the discipline's progress as a whole. It also

calls for an awareness of the current positioning of each methodology

and how they can interact with each other. Finally, the approach

requires the drive to innovate new research solutions, through

engagement with all these methods.

1. Look both at our own species and at the rest of the animal king-

dom (prioritizing likely candidates for property, as indicated by

290 TIBBLE AND CARVALHO

comparative cognition) and collate what studies need to be done

to fill in the blanks in our knowledge.

2. Draw on evolutionary game theory: Evolutionary game theory has

explained the factors that control the outcome of raids, such as

the value of the object, the strength of competitors, and so

on. Collaboration with comparative cognition is helping us to

understand how biases affect competition outcomes. Game the-

ory and developmental psychology can look theoretically at how

real-world competitions in human children develop toward prop-

erty behavior. Game theory can produce testable predictions to

be tested using ethological research.

3. Draw on ethology: Ethology has identified specific rules of posses-

sion in a number of animal species. In addition to broadening our

understanding of possession rules, ethology also actively tests

animal species for property behavior. The selection of likely spe-

cies to test for property can become more systematic with the

collaboration of ethology with comparative cognition and evolu-

tionary game theory. Finally, ethology can apply the methodology

of developmental psychology to gain a clearer understanding of

the ontogeny of possessive and property behavior in other ani-

mals. We suggest this research could begin with species-neutral

definitions and progress to more operationalized definitions of

what behavior, we would expect to see in that particular species.

4. Draw on comparative cognition: Comparative cognition has tested

cognitive skills and biases. This research can help to determine what

cognitive characters would be necessary for property behavior to

occur. It can then be followed up with systematic and standardized

research to confirm if these skills and/or biases exist in a range of ani-

mals. The selection and testing of target species would naturally

involve collaboration with ethology. Game theory and psychological

methods can be used to consider factors such as group composition,

including individuals' ontogenetic development.

5. Draw on developmental psychology: Developmental psychology has

investigated property behavior in children of different ages and

cultural backgrounds. This research helps us to understand the

extent to which the development of property behavior is innate in

humans. It also helps us to identify critical periods of development

for the behavior. Developmental research can be validated by

cross-cultural investigation. Subsequently confirmed innate fea-

tures, if confirmed at all, can be compared to other species' devel-

opment, through integration with ethology and comparative

cognition.

6. Integrate the findings from the above-mentioned disciplines in

new ways. For example, psychology of human behavior can pro-

vide ideas for property and possession biases, which we can test

in other animals under experimental settings using comparative

cognition. Another example is that comparative cognition can

guide ethology with regards to the selection of species.

With clear definitions, direction, and a multitude of methodolo-

gies at their disposal, research in this field is at a turning point. At this

critical position, interdisciplinary collaborations stand to accelerate

our learning about the evolution of property.

9 | A PROPOSITION FOR FUTURE EMPIRICAL RESEARCH: A QUASI- EXPERIMENTAL APPROACH CENTERED ON THE CHIMPANZEE, PAN TROGLODYTES

To conclude this review, we illustrate an example of how creative

approaches, informed by the interdisciplinary findings above, can be used

to empirically test for property in wild animal species.3 We suggest that

this approach can be used to contribute to Tinbergen's “evolution” and

“function” questions. Comparative cognition formed the basis for our spe-

cies selection and could be used to conduct important follow-up studies,

one of which is suggested below. Our selection of target object was

guided by the findings of evolutionary game theory and psychology. We

also used evolutionary game theoretic models to identify variables to test.

Finally, we took consideration of developmental changes in behavior

toward the target object in order to select an appropriate age group for

each test. Thus, this approach leverages interdisciplinary findings to gen-

erate novel tests for property in a hominin model species in the wild.

One of our closest living relatives, the chimpanzee, shares cogni-

tive skills and biases toward objects with humans (as discussed in the

“Comparative Cognition” section above). Our other closest living rela-

tive, the bonobo (Pan paniscus) does not show respect for possession

norms to the same extent as chimpanzees, likely due to lower punish-

ment costs.72 Chimpanzees are more manipulative with objects than

bonobos, which we suggest makes chimpanzees more likely to illus-

trate property behaviors in the wild.73 In addition, chimpanzees are a

suitable model species for stone tool using hominins that may have

had a patriarchal, rank based society. Therefore, a chimpanzee model

allowed us to indirectly test whether early hominins could have shown

property behavior, without requiring complex language, formal institu-

tions or a level of cooperation existent in modern humans.

Evolutionary game theory and psychology findings have identified a

list of factors which may affect whether an object will be a target for

property behavior.9 Research into the psychology of property suggests

that a target object are likely to be: durable, controllable, of functional util-

ity, capable of creation (in the sense that you can find one and create a

functional purpose for it), and necessary to access high energy resources.9

Evolutionary game theory research highlights the importance of object

value, availability, and distribution.30 We selected stone tools are our tar-

get object. Stone tools are functional, valuable objects that enable chim-

panzees to access high energy nuts. Stone tools are durable, can be

shaped or “created” and are controllable. Stone tools are not so easily

available and dispersed, that individuals would not compete for them.

On the other hand, they are not so rare, valuable or clustered that

one high-ranking individual would simply horde them. This means

that individuals in a chimpanzee population could have repeated stra-

tegic interactions over the same stone tools for relatively long period

of time. Therefore, we selected stone tools as a target object because

they are plausible and testable targets for property behavior in wild

chimpanzee populations.

Ethologists testing for possession or property behavior in any

land-dwelling species would benefit from the example set by the wild

laboratory, set up by Matsuzawa and colleagues in Bossou forest,

Guinea, in 1988.74 This method captures natural chimpanzee

TIBBLE AND CARVALHO 291

behavior, in a context which allows researchers to manipulate a quasi-

laboratory context. The laboratory is a natural clearing (7 m by 20 m),

situated at the top of Mount Gban, and chimpanzees voluntarily visit

the so-called “outdoor laboratory” (7�390N, 8�300W) to crack nuts,

among other activities. Therefore, we suggest that the wild laboratory

would be the ideal location for this novel approach.

Next, we needed to assess that this combination of species, target

object and further, specific population, were worthy of investigation. We

created simple hypotheses for chimpanzee tool use behavior, assuming

that they did not display possession or property. Then, we applied avail-

able data from Bossou to determine whether the chimpanzee behavior

could be explained without possession or property behavior (see Table 2).

Having confirmed that these explanations do not dismiss a posses-

sion or property explanation a priori, we considered what observable

behavior would be sufficient to constitute evidence for possession and

beyond this, property. We formed three testable hypotheses for posses-

sion and property behavior (Table 3, see also Illustration 1). The first two

hypotheses test for support or rejection of the existence of property

behavior over stone tools in chimpanzees. The variables used in these

hypotheses reflect the importance of signals and the number of individ-

uals present, as suggested by evolutionary game theory findings.3,29

These hypotheses contribute to Tinbergen's “evolution” question.25 The

third hypothesis targets a possible evolutionary advantage for the selec-

tion of property behavior over stone tools: the increase of tool skill, spe-

cifically efficiency, due to increased time spent with particular tools,

which may lead to increased energetic intake. Therefore, this hypothesis

contributes to Tinbergen's “function” question.25

9.1 | Test 1: Looking and taking: Investigating observation, group presence, and takeover events

A test of how much time adults spend observing conspecifics during

nut-cracking, without “taking over” toolsets, could illuminate whether

or not adults are monitoring conspecifics' preferences, relating to

respect for stone tool property norms.77 Stone tools preferred by

highly skilled individuals may be more subject to “takeovers” because

individuals prefer to observe the most proficient nut-crackers of the

group as observers get older.78

If possessive or property behavior is present, the likelihood of a

departure event being followed by a “takeover event” will be higher

with increased monitoring. This is because individuals, who are about

to takeover a toolset, should monitor the user carefully, so that they

use the object out of sight of the possessor/property holder, to avoid

direct punishment. Second, in property behavior, but not in posses-

sion behavior, individuals should be wary of third party punishment.

Therefore, takeover events will negatively correlate with group size.

Takeover events may be more likely in smaller groups because the risk

to the “thief” of being punished would be reduced (for operational

definitions, see Table 4).

TABLE 2 Alternative hypotheses for stone tool use reuse at Bossou, Guinea

Hypotheses Predictions Preliminary analysis

(1) Chimpanzees do not have stone tool preferences, they use what is available.

They will use toolsets (hammers and anvils) at random, based on availability.

A chimpanzee will use a less preferred, but available stone tool rather than go hungry. Nevertheless, there is evidence that they use the same toolsets more than expected by chance.75

(2) Chimpanzees do not defer to others' preferences for stone tools beyond proximity and/or rank.

Respect for possession exists only if the property holder is within arm's reach of the object. Respect for possession is also dependent on the respective ranks of the two “players.” When the property holder is absent and their stone tool becomes available, individuals with less efficient stone tools will use them.

As they grow, juveniles are given harsher punishments than infants, if they try to use other individuals' tools while they are present. This indicates that individuals are encouraged to find and use their own preferred tool. they are discouraged from using others' tools even if they available. A new study shows that juveniles are the main tool “recyclers” and this affects their tool-use efficiency.76

(3) Due to associative mechanisms between food and tools, chimpanzees prefer particular tools simply out of familiarity. They have no real sense of property.

With newly introduced tools, individuals will show no preference, as they are not familiar with the tools.

At Bossou, new tools were introduced in 2008 and preferences emerged quickly. This is an anecdotal report; however, it makes clear that familiarity is not the only factor driving use patterns. Further, an associative explanation is not necessarily mutually exclusive to a property explanation.

TABLE 3 Possession and property hypotheses testing using chimpanzee stone tool behavior (see Figure 1)

Hypotheses Predictions

(1) Looking and taking: length of observation and number of adults present in a party explains the occurrence of takeover events.

The likelihood of an individual taking a tool from another individual— a “takeover event” -increases with (a) time spent monitoring the individual before the “takeover” and (b) with a lower number of individuals present capable of punishing the “thief” (i.e., adults), compared to instances when takeover events do not take place following the abandonment of toolsets.

(2) The presence of possession and property norms explains the severity of punishments given, including third party punishment.

Punishment will last longer and will be more severe when the punished individual is using the “preferred tool-set” of another individual. this rule will apply even when the punishing individual is a third party.

(3) Higher reutilization of the same stone-tool set explains higher individual efficiency.

Efficiency of nut-cracking increases when individuals use a very limited number of toolsets and a negative correlation should appear between toolsets used and efficiency.

292 TIBBLE AND CARVALHO

9.2 | Test 2: Norm violation: Comparing different punishment contexts

Adults, even third party individuals, begin to punish infants as they

grow into juveniles when the juvenile attempts to “takeover” adults'

stone tools.77 This could suggest the presence of a social norm regard-

ing either possession or property. We would test the prediction that

third parties, with no preference for the stone toolset involved, will

punish individuals more severely, and for longer periods of time, in

contexts where the stone toolset is the preferred toolset of another

individual compared with stone toolsets that are preferred by no indi-

vidual. This test targets property behavior because this goes beyond a

possessor maintaining physical control over the target object. We

would need to control for two variables in this test: the ID of the pun-

isher and whether the punished individual is alone or not. The first

variable controls for individual differences in the severity of the third

party's punishment. The second variable controls for the explanation

that individuals who are found nut-cracking alone at the outdoor lab

are punished by arriving group members simply because the individual

is accessing the nuts before others. The focus of the test would be to

understand if the “status of the toolset used” can explain a significant

amount of variation in a composite measure of punishment severity

(for operational definitions, see Table 5).

9.3 | Test 3: Efficiency and toolset repertoire: Testing an adaptive explanation

If property behavior, or rather restraining the use of absent individ-

uals' tools, were to be favored in hominin evolution, there may have

been an adaptive benefit for doing so. Property behavior toward

stone tools may benefit individuals through improved knowledge and

efficient use of their particular tool or toolset, which can translate into

higher energetic gains. Property behavior may also help individuals to

avoid punishment by property holders or by group members enforcing

punishment for reneging on property norms. If property behavior

were to confer an adaptive advantage, we should expect individuals

who limit their tool repertoire, rather than using other individuals' pre-

ferred tools, to be the most efficient tool users. Therefore, a support-

ive result would be a significantly negative association between

toolset repertoire and skill (for operational definitions, see Table 6).

These three tests described above illustrate that researchers

should not be dissuaded by initial challenges of designing tests for

researching property behavior in wild species. Here, we have selected

a species and target object based on interdisciplinary findings and evi-

denced that the problem of operational definitions is surmountable. In

addition, we illustrate an empirical way to test a hypothesis for a sig-

nificant catalyst in the evolution of property: the use of stone tools by

TABLE 4 Operational definitions for Test 1 “Looking and taking: investigating observation, group presence and takeover events”

Term Definition

Departure events

An event in which an individual, who had been cracking nuts, ceased using its tools and moved away from the objects, leaving the nut-cracking area.76

Highly skilled adults

Individuals who, on average, crack nuts with three strikes or less.77 Previous research uses the number of blows to crack a nut.78

Takeover events During the subsequent 60s to departure event, another individual begins to use the abandoned tools.76

Observation An event in which a group member approaches another group member within a distance of 1 m and remains with fixed gaze upon the target individual's face or hands for 3 or more continuous seconds.79

Group presence Third party presence is established by the number of adults present at the moment of a departure event. This excludes infants and juveniles because relative to a “thief,” they present a low punishment threat.

TABLE 5 Operational definitions for Test 2 “Norm violation: comparing different punishment contexts”

Term Definition

Punishment events

A punishment event starts when an individual or groups of individuals initiate(s) an agonistic interaction with another individual.

Punishment severity

Severity multiplied by punishment length (in seconds). Severity is defined as 1 (0 contact made), 2 (1 or 2 contacts made), or 3 (3 or more contacts made).

Status of toolset The last five uses of the toolset in question are recorded. A use is defined as an individual nut-cracking with the stone tool set or one of the stone tools for 10 or more continuous seconds.

The toolset is defined as “own or unused” if: 1. The toolset has not been used before 2. OR the punished party has been the “user” for at least three of the last five uses 3. OR the toolset has been used by various individuals, with no individual using the toolset more than twice. The toolset is defined as “other's” if: 1. Another individual has used the toolset for three or more of the last five uses.

FIGURE 2 Selection of stone tools used by the Bossou chimpanzee

population for cracking nuts. (Credit: Susana Carvalho/KUPRI Japan) [Color figure can be viewed at wileyonlinelibrary.com]

TIBBLE AND CARVALHO 293

a model species. This single case study is not the answer to the evolu-

tion of property, but it is a starting point toward methodology

designed to capture natural behavior.

Like any realistic example there are limitations to this methodol-

ogy, and awareness of this fact enables us to be constructive. Pilot

studies would be carried out, to test if operational definitions being

coded successfully capture the intended behavior and thus, to avoid

the risk of type I and type II errors. For example, in the “Looking and

taking” test, there is a risk of false negative results because the opera-

tional definition for an observation event is strict; subjects may be

performing more inconspicuous observations that are not captured by

the definition. Further, follow up studies would be important. For

example, if a relationship were found between repertoire size and skill

in Test 3, this would prove an association, but not causation. Labora-

tory research could help to confirm or reject a causal relationship

between toolset repertoire and efficiency. In this way, we hope that

this realistic methodological example evidences the value of the inter-

disciplinary approach throughout the research process.

10 | CONCLUSIONS

Overall, we hope to have demonstrated that this field is at a crucial

point. Creative approaches to the study of possession and property in

nonhuman species represent a major step forward for the discipline.3

To illustrate this point, we have presented an example of how the inter-

disciplinary approach can be used to test for property behavior in a

hominin model species in the wild. Contemporary researchers have

clear terminology at their disposal and a broad range of methodologies

with which to expand knowledge of this subject. This represents a turn-

ing point in the discipline, a clear opportunity for interdisciplinary, crea-

tive research to systematically fill in the gaps of our evolutionary

knowledge of the origin of property.

Glossary

Barter In barter, an individual can trade an object or service for another which is possessed or owned by another individual.17

Creator rule Property holder status is given to the individual or individuals who created or modified the object.18

Empirical examples of this rule include individuals feeling ownership toward their own work, their organization, the products they create and their jobs.9,11,19–21

Endowment effect

The overvaluing of an object due to possession and/or property exhibited by an individual or group of individuals. For example, in trade for the same two objects, a seller may attach higher sentimental, and thus economic, value to an object in their possession and/or property compared with an identical item which is not in their possession/property.

First possession rule

Property holder status is given to the first possessor of the object as shown in developmental studies and throughout the animal kingdom.4,8,17

Loss aversion Preference to avoid losses rather than acquire gains, even when the variable of object value is held constant.

Proximity principle

Respect for possession is dependent on the physical proximity of the possessor to the object.

Rank principle Respect for possession is dependent on hierarchy and correlates with the relative ranks of the present possessor and prospective possessor.

Reference dependence bias

Individuals evaluate outcomes relative to a reference point, leading to context-specific gain and loss classification.22

Winner and loser effects

Individuals are more likely to win raids (such as those over possessions or property) at time T, based on victories at time T-1, T-2, and more likely to lose at time T, based on losing at time T-1, T-2.23

FIGURE 3 Adult chimpanzee nut-cracking, with an observing infant, at

the outdoor laboratory, Bossou (Credit: Susana Carvalho/KUPRI Japan) [Color figure can be viewed at wileyonlinelibrary.com]

TABLE 6 Operational definitions for Test 3 “Efficiency and toolset repertoire: testing an adaptive explanation”

Term Definition

Toolset repertoire

Divided into three categories: “Diverse toolset user” individuals that use more than four toolsets, “toolset user” individuals that use three or four toolsets and “limited toolset user” individuals that use only one or two toolsets.

Skill The average number of blows the individual performs to crack a nut.78 Skill is categorized into four levels: (0) 10 or more blows, (1) 7–9 blows, (2) 4–6 blows, and (3) 1–3 blows. The higher the category the more skilled the individual.78

294 TIBBLE AND CARVALHO

ACKNOWLEDGMENTS

L. Tibble thanks Susana Carvalho for her encouragement and support

in pursuing this research and William C. McGrew for helpful com-

ments and suggestions. L. Tibble is grateful for the guidance on inter-

disciplinary approaches received from Emma Cohen and Claire El

Mouden, at the Institute of Cognitive and Evolutionary Anthropology,

Oxford. L. Tibble and S. Carvalho are sincerely grateful to the three

anonymous reviewers for their exemplary and extremely useful guid-

ance and constructive suggestions. S. Carvalho thanks the Leverhulme

Trust for the Prize (PLP-2016-114) that is providing the funds and

time to advance research in 2016–2018.

ORCID

Susana Carvalho https://orcid.org/0000-0003-4542-3720

REFERENCES

[1] Ellis L. 1985. On the rudiments of possessions and property. Soc Sci Inf 24:113–143.

[2] Maynard Smith J. 1982. Evolution and the theory of games, Cam- bridge: Cambridge University Press.

[3] Hare D, Reeve HK, Blossey B. 2016. Evolutionary routes to stable ownership. J Evol Biol 29:1178–1188.

[4] Stake JE. 2004. The property “instinct”. Philos Trans R Soc Lond Ser B Biol Sci 359:1763–1774.

[5] Sherratt TN, Mesterton-Gibbons M. 2015. The evolution of respect for property. J Evol Biol 28:1185–1202.

[6] Gintis H. 2007. The evolution of private property. J Econ Behav Organ 64:1–16.

[7] Kummer H, Cords M. 1991. Cues of ownership in long-tailed macaques, Macaca fascicularis. Anim Behav 42:529–549.

[8] Friedman O. 2008. First possession: An assumption guiding infer- ences about who owns what. Psychon Bull Rev 15:290–295.

[9] Pierce JL, Kostova T, Dirks KT. 2003. The state of psychological own- ership: Integrating and extending a century of research. Rev Gen Psy- chol 7:84–107.

[10] Kummer H. 1991. Evolutionary transformations of possessive behav- ior. J Soc Behav Pers 6:75–83.

[11] Beaglehole E. 1932. Property: A study in social psychology, New York: Macmillan.

[12] Itakura S. 1992. Symbolic association between individuals and objects by a chimpanzee as an initiation of ownership. Psychol Rep 70:539–544.

[13] Itakura S. 1994. Symbolic representation of possession in a chimpanzee. In: Taylor Parker S, Mitchell RW, Boccia ML, editors. Self-awareness in animals and humans, Cambridge: Cambridge University Press.

[14] Kanngiesser P, Santos LR, Hood BM, Call J. 2011. The limits of endowment effects in great apes (Pan paniscus, Pan troglodytes, Gorilla gorilla, Pongo pygmaeus). Int J Comp Psychol 125:436–445.

[15] Drayton LA, Brosnan SF, Carrigan J, Stoinski TS. 2013. Endowment effects in gorillas (Gorilla gorilla). J Comp Psychol 127:365–369.

[16] Flemming TE, Jones OD, Stoinski TS, Mayo L, Brosnan SF. 2012. The endowment effect in orangutans. Int J Comp Psychol 25:285–298.

[17] Brosnan SF. 2011. Property in non-human primates. New Dir Child Adolesc Dev 132:9–22.

[18] Rochat P, Robbins E, Passos-Ferreira C, Oliva AD, Dias MD, Guo L. 2014. Ownership reasoning in children across cultures. Cognition 132:471–484.

[19] Dirks KT, Cummings LL, Pierce JL. 1996. Psychological ownership in organizations: Conditions under which individuals promote and resist change. In: Woodman RW, Pasmore WA, editors. Research in organiza- tional change and development. vol 9, Greenwich, CT: JAI Press. p 1–23.

[20] Das G. 1993. Local memoirs of a global manager. Harv Bus Rev 71:38–47. [21] Peters T, Austin N. 1985. A passion for excellence, New York: Ran-

dom House.

[22] Kahneman D, Tversky A. 1979. Prospect theory: An analysis of deci- sion under risk. Econometrica 47:263–291.

[23] Dugatkin LA, Reeve HK. 2014. Winning, losing, and reaching out. Behav Ecol 25:675–679.

[24] Searle JR. 2010. Making the social world, Oxford: Oxford University Press.

[25] Tinbergen N. 1963. On aims and methods of ethology. Z Tierpsychol 20:410–433.

[26] Friedman D. 1998. On economic applications of evolutionary game theory. J Evol Econ 8:15–43.

[27] Morrell LJ, Kokko H. 2003. Adaptive strategies of territory formation. Behav Ecol Sociobiol 54:385–395.

[28] Grafen A. 1987. The logic of divisively asymmetric contests: Respect for ownership and the desperado effect. Anim Behav 35:462–467.

[29] Hinsch M, Komdeur J. 2017. Punish the thief – Coevolution of defense and cautiousness stabilizes ownership. Behav Ecol Sociobiol 71:102.

[30] DeScioli P, Wilson BJ. 2011. The territorial foundations of human property. Evol Hum Behav 32:297–304.

[31] Dionisio F, Gordo I. 2007. Controlling excludability in the evolution of cooperation. Evol Ecol Res 9:365–373.

[32] Rankin DJ, Bargum K, Kokko H. 2007. The tragedy of the commons in evolutionary biology. Trends Ecol Evol 22:643–651.

[33] Hardin G. 1968. The tragedy of the commons. Science 162:1243–1248. [34] Ostrom E, Gardner R, Walker J. 1994. Rules, games, and common-pool

resources, Ann Arbor, MI: University of Michigan Press. [35] Reeve HK, Emlen ST, Keller L. 1998. Reproductive sharing in animal

societies: Reproductive incentives or incomplete control by dominant breeders. Behav Ecol 9:267–278.

[36] Hammerstein P, Parker GA. 1982. The asymmetric war of attrition. J Theor Biol 96:647–682.

[37] Kummer H, Götz W, Angst W. 1974. Triadic differentiation: An inhib- itory process protecting pair bonds in baboons. Behaviour 49:62–87.

[38] Brosnan SF, Beran MJ. 2009. Bartering behavior between conspe- cifics in chimpanzees, Pan troglodytes. J Comp Psychol 123:181–194.

[39] Brosnan SF, Jones OD, Mareno MC, Richardson AS, Lambeth SP, Schapiro SJ. 2007. Endowment effects in chimpanzees. Curr Biol 17:1–4.

[40] Perry S. 1997. Male-female social relationships in wild white-faced capuchin monkeys, Cebus capucinus. Behavior 134:477–510.

[41] Russ BE, Comins JA, Smith R, Hauser MD. 2010. Recognizing and respecting claims over resources in free-ranging rhesus monkeys, Macaca mulatta. Anim Behav 80:563–569.

[42] Bachmann G, Kummer H. 1980. Male assessment of female choice in hamadryas baboons. Behav Ecol Sociobiol 6:315–321.

[43] Sigg H, Falett J. 1985. Experiments on respect of possession in hama- dryas baboons (Papio hamadryas). Anim Behav 33:978–984.

[44] Thierry B, Wunderlich D, Gueth G. 1989. Possession and transfer of objects in a group of brown capuchins (Cebus apella). Behavior 110: 294–305.

[45] Popp and DeVore. 1979. Hamburg D, ER MC, editors. Aggressive competition and social dominance theory: synopsis. In D. A. Hamburg & E.R. McCown (eds.), The great apes. pp 316-338. Menlo Park, CA: Benjamin/Cummings.

[46] Dowds BM, Elwood RW. 1983. Shell wars: Assessment strategies and the timing of decisions in hermit crab shell fights. Behaviour 85:1–24.

[47] Lefebvre L. 1982. Food exchange strategies in an infant chimpanzee. J Hum Evol 11:195–204.

[48] Lefebvre L, Hewitt TA. 1986. Food exchange in captive chimpanzees. In: Taub DM, King FA, editors. Current perspectives in primate social dynamics, New York: Van Nostrand Reinhold.

[49] Brosnan SF, de Waal FBM. 2004. A concept of value during experi- mental exchange in brown capuchin monkeys. Folia Primatol 75: 317–330.

[50] Lakshminarayanan V, Chen MK, Santos LR. 2008. Endowment effect in capuchin monkeys. Philos Trans R Soc Lond Ser B Biol Sci 363: 3837–3844.

[51] Ross H, Friedman O, Field A. 2014. Toddlers assert and acknowledge ownership rights. Soc Dev 24:341–356.

[52] Ardrey R. 1966. The territorial imperative: A personal inquiry into the animal origins of property and nations, New York: Dells.

TIBBLE AND CARVALHO 295

[53] Burk C. 1900. The collecting instinct. Pedagog Semin J Genet Psychol 7:179–207.

[54] Darling FF. 1937. A herd of red deer, London: Oxford University Press. [55] Hall G, Wiltse S. 1891. Children's collections. Pedagog Semin J Genet

Psychol 1:234–237. [56] Kline L, France CJ. 1899. The psychology of mine. Pedagog Semin J

Genet Psychol 6:421–470. [57] McDougall W. 1908. An introduction to social psychology, London:

Methuen. [58] Porteous JD. 1976. Home: The territorial core. Geogr Rev 66:

383–390. [59] Weil S. 1952. The need for roots: Prelude to a declaration of duties

towards mankind, London: Routledge and Kegan Paul. [60] Brown DE. 1991. Human universals, New York: McGraw Hill. [61] Heine. 1997. Possession: Cognitive sources, forces and Grammaticali-

zation, Cambridge: Cambridge University Press. [62] Blumenthal JA. 2010. Property law: A cognitive turn. Psychon Bull

17:186–191. [63] Friedman O, Neary KR. 2008. Determining who owns what: Do chil-

dren infer ownership from first possession. Cognition 107:829–849. [64] Gergely G, Csibra G. 2005. The social construction of the cultural

mind: Imitative learning as a mechanism of human pedagogy. Interact Stud 6:463–481.

[65] Tomasello M. 2011. Human culture in evolutionary perspective. In: Gelfand M, editor. Advances in culture and psychology, Oxford: Oxford University Press.

[66] Dittmar H. 1992. The social psychology of material possessions: To have is to be, New York: St. Martin's Press.

[67] Rochat P. 2010. Possession and morality in early development. New Dir Child Dev 132:23–38.

[68] Rochat P. 2009. Others in mind – The social origin of self-consciousness, Cambridge: Cambridge University Press.

[69] Taylor C. 1989. Sources of the self: The making of modern identity, Cambridge: Harvard University Press.

[70] Apicella CL, Azevedo EM, Christakis NA, Fowler JH. 2014. Evolution- ary origins of the endowment effect: Evidence from hunter-gatherers. Am Econ Rev 104:1793–1805.

[71] Kanngiesser P, Rossano F, Tomasello M. 2015. Late emergence of the first possession heuristic: Evidence from a small-scale culture. Child Dev 86:1282–1289.

[72] Byrnit JT, Høgh-Olesen H, Makransky G. 2015. Share your sweets: Chimpanzee (Pan troglodytes) and bonobo (Pan paniscus) willingness to share highly attractive, monopolizable food sources. J Comp Psy- chol 129:218–228.

[73] Koops K, Furuichi T, Hashimoto C. 2015. Chimpanzees and bonobos differ in intrinsic motivation for tool use. Sci Rep 5:11356.

[74] Matsuzawa T. 1994. Field experiments on use of stone tools by chim- panzees in the wild. In: Wrangham RW, WC MG, de Waal FBM, Heltne PG, editors. Chimpanzee Cultures, Cambridge: Harvard Uni- versity Press. p 351–370.

[75] Carvalho S, Biro D, McGrew WC, Matsuzawa T. 2009. Tool-composite reuse in wild chimpanzees (Pan troglodytes): Archaeo- logically invisible steps in the technological evolution of early homi- nins. Anim Cogn 12:103–114.

[76] Biro D, Carvalho S, Gwyther A, Matsuzawa T. In review. Tool “recy- cling”: Reuse of adults' tools increases nut-cracking efficiency in wild juvenile chimpanzees.

[77] Biro D, Sousa C, Matsuzawa T. 2006. Ontogeny and cultural propaga- tion of tool use by wild chimpanzees at Bossou, Guinea: Case studies in nut cracking and leaf folding. In: Matsuzawa T, Tomonaga M, Tanaka M, editors. Cognitive development in chimpanzees, Tokyo: Springer Press. p 476–508.

[78] Carvalho S, Marques R, Gama Mota P, Biro D, Braun DR, Matsuzawa T. 2016. Skilled individuals drive social learning in wild chimpanzee tool use: Implications for the Pleistocene archaeological record. Paleoanthropology Society Meeting Abstracts, Atlanta, GAA7. PaleoAnthropology Society. https://doi.org/10.4207/PA.2016.ABS14.

[79] Biro D, Inoue-Nakamura N, Tonooka R, Yamakoshi G, Sousa C, Matsuzawa T. 2003. Cultural innovation and transmission of tool use in wild chimpanzees: Evidence from field experiments. Anim Cogn 6: 213–223.

AUTHOR BIOGRAPHIES

Lucy M. Tibble is a recent graduate from the Masters program in

Cognitive and Evolutionary Anthropology at the University of Oxford.

Her previous undergraduate research in the department of Biological

Anthropology, at the University of Cambridge, involved studying

bushbaby behavior both in the field and remotely. Her graduate

research at the Institute of Cognitive and Evolutionary Anthropology,

Oxford, focused on applying the primate modeling approach to the

evolution of property behavior.

Susana Carvalho is an Associate Professor in the Institute of Cogni-

tive and Evolutionary Anthropology and an Associate Director for

Palaeoanthropology and Primatology at the Gorongosa National Park,

Mozambique where she currently coordinates the “Paleo-Primate Pro-

ject Gorongosa.” Her past research includes foundational work in the

novel field of Primate Archeology, and her new lab “Primate Models

for Behavioral Evolution” investigates the evolution of human behav-

ior using a variety of extant primates as models

How to cite this article: Tibble L, Carvalho S. Rethinking the

evolution of property and possession: A review and methodo-

logical proposition. Evol Anthropol. 2018;27:285–296. https://

doi.org/10.1002/evan.21748

296 TIBBLE AND CARVALHO