Maternal behaviors that mediate skill development in Sumatran orangutans

Abstract

Although orangutans are closely related to humans, very little is known about their ontogenetic development. In particular, there is a lack of systematic research on the maternal behaviors that mediate skill development in early infancy. To address this topic, we conducted a longitudinal study in which a Sumatran orangutan (Pongo abelii) mother-infant dyad was systematically observed across 28 months, starting with the infant's birth. Our data revealed several classes of maternal behavior that potentially influenced infant skill development. The timing of these behaviors was contingent upon infant competence level, as active interventions were intense during periods of skill acquisition. The same behaviors were flexibly deployed independent of whether the infant was in the process of acquiring foraging, locomotor or social skills. Our findings suggest that the maternal behaviors that mediate infant skill development in Sumatran orangutans have features reminiscent of human scaffolding, and raise questions about intentionality in such behaviors.

Full text

UC Merced
Proceedings of the Annual Meeting of the Cognitive Science
Society
Title
Maternal behaviors that mediate skill development in Sumatran orangutans
Permalink
https://escholarship.org/uc/item/35h6z1hg
Journal
Proceedings of the Annual Meeting of the Cognitive Science Society, 43(43)
ISSN
1069-7977
Authors
Sauciuc, Gabriela-Alina
Luna, Adriana
Wester, Anna Zara Louise
et al.
Publication Date
2021
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California

Maternal behaviors that mediate skill development in Sumatran orangutans
Gabriela-Alina Sauciuc (gabriela-alina.sauciuc@lucs.lu.se)
Lund University, Department of Philosophy, Cognitive Science
Box 192, 221 00, Lund
Adriana Luna Martinez (adriana.luna@estudiante.uam.es)
Lund University, Department of Philosophy, Cognitive Science
Box 192, 221 00, Lund
Anna Zara Louise Wester (alzwester@gmail.com)
Radboud University, Nijmegen, Netherlands
Tora Hellgren (7555he-s@student.lu.se)
Lund University, Department of Philosophy, Cognitive Science
Box 192, 221 00, Lund
Tomas Persson (tomas.persson@lucs.lu.se)
Lund University, Department of Philosophy, Cognitive Science
Box 192, 221 00, Lund
Abstract
Although orangutans are closely related to humans, very little
is known about their ontogenetic development. In particular,
there is a lack of systematic research on the maternal behaviors
that mediate skill development in early infancy. To address this
topic, we conducted a longitudinal study in which a Sumatran
orangutan (Pongo abelii) mother-infant dyad was
systematically observed across 28 months, starting with the
infant’s birth. Our data revealed several classes of maternal
behavior that potentially influenced infant skill development.
The timing of these behaviors was contingent upon infant
competence level, as active interventions were intense during
periods of skill acquisition. The same behaviors were flexibly
deployed independent of whether the infant was in the process
of acquiring foraging, locomotor or social skills. Our findings
suggest that the maternal behaviors that mediate infant skill
development in Sumatran orangutans have features
reminiscent of human scaffolding, and raise questions about
intentionality in such behaviors.
Keywords: orangutan ontogeny, skill acquisition, scaffolding,
teaching
Introduction
Orangutans are among our closest genetic relatives, sharing
96,4% of our genetic makeup. Surprisingly, in spite of this
close relatedness to humans, very little is known about their
ontogenetic development. In particular, research on skill
development in Sumatran orangutans (Pongo abelii) during
early infancy is lacking. It is therefore unknown how
orangutans’ journey towards becoming skillful foragers and
competent social individuals begins. In the most informative
study to date (Chevalier-Skolnikoff, 1983), four orangutan
infants were periodically assessed using behavioral
parameters derived from Piaget’s sensorimotor stages for
human ontogenetic development (Piaget, 1952). In humans,
these stages include skills that are acquired between birth and
two years of age through trial and error, and range from
reflexive motor activity (first stage, at one month) to insight-
based problem solving (sixth stage, at 24 months). In this
study, orangutans achieved the first piagetian stages earlier
than humans, but lagged at reaching the later stages. Stage
five behaviors, which involve exploration through relational
experimentation (e.g., flailing ropes, tossing objects in the
air), were first displayed at 12 months of age, while the sixth
stage was completed between five-eight years of age.
Additional observations come from less systematic reports
of orangutan development in early infancy, which mention
first attempts at hanging, standing or back-riding (seemingly
assisted by the mother), food manipulation, first contact
breaking, infant genital exploration by the mother, instances
of social play, and solitary play (e.g., Maple, Wilson, Zucker
et al., 1978; Miller & Nadler, 1981).
A first limitation of some of these studies is that there is no
distinction between orangutan species. Yet this is highly
relevant, since recent research documents significant socio-
cognitive differences between Bornean (Pongo pygmaeus)
and Sumatran (Pongo abelii) orangutans, which seem to be
related to different sociality styles (Forss, Willem, Call et al.,
2016). As such, Sumatran orangutans are more gregarious
and socially tolerant than their Bornean counterparts, show
superior problem-solving skills (Forss et al., 2016), and have
broader repertoires of culturally transmitted behaviors, which
encompass both social and ecological skills (e.g., Schuppli,
1187

Meulman, Forss et al., 2016). It is therefore imperative to
study the early development of skills separately for each of
the orangutan species.
Additional limitations derive from the impoverished
housing conditions of the involved individuals. Rearing
conditions have a major impact on the socio-emotional and
cognitive development of primates (e.g., Maestripieri, 2018).
Continuous contact with the mother - so that the physical and
social needs of the infants are appropriately attended to -
nurtures the development of social and communicative skills,
and ensures optimal cognitive, social and emotional
outcomes. In contrast, social deprivation has a tremendously
negative impact on all these developmental outcomes. As
such, only tentative conclusions can be drawn from studies
on nursery-raised individuals, such as, e.g., the study on
orangutan sensorimotor development by Chevalier-
Skolnikoff (1983). In addition, restrictive housing conditions
(as in Maple et al., 1978; Miller & Nadler, 1981) negatively
impact the normal development of young primates. For
example, in the study by Miller and Nadler (1981), the
studied individuals were housed in 2,4 m high cages with an
area of 5,76 m2 indoors and 9,6 m2 outdoors. This drastically
limits the expression of species-specific behaviors by the
mother, thus limiting the repertoire of behaviors and skills
that the infant is exposed to, and has the opportunity to
acquire through social learning. Restrictive housing
conditions are also likely to affect maternal physical and
psychological well-being, leading to increased levels of stress
and depression. In turn, insufficient maternal welfare
negatively impacts infant socio-emotional and socio-
cognitive development (e.g., Maestripieri, 2018). More
recent studies, especially if conducted in the wild, bypass
these issues. However, these studies usually do not include
the early stages of infancy (e.g., Bard, 1995; Schuppli et al.,
2016). If young infants are included, Sumatran orangutans
are not represented, or data recorded from infants is minimal
(e.g., Scheider, Call & Liebal, 2012).
In this paper, we present a longitudinal study of skill
development in the early infancy (0-28 months of age) of a
Sumatran orangutan female, in which we focus on maternal
behaviors that mediate the maturation of locomotor, foraging
and social skills. In particular, our aims are (1) to report the
findings of a systematic qualitative analysis of such
behaviors, as documented in over 300 hours of observations
of the mother-infant dyad, and (2) to outline the details of a
method for the systematic investigation of such behaviors.
There is widespread agreement that, in all primate species,
the mothers constitute the most important source of influence
on infant development, and that such influence is exerted on
several levels: physiological, reproductive, behavioral,
emotional, cognitive and social (for recent reviews see
Maestripieri, 2018; Whiten & de Waal, 2018). This is
especially the case for Sumatran orangutans, who display the
longest period of maternal dependency in the animal
kingdom (of up to 10 years), as well as the longest interbirth
interval of all mammalian species, since females with
offspring do not reproduce (van Noordwijk & van Schaik,
2005). Moreover, given the semi-solitary nature of
orangutans, developmental influences from other
conspecifics than the mother are nearly absent during
orangutan early infancy, although, starting with three years
of age, immatures begin to increasingly socialize with others
(Schuppli et al., 2016).
Available data suggests that, in nonhuman apes, maternal
influences on skill development are passively exerted, with
the offspring being primarily responsible for skill acquisition.
For great apes, this relationship has been described using an
apprenticeship analogy, whereby the novice’s presence (the
immature offspring) is tolerated by the skilled master (the
mother), and the novice is motivated to acquire the skill
through observation learning (Matsuzawa, 2001). By keeping
close proximity to the mother, immatures are provided with
learning opportunities by mere exposure to species-
appropriate actions in social and non-social contexts. As
such, infants will explore and eventually consume the food
items consumed by their mothers, thus acquiring the species-
appropriate foraging repertoire, while also learning the
location of food sources (e.g., Schuppli et al., 2016). By
watching their mothers interact with conspecifics, infants will
acquire their social skills (e.g., Maestripieri, 2018). The
cognitive mechanisms implicated in such learning processes
are likely to be stimulus and local enhancement (e.g., More,
2013; Whiten & de Waal, 2018). In great apes, who
frequently engage in complex forms of food processing, more
advanced mechanisms such as emulation or even imitation
may be also involved (as reviewed by Moore, 2013; Whiten
& de Waal, 2018). To date, imitation learning in great apes is
primarily documented in juveniles, although Russon and
Galdikas (1995) also report a small number of occurrences in
younger Bornean orangutans starting with three years of age.
Note that an active maternal involvement in this learning
process is not reported in any of these cases.
The literature, however, is not completely devoid of
examples in which primate mothers appear to take a more
active role in shaping infant skill acquisition. Such examples
are reminiscent of interventions that in the human literature
are regarded as scaffolding, whereby the expert (e.g., the
mother) engages in activities that mediate skill acquisition in
the pupil. An important feature of scaffolding is that skill
acquisition should take place in the so-called zone of
proximal development of the pupil, meaning that the expert
needs to be sensitive to the knowledge level of the pupil.
Another relevant feature is that such facilitation (the
‘scaffold’) is gradually removed by the expert, as the pupil
becomes more proficient (e.g., Palincsar, 1986).
In the animal literature, such scaffolding behaviors have
been described as a form of functional teaching. This notion
is meant to capture the functional consequences of facilitating
learning in others, as opposed to intentional teaching, which
is based on an explicit intention to cause an individual to learn
1188

(e.g., Caro & Hauser, 1992). In functional teaching, a skilled
individual modifies its behavior in the presence of novices,
thereby actively facilitating learning in the novices, i.e. the
latter learns a skill earlier, more rapidly or efficiently than in
the absence of the expert’s behavior, or even bypasses the risk
of not learning that skill at all (Caro & Hauser, 1992).
Examples of functional teaching have been described for a
number of species, including ants, birds, and a number of
mammalian taxa, such as felids, meerkats, cetaceans,
pinnipeds, and primates. In primates, the most prominent
examples of functional teaching include maternal
encouragement and discouragement, as well as tool transfer
in chimpanzees (as reviewed, e.g., by Maestripieri, 2018;
Moore, 2013; Whiten & de Waal, 2018).
Methods
Site and participants
Data collection was conducted at Lund University Primate
Research Station Furuvik, where a mother-infant dyad
(Dunja-Indah) was observed for 28 consecutive months,
starting with the infant’s birth, in November 2017 and ending
with March 2020, when travel to the research station was
restricted due to the pandemic. Dunja was 25 years old at
parturition, and Indah is her firstborn. The mother-infant
dyad is housed in a semi-naturalistic enclosure that consists
of an indoor area and an outdoor area. The indoor enclosure
has natural flooring, and is enriched with logs, platforms,
hammocks, and ropes. Adjacent to it, there are three smaller
rooms, which have been built for research purposes. Glass
walls separate the indoor enclosure from the visitor area,
which runs as a long corridor parallel to the enclosure. The
outdoor area consists of two islands, which are cultivated
with edible plants. On the islands, there are also wooden
platforms, logs, trees and bushes. In the outdoors, the visitors
can watch the orangutans from several wooden platforms that
are built around the enclosure.
The mother-infant dyad shares this enclosure with the male
that sired Indah (Naong), who was 26 years old at Indah’s
birth. Naong was separated from Dunja a few weeks prior to
parturition, and joined the dyad about four weeks after
Indah’s birth. During this separation time, the orangutans had
visual access to each other. Occasionally, they could also
come in close proximity, which enabled tactile interactions
through the mesh. The orangutans are fed a varied diet,
composed of vegetables, fruits, seeds, nuts, vitamin-enriched
pellets and various protein sources. To limit the intake of
soluble sugars, fruit is primarily provided as reinforcement
during husbandry training and experimental research. Several
food items are provided on a daily basis as part of various
enrichment routines, based on a weekly updated schedule.
Data collection
Data collection relied on continuous observation, with
systematic video-recording of mother-infant activities for
subsequent micro-analysis. The observations were taken
primarily from the visitor areas, and occasionally from
restricted areas. An all-occurrence sampling approach was
used, which targeted the dyad’s activities. Written notes were
also taken if activity onset was sudden, or if relevant activities
were in progress at the start of a session of data collection,
before the filming equipment was set up. Whenever the dyad
was inactive or out of sight (hidden behind a log, under a
blanket, up in a hammock, etc.), data collection was
complemented with five-minute scan sampling.
The observation schedule varied with the infant’s age,
since the level of activity and the predictability of the
orangutans’ activity budget varied as a function of age. Thus,
the observation schedule was very intense during the first
month of the infant’s life, amounting to six hours daily for at
least 24 hours / week. During this month, there was a high
risk of missing relevant behaviors, as the orangutans’ activity
level was low, and the occurrence of activity periods highly
unpredictable. Toward the end of the first month, however,
activity periods had become predictable, and relevant
behaviors re-occurred with sufficient frequency to justify a
reduced observation schedule. From now on, the dyad was
followed for a minimum of two hours / weekly, thus totaling
337 hours and 27 minutes of observation for the overall 28-
month period of data collection.
Data analysis
Given the absence of previous systematic research on
maternal behaviors that support skill development in
orangutan infants, the first stage of data analysis consisted of
systematic qualitative analysis. The filmed material was
screened in its entirety, to extract suitable themes and
categories of behavior. In order to study the kind of maternal
responses (henceforth maternal behaviors) that influence
skill development, it was important to also record the infant’s
activities that were likely to trigger a maternal response. To
determine the immediate effects of maternal behaviors on the
infant, we also examined the infant’s response whenever an
active maternal behavior was recorded.
To follow up with a quantitative treatment of target
behaviors (i.e., maternal behaviors, infant responses), the
filmed material is annotated using the video-annotation
software ELAN. To this end, relevant tiers have been set up
in ELAN, including Indah’s activity, Dunja’s activity,
maternal behavior, infant response. Annotations
corresponding to these categories are entered in the
respective tiers whenever relevant behaviors are identified on
film. Indah’s activity is meant to capture infant actions that
could elicit a maternal response, while Dunja’s activity is
meant to capture the actions carried out by the mother
concomitant with the activities of the infant, and, thus, prior
to a maternal behavior. The categories annotated as activities
are based on commonly used ethograms and include: Rest
seated, Rest lying / sleeping, Eat seated / lying, Forage,
Nursing, Travel horizontal, Travel vertical, Nesting, Solitary
1189

locomotor (play), Social play, Agonistic interaction (with
Naong), Grooming, Object manipulation, Oral exploration,
Other social interaction, Other activity.
Each annotated episode is also recorded in a spreadsheet,
where we also record information about the date and duration
of each behavior, and codify relevant variables, such as
proximity, gaze direction and timing of facial expressions.
Findings
In this paper, we present the main findings of the qualitative
analysis, in particular an inventory of maternal behaviors that
have the potential to influence skill development in orangutan
infants, as well as an inventory of infant responses triggered
by active forms of maternal behavior. We also delineate the
longitudinal progression of target behaviors, and discuss a
number of examples. Finally, we also present preliminary
results from the quantitative analysis of maternal behaviors
targeting infant object-directed actions.
Maternal behaviors
Based on the qualitative data analysis carried out for the
purposes of this study, maternal behaviors with a potential to
mediate infant skill acquisition were classified in the
following categories: tolerate, interference, approach,
ignore, molding, take-over, co-action, social visits.
Tolerate comprises a class of situations in which the
infant’s activity is carried out within the mother’s visual field,
so that the mother can observe it, but the mother does nothing
to respond to it, other than monitoring the infant’s activity.
This behavior was recorded in response to a great variety of
actions performed by the infant, including locomotion,
foraging, object manipulation, social interaction, etc.
Developmentally, tolerate was often preceded by
interference, and succeeded by ignore. In this category, we
have also included a number of examples that in previous
anecdotal observations have been presented in more active
terms, as encouraging or supporting infant locomotion. For
example, we observed that the infant used the mother’s
stretched arm as a support during her first attempts to stand,
or used the mother’s body as a ladder or bridge for climbing
or crossing a gap. We classified these examples as tolerate,
since it was not possible to ascertain that the mother was
actively supporting the infant’s actions.
Interference refers to situations in which the mother
intervenes, and stops the infant’s activity within the first 30
seconds of infant engagement in that activity. For example,
Dunja prevents Indah to continue climbing on the mesh, by
removing her from the mesh, or prevents Indah to put objects
(bark, food, toys) into her mouth, by removing them from her
hand or mouth. This type of intervention (also called
discouragement) is considered to be the most common
expression of active maternal influence on infant skill
development (e.g., Maestripieri, 2018). Our data suggests a
developmental progression with respect to interference
targets. Initially, interference seemingly targeted resting, i.e.
the mother actively intervened to activate her infant after a
period of rest. Once the infant was able to grasp, and achieved
hand-mouth coordination, interference was frequently
observed when the infant attempted to explore objects,
including food items, bark pieces, small containers, blankets,
sheets, etc. Interference that targeted object exploration was
intensively documented during a short period of time (of
several weeks), after which, in these very same situations, the
dominating response was tolerate. Interference targeting
locomotor attempts (climbing, walking away) was
manifested for a longer period of time (of several months). A
final target of intensive interference documented in this
study was social interaction, in particular attempts to
socialize with the male Naong.
Approach captures situations in which the activity of the
infant triggers the mother to approach from some distance,
and to monitor the infant’s activity, but without any further
intervention. It can be argued that, in such situations, the
mother’s interest acts to increase the salience of the activity
carried out by the infant, thereby reinforcing it.
Ignore includes situations in which the mother does not
appear to monitor the infant’s activity in any way. This also
includes situations in which the infant’s activity is carried out
in a location that is not within the mother’s visual field, and
the mother does not make any effort to gain visual access to
the activity in which the infant is engaged. Ignore was
frequent in the second year of the infant’s life, when the
infant achieved good proficiency at foraging, feeding,
locomoting, and manipulating certain objects. The last
category of infant activity to be responded to with ignore, at
about two years of age, was engagement in social interactions
with Naong.
Molding refers to situations in which the mother places the
infant in a specific position or location. Molding has been
previously described in the literature, and anecdotal
observations have been reported in chimpanzees. For
example, the sign-language trained chimpanzee Washoe was
observed to mold her foster infant’s hand as to sign FOOD
(Fouts, Fouts & van Cantfort, 1989). In our data, molding was
represented by situations in which the mother actively placed
the infant into a position that facilitated mutual gaze (a
behavior described in chimpanzees by Bard, Myowa-
Yamakoshi, Tomonaga et al., 2005), placed the infant in a
nursing position, attempted to place the infant in a back-
riding position, or placed the infant on the ground in the
context of independence training. Mutual gaze facilitation
was observed during the first month of Indah’s life only, with
mutual gaze being achieved by the infant within this period.
The first back-riding attempt was first documented when
Indah was two-month old, and then again at five, six and 16
months of age. All these attempts were unsuccessful.
Independence training was serendipitously captured across
two consecutive days, when Indah was nine-month old. The
first occurrence was documented on an outdoor island, where
Dunja was seen placing Indah on the ground, among tall
1190

grasses, and moving away from her while keeping eye
contact. Indah reached to her mother, and clung onto her.
Dunja, again, placed Indah on the ground, then slowly backed
away, while keeping eye contact with Indah. Shortly after she
returned to Indah. On the following day, in the indoor
enclosure, the dyad sat in close proximity when Dunja slowly
pushed Indah away. A gap of several meters was created
between the two, before Dunja returned to Indah and patted
her with the back of the hand. Dunja moved away from Indah
once more, but Indah no longer seemed to react to the
separation.
Take-over captures situations in which the mother
interrupts the infant’s activity, and engages herself in that
activity. Situations that prompted take-over were situations
in which the infant manipulated, e.g., a food item, which was
then removed by the mother, who consumed it. Based on
current knowledge and data, it is difficult to determine if such
behavior could be interpreted as a scaffolding behavior, or if
the mother behaves selfishly. Either way, the behavior has
stimulus enhancing effects, and is, thus, likely to contribute
to skill development in the infant.
Co-action refers to situations in which the mother and the
infant are simultaneously or sequentially (i.e., taking turns)
engaged in the same activity or act on the same or similar
material. Here, we present two examples from our data, in
which, during co-action, the mother performed a more
skillful version of the infant’s activity, which, in turn,
resulted in apparent behavior copying by the infant. The first
case occurred when Indah was seven-month old, and several
structures in the enclosure had been baited with porridge as
part of an enrichment activity. Traveling around to lick, and
dip for porridge was a novel experience for the infant. After
foraging for porridge with the mouth or fingers at several
stations, the mother-infant dyad stopped in the vicinity of a
rock formation that was also baited. The travel was initiated
from a nearby platform, where they foraged using mouth and
fingers. Before initiating travel to the ground, Dunja collected
an approximately 25 cm long stick from the platform. Upon
arriving at the rock formation, Dunja slightly modified one
end of the stick, and two larger splinters of about 8 cm each
fell on her chest, within Indah’s reach. Dunja glanced briefly
to Indah as she picked one of the splinters with her lips and
the other with her hand. Subsequently, Dunja turned her gaze
to the porridge and leaned forward touching the rock with the
stick she held between her lips. Indah watched her mother’s
action, and almost touched the rock with the splinter she
carried in her hand. As her mother leaned back, Indah’s
attention moved to the splinter. As Dunja inserted the other
end of the stick into her mouth, Indah did the same with her
splinter. Dunja bit off a splinter for herself, and, as she placed
the larger stick on the ground looked toward Indah who was
intently looking at her own splinter and the porridge.
Subsequently, Dunja leaned forward to collect porridge,
causing Indah to also lean forward and attend to the foraging
process. Indah too attempted to touch the rock with her own
splinter, but she lost the tool. Dunja glanced briefly toward
Indah then to the baited crevice, released her own splinter and
foraged with the lips. Indah watched intently, then both
continued foraging with the lips, yet on different, but nearby
located spots. Dunja, once more, dipped her splinter into a
crevice on Indah’s side, while Indah watched intently. As
Dunja leaned back, Indah (who was attached in ventrolateral
position), leaned forward, collected her lost splinter and
placed it between the lips. Dunja glanced at Indah during this
time, then switched to monitoring the male (Naong), who was
foraging at another station in the vicinity. Indah turned
around, and looked away in the same direction as Dunja.
Dunja dipped her splinter into the porridge pool once more,
but Indah no longer showed interest in foraging. Within
seconds, they traveled away from the baited station. Although
tool use is prominent in the episode described above, it has to
be noted that Dunja is typically not a tool user. She
commonly approaches this type of enrichment by using her
fingers or her lips to the bait (porridge, honey, peanut butter,
etc.).
The second case occurred when Indah was 12-month old,
and she found herself, again, in a novel situation. It was the
first fall day when the outdoors temperature had fallen below
the freezing point. Frost in the outdoors occasionally causes
one of the windows in the enclosure to be covered with a thin
sheet of ice. Dunja (carrying Indah) approached the frosty
window, and collected ice, either by scraping it with her
fingers or by licking it. Indah watched Dunja intently for
nearly two minutes, after which, when Dunja stopped,
approached the window and attempted to scrape ice using her
fingers. After a while, Dunja began watching Indah’s
attempts, and briefly interrupted her scraping actions to lick
the ice collected on Indah’s nails. They left shortly after, but
Indah returned alone about ten minutes later, and began
foraging ice from the window. Rather than scraping the
window without interruption as previously, she would now
stop to lick the ice collected on her nails after a couple of
scraping movements.
Social visits. Shortly after being reunited with Naong,
Dunja (carrying Indah) began making short visits to wherever
Naong was located at that time of the day. There was no
obvious trigger to these visits, and they did not happen on a
precise schedule. Once in proximity to Naong, Dunja would
sit close to him for several minutes. Body contact between
Naong and Indah was, however, not allowed during these
visits, at least not until Indah had achieved locomotor
proficiency.
Preliminary results: maternal responses to the
infant’s object-directed actions
A total of 336 bouts of object-directed actions were
identified through systematic data coding (as described
above) in the data analyzed so far, which spans the interval
between birth (November 21st 2017) and December 30th
2018. Of these, 124 bouts are manual actions (reaching,
1191

exploration) and 212 are bouts of oral exploration of objects.
During this period of 12 months and 10 days, the great
majority of maternal responses to the infant’s object-directed
actions took the form of interference (N=77) and tolerate
(N=240). An additional number of 19 responses were
approach, co-action or take-over. Of the 77 interference
responses, 32 were in response to oral object exploration and
45 were in response to manual actions directed at objects. Of
the 240 tolerate responses, 163 were in response to oral
exploration and 77 were in response to manual actions.
Nearly half of the interference responses documented
between birth and December 2018 (i.e. 49.35%) was
deployed in February-April, i.e. during the period when oral
and manual exploratory skills were under development
(between 3-5 months of age). For comparison, only 61
(25.42%) tolerate responses were documented during the
same period. In addition, the ratio of interference-to-
tolerate responses decreased constantly from 1.1 in March to
0.5 during April, to then decline to 0.3 in May and June, then
to 0.2 and 0.1 in July and August respectively. Finally,
preliminary data show that the mean latency of maternal
interference had significantly increased (t(35)=3.359,
p=0.002) from March (M=7.91 seconds, SEM=1.866) to
April (M=17.19 seconds, SEM=1.610), suggesting that the
amount of time that the mother allowed the infant to
(manually or orally) explore objects had gradually increased
during this period of skill maturation.
Infant responses
The qualitative analysis revealed five categories of infant
responses to active forms of maternal behaviors: (1)
compliance: infant ceases activity, and does not engage in a
new activity; (2) compliance followed by mother-initiated
activity: infant ceases activity and engages in an activity
initiated by the mother; (3) compliance followed by self-
initiated activity: infant ceases ongoing activity and directly
engages in a new one; (4) re-start: the infant re-engages in
the same activity as previously, after a brief (up to one
minute) interruption; (5) protest: infant does not adjust the
ongoing activity, and may display protesting gestures (e.g.,
push, charge, bite). Since the main focus of this paper is on
maternal behaviors that influence skill development, here we
discuss only category (2) compliance followed by mother-
initiated activity. This was observed to occur especially if
maternal interference was unsuccessful, and the infant re-
started the activity shortly after the interference. In such
cases, immediately after a new interreference the mother
engaged the infant in finger-sucking, play, foraging, or
manipulating an object, as if to distract the infant from the
activity that she attempted to discourage.
Conclusions
In this study we sought to examine maternal behaviors that
potentially mediate skill development in Sumatran orangutan
infants, by conducting a systematic qualitative analysis on the
activities of a mother-infant dyad across 28 months, starting
with the infant’s birth. The findings of this study suggest
several important conclusions. First, we found that orangutan
mothers begin to deploy maternal behaviors that have a
formative effect on the infant already from birth, and several
such behaviors are displayed already in the first few months
of life. Second, from early on, these behaviors target
locomotor and foraging skills (as noted by previous anecdotal
observations), but also social skills. Thirdly, the timing of
these maternal behaviors appears contingent upon the
infant’s competence level in a given domain, thus exhibiting
similar timing features as described for human scaffolding.
We documented the intensive occurrence of more active
forms of maternal behaviors (molding, interference, take-
over, co-action) during periods when infant skill in the
targeted domain (e.g., mutual gaze, object manipulation,
locomotion, food repertoire, foraging techniques) was under
development. As the infant’s skills developed, such active
forms were supplanted by passive ones, i.e. tolerate, and
eventually ignore. This conclusion is also supported by
preliminary results from a quantitative treatment of responses
to infant object-directed actions, which show that active
interference peaked during a period of skill acquisition
between 3-5 months of age, being subsequently replaced by
tolerate. By the end of the observation period, visual
monitoring of the infant’s activities was rare, and ignore was
the dominating maternal behavior documented.
As mentioned in the Introduction, examples of functional
teaching have been described for several non-primate
species. In all these species, however, the behaviors described
as teaching were limited to one domain and one skill only. In
contrast to these species, our data show that such scaffolding
behaviors exhibit context flexibility, targeting locomotor,
foraging and social skills. This finding is intriguing
considering that such flexibility (and generalizability) is
discussed as a key feature of human intentional teaching (e.g.,
Caro and Hauser, 1992; Moore, 2013), thus raising the
question of intentionality in the maternal behaviors that
scaffold skill development in orangutans, and in great apes,
in general. Even more so in the light of recent evidence
suggesting that great apes are capable of understanding
others’ goals (Kano & Call, 2014) and intentions (Frölich,
Wittig & Pika, 2019), are sensitive to others’ knowledge
states (Krupenye, Kano, Hirata et al., 2016) and competence
(Kendal, Hopper, Whiten et al., 2015), and may exhibit
shared intentionality, i.e. a motivation to promote shared
psychological states with others (as discussed in Persson,
Sauciuc & Madsen, 2018).
We hope that the method and data presented here will
stimulate similar research in additional mother-infant dyads
(to address the inherent limitations of cases studies, such as
the one presented here), and in other primate species. Data
from detailed and systematic observations will enable more
systematic and bias-free comparisons with human parenting
behaviors, and will further our understanding of processes
and mechanisms that support skill development in Hominids.
1192

Acknowledgments
We are thankful to Furuvik Zoo, and especially to Rickard
Beldt, Linda-Marie Lenell, Elina Lundholm, Torsten Lönn,
Natalie Magnusson, and Lotta Widlund for their assistance.
This work was supported by research grants from Thora
Ohlssons stiftelse (2019-2022), and by a travel grant from
Stiftelsen Fil dr Uno Otterstedts fond (RFh2019-0036).
References
Bard, K., Myowa-Yamakoshi, M., Tomonaga, M., Tanaka,
M., Costall, A., & Matsuzawa, T. (2005). Group
differences in the mutual gaze of chimpanzees (Pan
troglodytes). Developmental psychology, 41 4, 616-24.
Bard, K. A. (1995). Sensorimotor cognition in young feral
orangutans (Pongo pygmaeus). Primates, 36(3), 297–321.
Caro, T., & Hauser, M. (1992). Is There Teaching in
Nonhuman Animals? The Quarterly Review of Biology, 67,
151 - 174.
Chevalier-Skolnikoff, S. (1983). Sensorimotor development
in orang-utans and other primates. Journal of Human
Evolution, 12, 545-561.
Forss, S., Willems, E., Call, J., & Schaik, C.V. (2016).
Cognitive differences between orang-utan species: a test of
the cultural intelligence hypothesis. Scientific Reports, 6,
30516.
Fröhlich, M., Wittig, R., & Pika, S. (2019). The ontogeny of
intentional communication in chimpanzees in the
wild. Developmental science, 22, e12716 .
Fouts, R. S., Fouts, D. H., & Van Cantfort, T. E. (1989). The
infant Loulis learns signs from cross-fostered
chimpanzees. In R. A. Gardner, B. T. Gardner, & T. E. Van
Cantfort (Eds.), Teaching sign language to chimpanzees.
State University of New York Press.
Kano, F., & Call, J. (2014). Great Apes Generate Goal-Based
Action Predictions. Psychological Science, 25, 1691-
1698.
Kendal, R., Hopper, L.M., Whiten, A., Brosnan, S., Lambeth,
S., Schapiro, S.J., & Hoppitt, W. (2015). Chimpanzees
copy dominant and knowledgeable individuals:
implications for cultural diversity. Evolution and human
behavior, 36 1, 65-72.
Krupenye, C., Kano, F., Hirata, S., Call, J., & Tomasello, M.
(2016). Great apes anticipate that other individuals will act
according to false beliefs. Science, 354, 110 - 114.
Maestripieri, D. (2018). Maternal influences on primate
social development. Behavioral Ecology and
Sociobiology, 72, 1-12.
Maple, T., Wilson, M.E., Zucker, E., & Wilson, S.F. (2006).
Notes on the development of a mother-reared orang-utan:
The first six months. Primates, 19, 593-602.
Matsuzawa, T., Biro, D., Humle, T., Inoue-Nakamura, N.,
Tonooka, R., & Yamakoshi, G. (2001). Emergence of
culture in wild chimpanzees: Education by master-
apprenticeship. In T. Matsuzawa (Ed.), Primate origins of
human cognition and behavior. Springer-Verlag
Publishing.
Miller, L.C., & Nadler, R.D. (2007). Mother-infant relations
and infant development in captive chimpanzees and orang-
utans. International Journal of Primatology, 2, 247-261.
Moore, R. (2013). Social learning and teaching in
chimpanzees. Biology & Philosophy, 28, 879-901.
Noordwijk, M., & Schaik, C. (2005). Development of
ecological competence in Sumatran orangutans. American
Journal of Physical Anthropology, 127, 79-94.
Palincsar, A.S. (1986). The Role of Dialogue in Providing
Scaffolded Instruction. Educational Psychologist, 21, 73-
98.
Persson, T., Sauciuc, G., & Madsen, E. (2017). Spontaneous
cross-species imitation in interactions between
chimpanzees and zoo visitors. Primates, 59, 19-29.
Piaget, J. (1952). The Origins of Intelligence in Children.
(Trans. M. Cook.) New York: International Universities
Press.
Russon, A., & Galdikas, B.M. (1995). Constraints on great
apes' imitation: model and action selectivity in rehabilitant
orangutan (Pongo pygmaeus) imitation. Journal of
Comparative Psychology, 109, 5-17.
Schneider, C., Call, J., & Liebal, K. (2012). Onset and early
use of gestural communication in nonhuman great
apes. American Journal of Primatology, 74.
Schuppli, C., Meulman, E., Forss, S., Aprilinayati, F., &
Schaik, C. (2016). Observational social learning and
socially induced practice of routine skills in immature wild
orang-utans. Animal Behaviour, 119, 87-98.
Whiten, A., & Waal, E.V. (2018). The pervasive role of social
learning in primate lifetime development. Behavioral
Ecology and Sociobiology, 72, 80.
1193