Learning from others is good, with others is better: the role of social interaction in human acquisition of new knowledge

Abstract

Learning in humans is highly embedded in social interaction: since the very early stages of our lives, we form memories and acquire knowledge about the world from and with others. Yet, within cognitive science and neuroscience, human learning is mainly studied in isolation. The focus of past research in learning has been either exclusively on the learner or (less often) on the teacher, with the primary aim of determining developmental trajectories and/or effective teaching techniques. In fact, social interaction has rarely been explicitly taken as a variable of interest, despite being the medium through which learning occurs, especially in development, but also in adulthood. Here, we review behavioural and neuroimaging research on social human learning, specifically focusing on cognitive models of how we acquire semantic knowledge from and with others, and include both developmental as well as adult work. We then identify potential cognitive mechanisms that support social learning, and their neural correlates. The aim is to outline key new directions for experiments investigating how knowledge is acquired in its ecological niche, i.e. socially, within the framework of the two-person neuroscience approach.
This article is part of the theme issue ‘Concepts in interaction: social engagement and inner experiences’.

Full text

royalsocietypublishing.org/journal/rstb
Opinion piece
Cite this article: De Felice S, de C Hamilton
AF, Ponari M, Vigliocco G. 2022 Learning from
others is good, with others is better: the role
of social interaction in human acquisition of
new knowledge. Phil. Trans. R. Soc. B 378:
20210357.
https://doi.org/10.1098/rstb.2021.0357
Received: 17 January 2022
Accepted: 20 July 2022
One contribution of 23 to a theme issue
‘Concepts in interaction: social engagement
and inner experiences’.
Subject Areas:
cognition, neuroscience, behaviour
Keywords:
social learning, social interaction, interactive
learning, hyperscanning, ecological
neuroscience, two-person neuroscience
Author for correspondence:
Sara De Felice
e-mail: sara.felice.16@ucl.ac.uk
Learning from others is good, with others
is better: the role of social interaction in
human acquisition of new knowledge
Sara De Felice
1
, Antonia F. de C. Hamilton
1
, Marta Ponari
3
and
Gabriella Vigliocco
2
1
Institute of Cognitive Neuroscience, University College London (UCL), 17–19 Alexandra House Queen Square,
London WC1N 3AZ, UK
2
Experimental Psychology, 26 Bedford Way, London WC1H 0AP, UK
3
School of Psychology, University of Kent, Canterbury CT2 7NP, UK
SDF, 0000-0003-3065-3363; GV, 0000-0002-7190-3659
Learning in humans is highly embedded in social interaction: since the very
early stages of our lives, we form memories and acquire knowledge about
the world from and with others. Yet, within cognitive science and neuro-
science, human learning is mainly studied in isolation. The focus of past
research in learning has been either exclusively on the learner or (less
often) on the teacher, with the primary aim of determining developmental
trajectories and/or effective teaching techniques. In fact, social interaction
has rarely been explicitly taken as a variable of interest, despite being the
medium through which learning occurs, especially in development, but
also in adulthood. Here, we review behavioural and neuroimaging research
on social human learning, specifically focusing on cognitive models of how
we acquire semantic knowledge from and with others, and include both
developmental as well as adult work. We then identify potential cognitive
mechanisms that support social learning, and their neural correlates.
The aim is to outline key new directions for experiments investigating
how knowledge is acquired in its ecological niche, i.e. socially, within the
framework of the two-person neuroscience approach.
This article is part of the theme issue ‘Concepts in interaction: social
engagement and inner experiences’.
1. Introduction
Throughout our life, we acquire new information and form new conceptual
representations largely in social contexts: for example, babies learn from their
carers at home, pupils learn from teachers at school and by sharing their experi-
ences with other students. In the same way, adult learning typically occurs in
social contexts and in relation to peers, colleagues at work and/or mentors.
Researchers in anthropology and sociology (e.g. [1,2]), as well as in develop-
mental psychology (e.g. [3–7]) have emphasized in their work the importance
of social interaction for learning and for development. However, cognitive
psychology and neuroscience have traditionally studied cognition at the indi-
vidual level. The ‘single-brain’approach [8] studies brain and cognition using
experimental designs involving a sample of participants (children or adults)
completing a given task individually, and then makes inferences about how
the brain works more generally. It is only in the past decade that cognitive neu-
roscientists have begun to move to a ‘second-person neuroscience’approach [9],
which studies cognitive processes in interaction, including the back-and-forth
dynamics between two or more people.
A significant component of human learning lies in the ability to act on and
interact with the surrounding environment [10]. For example, during childhood,
we learn to encode features of objects via manipulation (action), and we
© 2022 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution
License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original
author and source are credited.

understand the space around us by mapping regions where
objects are within or outside manual reach [11]. Previous litera-
ture has focused on solitary individuals performing actions
(e.g. object manipulation) and showed how action supports
learning (e.g. [12]). Here, we discuss the notion and present
evidence showing that (solitary) action per se may not be
enough, but rather that action in interaction may be key to sup-
port human learning, as exercised via many forms including
gestures, object manipulation and language [13–16].
Here, we apply these ideas to the case of learning new
concepts and knowledge. Our discussion is limited in scope
as it considers especially cognitive neuroscience studies of
learning. We primarily focus on conceptual learning, broadly
defined (i.e. long-term memory for knowledge), but we
extend the review discussion to other forms (e.g. single
words, motor learning) when relevant. We first set out the
theoretical framework and rationale of this review, providing
some definitions and introducing some ideas relevant to the
study of human learning. We then move to review the evi-
dence in children and adults showing that learning benefits
from social interaction across the lifespan. We also present
neuroimaging studies to identify the neural signature of
social interactive learning. In our discussion, we identify
the possible cognitive mechanisms subserving interactive
learning. We conclude by highlighting some methodological
and theoretical issues and pose questions for future research.
(a) Theories and definitions
Human learning refers to any form of acquisition of new
knowledge and skills by an individual. One can learn new
information alone, e.g. memorizing events via reading a his-
tory book. However, often learning occurs with and from
other people. When such learning occurs via transmission
of information across members of a social group, it is defined
as ‘social learning’[17]. Importantly, there are many ways in
which learning can be social, depending on the role that the
social agent(s) has in the learning process of a given individ-
ual. Therefore, the term social learning is a broad term that
refers to any form of learning—such as motor, verbal and
knowledge-based learning—via any form of social context,
including observation of others [18,19], imitation [20] and
interactive learning (figure 1).
Observational learning refers to learning via attending to
someone else’s actions and/or listening to them delivering
information. Imitation refers to copying someone else [20].
Observational learning and imitation differ as observational
learning only requires attention to the teacher, without
immediate replication of their actions/words, while imitation
learning involves observation plus active performance. In
other words, imitational learning is defined by action.How-
ever, both observational and imitational learning can arise
when a person watches a video of another person’s actions
or speech, with no interaction between the watcher and the
video, and thus both involve the one-way transmission of
information from teacher to student, where the learner is
confined to the role of a receiver. By contrast, in interactive
learning,action occurs in interaction: both the teacher and the
student are concurrently engaged in the learning process
and they can both take full turns during the interaction.
While interactive learning can vary in terms of how inter-
active any given context is [21,22], with variations even
within single episodes (e.g. sessions), by definition any
given interaction draws in all agents as contributors: all can
participate to the interaction in both explicit (e.g. verbal
feedback) and implicit (e.g. body language) forms.
The impact of social learning for development has been lar-
gely acknowledged within sociocultural contexts, especially
with reference to the collaborative nature of learning in interaction
[23]. Within cognitive models, interactive learning is defined bya
two-way exchange of signals that includes subtle but critical
reciprocity from student to teacher (figure 1). These could indicate
understanding (or lack thereof) as well as attentiveness (or inat-
tentiveness) and thus allow the teacher to tune their lesson to
the student. Here, we refer to reciprocity as any reaction fed-back
during an online exchange that would inform the interlocutor(s)
about the quality of the exchange (e.g. nodding for understand-
ing, frowning for confusion etc.), and thus possibly allow for a
(re)direction of behaviour(s), as well as opportunities for the lear-
ner to elaborate what is being discussed. Thus, learning in
interaction requires mutual feedback between a student (or learner,
who is acquiring new knowledge) and a teacher (who is providing
new information). Importantly, independently of the type of
social context, for (social) learning to occur there must be an
enduring change in the learner’s action and/or knowledge as a
consequence of either observing, imitating or interacting with
others [24].
Such categories have been developed over the past dec-
ades especially in the context of action/motor learning, and
their applicability to knowledge-based learning may be less
obvious, although still useful to draw some conceptual dis-
tinctions. We employ these categories here to draw the
observational learning imitation learning interactive learning
Figure 1. Schematic of three types of social learning. Arrows indicate information flow between teacher and learner in three types of social learning. From leftto
right: Observational learning: learner attends to information that flows from teacher to learner. Imitation learning: learner attends to information that flows from
teacher to learner and repeats/imitates the teacher. Interactive learning: teacher and learner engage in social interaction and exchange reciprocal social signals.
Importantly, information flows back and forth from teacher to learner. (Online version in colour.)
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
2

distinction between learning from others (imitation and
observational learning) and learning with others (interactive
learning). There are only few studies that directly contrast
different types of learning. Examples include comparisons
of physical or observational learning [25] and comparisons
of sequence learning from imitation or verbal instruction
[26] or of observational versus interactive learning [27,28].
Pioneering work from developmental psychologist
Vygotsky [7,29] had long argued for a key role for the
environment, especially social environment, in learning and
development. His sociocultural theory of cognitive develop-
ment views conceptual learning as an intrinsic social
process. A number of other researchers in developmental
psychology have also emphasized the importance of social
interaction in cognitive and linguistic development (e.g.
[30]). It is the case, however, that learner–teacher (or lear-
ner–learner) social dynamics has been largely neglected by
modern cognitive neuroscience research, especially in adult-
hood. Methods adopted to study human learning have
often included single-user tasks, where participants were
required to memorize things from cards/screens, in very
repetitive and highly constrained experimental paradigms
(e.g. [31,32]). More recently, there has been a trend towards
studying human learning in more dynamic social contexts
(e.g. [33,34]). Despite the creditable effort to move away
from the traditional reductionist approach [35], this new
line of research has only partially included the social aspect
in the study of human learning. Namely, at best social context
has been included in the study design and data collection,
while the focus of data analysis has been almost exclusively
either on the learner [36] or (less often) on the teacher [37],
and only rarely on the interaction [38]. This is also the case
in many developmental studies that see the teacher (carer)
as providing an input to the learner (the child; e.g. [39]).
The problem with this is that conclusions may be based on
a partial view of what is happening during a real-world
student–teacher (or child–carer) interaction.
We propose to look at human social learning by consider-
ing three main elements from study design to data analysis
and interpretation of results: the learner, the teacher and the
interaction among the social agents involved in the learning
processes (e.g. learner–teacher and/or learner–learner). We
believe this approach will provide a systematic and compre-
hensive method to understand human learning in social
contexts, by identifying the mechanisms subserving learning
within individual cognitive and brain systems and as an inter-
active process. Importantly, we do not suggest that learning can
only occur socially, rather that interactive learning is a qualitat-
ively different phenomenon and involves different cognitive
and physiological mechanisms from learning alone. As such,
interactive learning must be studied taking social context into
careful consideration and within clear definitions.
2. Review of the evidence
(a) Human learning is social
(i) Social interaction is crucial for optimal development
Social interaction is crucial for optimal cognitive and brain
development [40–42]. This statement is relatively uncontrover-
sial and is supported by a large body of literature (for the latest
systematic review on the topic see [43]). Taking together the
results from their review, Ilyka and colleagues concluded that
an optimal development of cognitive functions—as measured
via heterogeneous neuropsychological test batteries—and
brain—as measured via structural and functional analysis of
selected regions and networks—is contingent on child–carer
interaction during the child’s first years of life. Out of 55 rel-
evant publications included in their systematic review, only
six looked at both child and carer and how the dyadic inter-
action impacted on cognitive and brain development [44–49].
Results coming from such an approach point to the importance
of ‘sensitivity’and ‘reciprocity’of both agents for optimal child
development, and in turn at the qualityof the overall interaction
to support cognition later in life.
Considering interactive learning specifically, the majority of
work on children comes from studies on language acquisition:
these robustly and consistently show that social interaction is a
critical and constraining factor for successful language devel-
opment [41]. In a pioneering study, Kuhl et al. [50] trained
nine-month-old American babies to distinguish Chinese Man-
darin sounds in three different conditions: in interaction with
a native speaker, or by exposure to either videos or sound
recordings from the same native speaker. Despite equivalent
exposure time and content of Chinese sounds, only the group
who engaged in live-interaction with the teacher showed learn-
ing, and being exposed to videos or sound recordings was
associated with no learning. While this study is not strictly
looking at knowledge-based learning (e.g. concepts), it pro-
vides strong evidence for the crucial role of interaction in
children’s learning over non-interactive learning methods.
More work on word acquisition during child–carer inter-
action has been conducted by Yu and Smith [51–54]. In their
experimental paradigm, the infant and the carer engage in a
series of free-play sessions during which they manipulate
and name various objects (toys), while both wear head-move-
ment sensors and eye-trackers. Crucially, in all their studies, the
parent (teacher) is not aware that their infant’s learning of the
objects’names will be tested after the free-play session. This
ensures that child–carer interactions are as natural as possible.
By conducting a series of dyadic analyses, Yu & Smith [52]
showed that 18-month-old infants were more likely to success-
fully learn objects’names if two things happened concurrently:
(1) the infant (learner) held the object closer so that it was visu-
ally dominant within their visual field (over other competitor
objects on the play table), and (2) the carer (teacher) named
the object. Overall, these papers demonstrate the importance
of social interaction in young children’s word-learning.
The critical role of social interaction for optimal language
development is relatively unsurprising, considering how heav-
ily human language relies on the ‘social brain’and vice versa
[41,55,56]. Also, in many developmental pathologies such
as autism spectrum disorder, social cognition deficits and com-
municative disorders are co-occurring [57,58]. Given the highly
interconnected nature of social cognition and language proces-
sing [59], learning language within a social context would be
expected to be beneficial. Therefore, because of the strong
relationship between communication and sociality, one may
argue that the social-interaction advantage is limited to
language development.
However, the beneficial effect of social interaction during
development is not limited to language. Evidence from a
variety of studies shows that social interaction supports
learning more generally across different domains, including
visuospatial categorization [60,61], procedural learning [62]
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
3

and mathematical reasoning [34]. In their study, Kostyrka-
Allchorne and colleagues [34] found that in a large group of
5-year-olds (n= 215), the physical presence of a teacher
(versus having the teacher on screen) was associated with the
highest learning, independent of whether children were
observing the teacher playing with a shape or they were play-
ing with it themselves. This study is a great example of why
student–teacher interaction should be studied as interdepen-
dent: it is certainly unlikely that the mere presence of the
teacher had something ‘magical’about it so that the child
learned more when the teacher was physically there. Equally,
audience effects—defined as ‘a change in behaviour caused
by being observed by another person’([63], pp. 160)—cannot
explain the results, as students were explicitly observed in
the on-screen condition too. Rather, there may well be some-
thing within the dynamic of learner–teacher interaction as it
occurs face-to-face that positively impacted learning. For
example, Marsh et al. [64] found that children showed more
overimitation—i.e. unnecessarily copying actions of others—
when an adult was demonstrating goal-oriented actions, com-
pared to when demonstration was presented through a
recorded video. This suggests that social factors may directly
give rise to different behaviours (e.g. overimitation) during
learning with others, and this may increase with age as people
think more about social norms ([65], see §3 below for a
discussion of the mechanisms subserving interactive learning).
(ii) Social interaction is a booster in adult human learning
Studies on interactive learning predominantly focus on chil-
dren and young people because they are considered as the
typical learners. Little is known about social learning in
adults, and even less in interactive learning specifically.
There is some evidence suggesting that, similar to what is
found in children, social interaction acts as a catalyst for learn-
ing in adults too [27,66]. Again similarly to the literature on
children, the majority of studies on interactive learning in
adults have considered the domain of language, the focus
being on second language acquisition [33,66–68]. In their
study, Jeong et al. [33] asked 36 Japanese adults to learn two
sets of unknown Korean spoken words via either translation
or videos depicting social situations. Words encoded from
the social condition showed significantly higher accuracy
rates and faster reaction times than words encoded from the
translation condition, and the social-video learning condition
was also associated with higher activity in the right temporal
parietal junction, right hippocampus and motor areas, as
measured with functional magnetic resonance imaging.
In their review, Verga & Kotz [66] reported evidence for the
importance of sociality in adult learning: specifically, learning
a second language in interaction with another person
significantly improves long-term retention of new vocabulary.
Considering domains other than language, studies on
adults have mainly looked at the impact of social interaction
in comparison to online/virtual learning environments.
Results are less conclusive: the majority of studies found no
difference in learning outcomes between teaching live versus
teaching through recorded videos [69–74]. In their intervention
study, Brokfeld et al. [69] divided 296 medical students into four
groups, three of which received 41 4 h lessons live, while the
last group watched videos of the same lessons. The group
assigned to the video condition changed daily, so that all stu-
dents saw both live and video lectures. The effectiveness of
the teaching method was evaluated by looking at students’per-
formance on 301 multiple-choice questions of the medical
exam. Similar approaches were adopted by the other studies
cited here, and all found that learning performance did not
differ across teaching methods. Despite no difference in objec-
tive performance, all these studies found that there were some
differences in subjective evaluation, with the majority of stu-
dents preferring live lessons. However, these studies did not
control for exposure time: recorded material could be replayed
multiple times while the live session was only live once.
Studies that controlled for content and exposure time across
conditions found a significant improvement in learning of
medical students during social interactive lectures compared
to recorded tutorials [75,76].
The studies reviewed above did not directly control the
social factor during learning. In a study from our group [27],
we designed two repeated-measure yoked-control experiments
and tested learning of over 50 adults during online sessions in
different conditions designed to specifically test different social
factors. People learned some facts about uncommon items
(musical instruments, ancient objects, exotic food and animals)
in interaction with a teacher (experimenter) and some other
facts from videos of another participant attending to the pre-
vious experimental session. Results robustly showed that
performance was better for items learned in live-interaction
with the teacher, and such advantages remained a week later.
In this study, all conditions were social: however, while in the
recorded-video condition, the student took part in observational
social learning (learning by watching a video of another tea-
cher–student interaction), during live sessions the student
directly engaged with the teacher (interactive social learning).
Overall, studies of interactive learning in adults yield simi-
lar results to those in children. However, scholars are less
unified on the notion that social factors matter in adult learn-
ing, possibly because adult learning—specifically of new
concepts and information—is generally under-studied com-
pared to child learning, and social factors in adulthood may
be less critical than during development. We presented studies
showing that, when exposure time is controlled for, there is an
advantage for social interactive learning in adults too, and this
is seen across a variety of stimulus types, including foreign
languages, motor skills and—central to this paper—conceptual
knowledge.
(b) The neural signature of interactive learning
Hyperscanning has become increasingly popular over the
past decade because it has the advantage of measuring
brain activity from more than one individual at the same
time, meaning the social brain can be studied while people
engage in social interaction rather than in isolation [77–80].
In a five-person electroencephalography (EEG) hyperscan-
ning study, Davidesco et al. [81] simultaneously measured
brain activity from four students and their teacher during a
science class. They found that alpha-band (8–12 Hz) brain-to-
brain synchrony (i.e. across individuals), but not intra-brain
alpha synchrony (i.e. within individuals), significantly pre-
dicted students’learning, as measured via performance in an
immediate and a delayed test a week after the class. Moreover,
moment-to-moment variation in alpha-band brain-to-brain
synchrony during the class specifically predicted what infor-
mation was retained by the students a week later. Alpha
frequency band is a well-established neural index of attention
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
4

[82], which suggests that learning was better predicted by
moments when students were attuned (or paying attention)
to the teacher, and concurrently the teacher was attuned (or
paying attention) to the students.
The same research group conducted another EEG hypers-
canning study where brain activity was recorded from 12
high school seniors simultaneously over a semester [83].
Recording took place during students’regular biology class
and was repeated over 11 sessions. Results showed that the
degree to which brain activity was synchronized across stu-
dents predicted student class engagement (quantified as
student appreciation ratings of different teaching styles and
student daily self-reported focus). In particular, they conducted
a group-based neural coherence analysis to link student-to-
group brain synchrony to different predictors. They found
that student focus predicted student-to-group synchrony
above and beyond teaching style, and also students who
were more focused on a given day showed higher synchrony
for that day.
Given the association between learner-to-group synchrony
and class engagement [83] and the link between engagement,
attentional processes and learning [84], Dikker’s group
extended their work to ask whether learner-to-group or lear-
ner-to-teacher neural synchrony predicts learner’s content
retention [85]. Using a similar real-world classroom scenario,
biology class materials were presented in either videos or live
lectures, and students completed a multiple-choice quiz after
each class. Results showed that brain-to-brain synchrony was
higher for video than for live lectures (as expected by greater
similarity in low-level processes during watching of the same
video content). However, for live lectures only, social closeness
to the teacher was related to learner–teacher brain synchrony:
in other words, when there was a contingent learner–teacher
interaction, this was reflected in their brain activity. In addition,
learning performance correlated with learner–teacher
closeness, but not with learner–teacher brain synchrony.
Using functional near-infrared spectroscopy (fNIRS),
Holper et al. [38] recorded prefrontal brain activity during the
Socratic dialog simultaneously in seventeen teacher–student
pairs. The Socratic dialog is a classical teaching model where
the teacher encourages learning by interrogating the students
via a set of structured questions. They found that learning—
as measured by students’correct responses—was associated
with higher correlation of student–teacher brain activity.
Similar findings were obtained by another group that also
used fNIRS to measure brain activity from learner and tea-
cher dyads during the acquisition of a music song [86].
They found that brain activity in the bilateral Inferior Frontal
Cortex showed learner–teacher synchronization. This was
specifically associated with moments when the learner
was observing the teacher and when learning was more
interactive (measured in terms of turn-taking). Importantly,
learner–teacher brain synchronization could predict
a student’s performance on the learned song. The same
research group conducted a further study to investigate the
causal role of such synchronization in learning [87]. They
used transcranial alternating current stimulation to induce
(or disrupt) brain synchrony in different conditions and
found that induced teacher–learner neural coupling facili-
tated motor coordination, which in turn was associated
with enhanced learning of novel Chinese songs. This intri-
guing work hints at many further interesting questions, and
it will be useful to see it replicated and extended.
Overall, these studies demonstrate that brain-to-brain
synchrony can be measured during interactive learning and
may correlate with learning performance either across
sessions [83] or even across individual events [81,86]. How-
ever, the presence of a correlation does not necessarily
reveal the causal mechanism behind the effect, and we
consider possible cognitive processes in the next section.
3. Learning from and with others: what is
special about interactive learning?
In the previous sections, we have reviewed evidence showing
that social interaction plays a key role in human learning
across the lifespan and in a variety of cognitive domains,
and also that it has a distinctive neural signature in the
brain. It remains unclear, however, what social and cognitive
mechanisms enhance learning in an interactive context. In
this section, we consider some of the possible mechanisms
that have been suggested and how these might be studied,
before moving on to the wider implications of the work.
(a) Cognitive mechanisms of interactive learning
A number of cognitive mechanisms have been proposed to
account for the advantage of interactive learning over non-
interactive learning, including stimulus saliency [88], social
arousal [89], internal motivation [90], sustained attention [51],
audience effects [63], eye-contact and gaze [91], joint attention
[92], common ground [93], attunement and shared intentional-
ity [94] and mutual predictions within inter-agents dynamics
[95]. These can be distinguished on the basis of whether they
describe effects within one individual alone (e.g. the learner)
or whether they describe the learner–teacher relational
dynamic [96]. Here, we discuss these systematically and
evaluate them in relation to interactive learning.
(b) Individual-based mechanisms: learner- and teacher-
based approaches
Individual-based mechanisms include stimulus saliency, social
arousal, internal motivation and sustained attention. Stimulus
saliency and social arousal have been proposed as possible
explanations for the social learning advantage on the basis of
the well-established effects that faces (and social stimuli more
generally) are processed differently from other types of non-
social stimuli [97], and that we get more aroused in social
than non-social situations [98]. In other words, according to
these accounts, interactive learning is not ‘special’because it
is social per se, but rather because social contexts share some fea-
tures that make encoding of information somehow more
memorable for future recalls [33]. In line with this, it has
been found that distinct neural patterns of activation are associ-
ated with encoding and retrieving information learned in
social contexts [33,67].
In addition to external bottom–up influences, the internal
motivation of the learner may be fundamental to direct sus-
tained attention, which in turn is an essential pre-requisite
of learning [84]. There is no doubt that engaging with the
learning material, by attending and processing the target
information, is a strong predictor of how well we may be
doing on a follow-up test. Yu et al. [51] demonstrated that
this may be a very early mechanism that we engage from a
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
5

young age. They found that in nine-month-olds, infant
sustained attention predicted the learning of new vocabulary
above and beyond joint attention between infant and their carer.
However, these factors seem to be telling only part of the
story, and specifically the part concerning the learner. For
example, Kostyrka-Allchorne et al. [34] found that five-year-
olds learned about atypical geometric shapes better when
there was a teacher physically present in the room with
them. The observed learning benefit may well reflect some
degree of arousal given by the physical presence of the tea-
cher. However, it cannot be excluded that the presence of
the teacher improved learning via mechanisms of relational
dynamics [99]. We simply cannot exclude either option, due
to the way the study was designed and the fact that the
focus of the analysis was limited to the learner.
In fact, the role played by the teacher is crucial in determin-
ing the learning outcomes, and yet it is often underrepresented
in the learning literature. Social communicative signals (both
verbal and non-verbal e.g. pointing, eye-gaze) are overtly
employed by an expert (the teacher) to transfer information
to a novice (the learner [100,101]). The teacher’s communi-
cation is functional to achieve successful teaching (and in
turn someone else’s learning), and as such is explicitly adjusted
to maintain the learner’s attention and assist information
transfer. The teacher’s communicative actions are therefore
the other fundamental aspects to consider in the study of
human learning.
The fact that teachers can adjust their verbal andnon-verbal
behaviour to assist the learner has been demonstrated in the
case of both children and adult learners. For example, in their
study, Brand et al. [102] showed that carers deliberately
modified both their language and their actions when
sharing information about novel object properties with infants
compared to adults (who presumably were not novice to
those objects). Similarly, Vigliocco et al. [103] further showed
that carers adapted their language and their actions when
presenting unknown versus known objects to their 2–3-year-
old children. Similar modification of action with pedagogical
intentions has been demonstrated in adults. In three
experiments, McEllin et al. [104] recorded movements of par-
ticipants playing simple xylophone melodies either alone, for
a learner watching them, or together with another participant.
They found that movement velocity was altered specifically in
the condition when participants were playing to demonstrate
musical sequence to a novice, compared to when they were
playing alone or with someone else who was expert in
the melody.
This literature demonstrates the importance of consider-
ing the teacher as well as the learner when studying how
humans learn in interaction. However, looking at one or the
other may not be enough. Studies that only consider one
side of the interaction may overlook the social dynamic
unfolding during interpersonal communication, and in turn
reach partial and/or inaccurate conclusions regarding the
mechanisms of human interactive learning.
In our work on social learning online [27], we found that
visual social cues (teacher—face and hands) impacted learn-
ing differently depending on whether learning was
interactive (student engaged in a live lecture) or observational
(student learned from a pre-recorded video of a previous ses-
sion): in two experiments, we showed a strong interaction
effect between social contingency (live versus recorded con-
trast) and social richness (whether the face of the teacher
was visible versus when just their hands or a slide was pre-
sented instead). To our knowledge, this was the first study
showing that rich social cues specifically improve interactive
but not observational learning. Our results point at the possi-
bility that different cognitive mechanisms may support
interactive and observational learning. In other words, there
may be qualitatively different processes that are involved
when we learn with others compared to when we are learning
from others.
(c) For every learner, there is a teacher, and vice versa:
interaction-based approaches
Real-time social interaction involves rich and complex behav-
ioural dynamics, with bi-directional responses and input
between two or more people [52,105–107]. Such a multifaceted
phenomenon is unlikely to rely on a single cognitive mechan-
ism but rather a number of cognitive processes, which may be
absent in a non-interactive situation. During interactive learn-
ing, learner–teacher dynamics may be characterized by joint
attention [99], common ground [93], shared intentionality
[94] or all these processes together [108]. These mechanisms
of attunement between two or more conversational partners
may allow information to be shared more effectively, and in
turn be advantageous in those situations when we learn
socially [109,110].
One approach to examining the rich and complex dynamics
of interpersonal interaction is to argue that social interaction is
more than just a context for social cognitive processes, but in
fact replaces individual mechanisms [111]. In such an enactive
model, the inter-personal relational dynamics become auton-
omous from the single individual parts making up the
interaction. This implies that traditional single-person models
have little relevance to the two-person interaction, and that
researchers need to find a new type of dynamic model to
understand interaction at a more abstract level.
However, we argue for a more incremental approach,
where social interaction is included as an additional necessary
element in the study of human cognition. As such, interperso-
nal interactions can be integrated into—and understood by
building on—models of the solo brain. For example, we
know that learning a new concept from a video will involve
processes of perception, language and memory that allow the
learner to integrate the new information into their existing
knowledge structures. Learning a concept in interaction is
likely to engage the same processes plus additional cognitive
systems (e.g. joint attention, common ground etc.), where the
moment-by-moment coordination of gaze and speech allows
these additional processes to function smoothly. Understand-
ing what these additional processes are and how they relate
to enactive models will be an important research area in
the future.
There is evidence that the quality and quantity of social
cues present in a given interaction substantially affect the
communicative outcome of that interaction [39,62]. Rich
visual cues may enable stronger attunement by providing
more information about the interaction partner’s gaze and
mental states [112,113]. Alksne [114] looked at what features
in teaching videos improved the quality of the lecture in a
group of young adults; they found that speaking over the
presentation and making eye-contact significantly improved
student engagement, which in turn has been positively
associated with learning outcomes [115].
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
6

The fact that we somehow use our body to achieve a
better attunment with our intelocutor(s) during social com-
munication has been recently well demonstrated by Fini
et al. [116]. In their study, they asked adults to guess concrete
and abstract concepts from some photos, while being in inter-
action with an avatar. The avatar moved following the
previously recorded kinematics of a real actor’s arm, from
which human movement was implicated to the avatar.
They found an association between sociality (as measured
by motor imitation and motor synchrony between the partici-
pant and the avatar) and guessing of abstract concepts. They
argued that greater motor imitation showed by the learner
specifically during more difficult trials (abstract words)
reflected a greater longing for help: participants would try
to attune more to the avatar to receive more hints and sup-
port in the guessing task. This interpretation is in line with
the argument that social attunement may be a way to support
efficient information transfer across interlocutors [117].
However, by looking only at the student, this study does
not tell us much about whether the direction of such syn-
chrony is unidirectional (from learner to teacher) or rather
bi-directional. For example, Davidesco et al. [81] found that
while learner-to-learner brain synchrony was instantaneous,
learner-to-teacher brain synchrony could best predict learn-
ing when adjusting for a temporal lag of approximately
200 ms. Specifically, student brain activity would ‘tune in’
to brain activity of the teacher only after a short delay,
suggesting a sequential, lagged transfer of information from
teachers to students. This type of data shows that, to fully
grasp the neural mechanisms of interactive learning, it may
be insufficient to focus on one social agent alone: dyadic
analysis may carry more interesting and comprehensive
information about these complex dynamics.
(d) Synchronization as a signature of social learning
A growing body of the literature is emerging showing that a
signature of interactive learning may be a bi-directional
synchrony during teacher–learner interaction (see §2 for a
review of the literature on this). When A interacts with B,
both A and B would share some processing linked to the
experience they are both part of, while brain of A would pro-
cess information about B and brain of B would process
information about A. By looking at individual brain systems
as part of an interaction, we can start to understand the full
temporal and behavioural dynamics that are reflected into
individual brain activity (of interactive agents). These
patterns of bi-directional coordination can be interpreted
within the framework of the mutual-prediction hypothesis
[95,118,119]. This claims that, when interacting with others,
we engage in social prediction all the time in order to antici-
pate other people’s actions and mental states [107,120].
Furthermore, when two people are both engaged in mutual
prediction, their brain states will correlate and thus the sig-
nals recorded from their brains will correlate, giving rise to
interbrain synchrony. Thus, predictive mechanisms present
in individual brains can give rise to a consistent cross-brain
signal that may predict learning [38,81,83,85,86].
However, claiming that brain to brain coupling on its own
can tell us something conclusive about the quality of
the social interaction, and even further, about the learning
mechanisms of teacher–student social exchange, is at best
ambitious—if not misleading (see [121] and [118] for a
discussion on this). In conjunction with studying interperso-
nal brain synchrony, it is critical to understand the
coordination of actions and how that relates to shared knowl-
edge states (see [122] for a comprehensive framework of
neural synchrony and its behavioural references). This may
be particularly useful when learning from or teaching to
someone else. In the case of interactive learning, the co-cre-
ation of knowledge and understanding is functional to the
learning process: ideally, the teacher would want to share
information, and the learner would want to tune in to their
teacher to receive and process that information, while both
would remain sensitive to feedback coming from their inter-
locutor to adjust their behaviour accordingly. It has been
proposed that the extent to which people synchronize may
be a proxy of ongoing exchanges during human social inter-
action [123,124]: in other words, high brain-to-brain
synchrony across social agents should reflect behavioural
inter-personal dynamics. Possibly, the objective is that of
reducing prediction errors and increasing affiliation and com-
municative benefits [125]. Therefore, integrating behavioural
data into hyperscanning studies is necessary to achieve a
more comprehensive and meaningful knowledge of how
humans learn from and with others.
In fact, being a form of social interaction, good pedagogy
would be therefore characterized by continuous reciprocity:
the teacher would monitor the audience’s engagement and
understanding, and use the audience feedback to adapt their
performance as needed. Such mutual prediction engages
the brain in a constant probabilistic estimate of occurrence of
external experiences based on expected outcome. These may
be plausible mechanisms underlying inter-personal synchrony
and shared neural representations typical of social situations
[96]. Studies have shown that interpersonal synchrony
manifests across multiple levels during social interaction,
including motor coordination [126,127], action coordination
and decision-making [128] and verbal coordination [129–
131]. In addition, person-to-person synchrony has been
reported even at the physiological [132,133] and neural levels
[122,125,134,135].
The study of interactive learning cannot answer questions
on how individual cognitive mechanisms work per se, unless
research considers the individual agents alone and as part of
an integrated social dynamic where they learn from (and/
or with) one another. Taking a second-person neuroscience
approach [9,118] across all stages of the experimental work
is particularly important as we are moving away from study-
ing learning in isolation to study learning in social contexts:
we must study interactive minds as they are found in the
real world in the context of rich interactions, to fully under-
stand interactive learning dynamics as they unfold [136].
We believe that this approach can give us a comprehensive
understanding of what factors influence learning and its
underlying cognitive mechanisms and neural markers.
4. Final remarks and future questions
The scope of this review was to look at the state-of-the-art in
the neuroscience of human learning as it most naturally
occurs, i.e. socially. Social interactions feature as the major
catalyst for the human ability to acquire and retain new infor-
mation [27,42,137], even in robot–human interaction [138].
We have presented evidence showing the crucial role that
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
7

social interaction plays in human learning across the lifespan.
We have argued that the relation between social interaction
and learning may be modulated by complex dynamics span-
ning across behaviour, physiology and the brain [9,139]. As
such, we believe that these complex dynamics are unlikely
to be fully grasped by experiments that look solely at either
the learner or the teacher in isolation, rather than as agents
who are part of an interaction. Social agents will inevitably
influence and be influenced by each other, and as such the
inter-personal dynamics need to be taken into account to
fully grasp the cognitive and neural mechanisms subserving
social interactive learning.
Taking the cited literature together, a few issues emerge.
First, behavioural and cognitive mechanisms involved in
learning are mainly studied in isolation: it remains largely
unexplored how the experience of learning from others modu-
lates the physiological, behavioural and neural response of the
people involved in the interaction, both individually and as
a coordinated system. Future studies should integrate
the multimodal experimental design and analysis pipeline
to grasp the complexity of interactive learning and the
mechanisms subserving it.
Second, there are disproportionally more studies on inter-
active learning in the domain of language acquisition, thus
specifically looking at childhood, than any other domain,
both in children and adults. Coordination in language use
has been demonstrated in adult speakers as well (e.g.
[105,106]) due to the tight link between language and social
interaction, given the social nature of communication. This
coordination may be particularly important in conceptual
learning at all ages because of the particular challenges
required in learning new concepts. Unlike learning an arbitrary
word list, a new concept must be integrated with other existing
knowledge of many different types (e.g. [140]). Interactive
teaching would allow a learner to try out a new concept and
explore how it relates to other concepts with immediate feed-
back, which is likely to provide richer and more robust
learning. Thus, the study of interactive conceptual learning
should be integrated with the study of interactive communi-
cation, to include coordination in a number of non-verbal
behaviours such as gestures and eye-gaze [107,141]. We
know very little about how coordination within and across
these different channels can support conceptual learning
across the lifespan (but see [27]). We also know very little
about how we learn in interaction in ways that do not involve
communication, for example during large in-person classes
where although there is co-presence, there is no or very little
active exchange. Studies addressing these questions are
needed in order to assess to what extent the benefit of social
interaction in learning is content-dependent.
Third, within the domain of interactive learning research,
more direct comparisons of observational and interactive
social learning are needed. This will be essential to disentangle
the contribution of specific factors associated with social
contextsthat benefit human learning. It is hard to separate indi-
vidual components because live interaction cannot be easily
deconstructed. Future studies using virtual reality might be
able to do so by experimentally manipulating which aspects
of interaction are most important to learning [142].
Educational neuroscience has recently emerged as a
growing discipline worldwide: it aims to identify the key com-
ponents of naturalistic social engagement during learning. By
adopting a data-driven multimodal approach [143], learner–
teacher interaction can be dissected to understand infor-
mation-transfer processes and how these are modulated by
interpersonal dynamics [40,144–146]. Here, we have reviewed
evidence that allows us to make two claims: first, social inter-
action is an integral part of human learning. Second, learning
in social (and interactive) contexts engages partially different
mechanisms from learning in non-social (and non-interactive)
contexts. Social interaction in fact employs a series of processes
unique to interactive situations, including (but not limited to)
joint attention, reciprocity and active attunement, that may be
key to support learning in humans.
In conclusion, we presented studies to show the role of social
interaction in the first years of life for optimal cognitive and
brain development, and demonstrate that social interaction
boosts learning in adulthood too. We reported studies that
have specifically looked at complex real-world learning (e.g.
from object features in children to complex medical curricula
in adults) and placed learning in its ecology:these have included
social interaction at the different stages of the experimental pro-
cess to fully grasp the multifaceted mechanisms of interactive
learning, and we have discussed their evidence in relation to
less comprehensive approaches. Taken together, these suggest
that learning via solitary action may not be enough, while learn-
ing in interaction with others may be a key factor supporting
acquisition of new knowledge in the real world.
Data accessibility. This article has no additional data.
Authors’contributions. S.D.F.: conceptualization, writing—original draft,
writing—review and editing; A.F.C.H.: conceptualization, supervi-
sion, writing—review and editing; M.P.: conceptualization,
writing—review and editing; G.V.: conceptualization, supervision,
writing—review and editing.
All authors gave final approval for publication and agreed to be
held accountable for the work performed therein.
Conflict of interest declaration. We declare we have no competing interests.
Funding. At the time of publication S.D.F. is supported by Leverhulme
award DS-2017-026. G.V. is supported by European Research Council
Advanced Grant (grant no. ECOLANG, 743035) and Royal Society
Wolfson Research Merit Award (grant no. WRM\R3\170016).
References
1. Schegloff EA. 2007 Sequence organization
in interaction: a primer in conversation
analysis. Cambridge, UK: Cambridge University
Press.
2. Sacks H, Schegloff EA, Jefferson G. 1974
A simplest systematics for the organization of
turn-taking for conversation. Language 50,
696–735. (doi:10.2307/412243)
3. Bruner JS. 1957 Going beyond the information given.
New York, NY: Norton.
4. Bruner JS. 1978 The role of dialogue in language
acquisition. In The child’s concept of language (eds
A Sinclair, RJ Jarvelle, WJM Levelt), pp. 241–256.
New York, NY: Springer.
5. Nomikou I, Leonardi G, Rohlfing KJ,
Rączaszek-Leonardi J. 2016 Constructing interaction:
the development of gaze dynamics.
Infant Child Dev. 25, 277–295. (doi:10.1002/
icd.1975)
6. Rohlfing KJ, Wrede B, Vollmer AL, Oudeyer PY.
2016 An alternative to mapping a word onto
a concept in language acquisition: pragmatic
frames. Front. Psychol. 7, 470. (doi:10.3389/fpsyg.
2016.00470)
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
8

7. Vygotsky LS. 1978 Mind in society: The development
of higher psychological processes. Cambridge, MA:
Harvard University Press.
8. Gazzaniga MS, Ivry RB, Mangun GR. 2002 Cognitive
neuroscience: the biology of the mind. New York, NY:
Norton.
9. Redcay E, Schilbach L. 2019 Using second-person
neuroscience to elucidate the mechanisms of social
interaction. Nat. Rev. Neurosci. 20, 495–505.
(doi:10.1038/s41583-019-0179-4)
10. Goodale MA, Milner AD. 1992 Separate visual
pathways for perception and action. Trends Neurosci.
15,20–25. (doi:10.1016/0166-2236(92)90344-8)
11. Rizzolatti G, Fadiga L, Fogassi L, Gallese V. 1997 The
space around us. Science 277, 190–191. (doi:10.
1126/science.277.5323.190)
12. Focaroli V, Iverson JM. 2017 Children’s object
manipulation: a tool for knowing the external world
and for communicative development. In Studies in
applied philosophy, epistemology and rational ethics,
vol. 38 (eds M Bertolaso, N Di Stefano), pp. 19–28.
Cham, Switzerland: Springer.
13. Fusaroli R, Bahrami B, Olsen K, Roepstorff A, Rees G,
Frith C, Tylén K. 2012 Coming to terms: quantifying
the benefits of linguistic coordination. Psychol. Sci.
23, 931–939. (doi:10.1177/0956797612436816)
14. Jant EA, Haden CA, Uttal DH, Babcock E. 2014
Conversation and object manipulation influence
children’s learning in a museum. Child Dev. 85,
2029–2045. (doi:10.1111/cdev.12252)
15. Nölle J, Fusaroli R, Mills GJ, Tylén K. 2020 Language
as shaped by the environment: linguistic construal
in a collaborative spatial task. Palgrave Commun. 6,
1–10. (doi:10.1057/s41599-020-0404-9)
16. Rocca R, Wallentin M, Vesper C, Tylen K. 2019 This
is for you: social modulations of proximal vs. distal
space in collaborative interaction. Sci. Rep. 9, 14967.
(doi:10.1038/s41598-019-51134-8)
17. Tomasello M. 2004 Learning through others. The
Cambridge Habermas Lexicon 133,51–58. (doi:10.
1017/9781316771303.200)
18. Cross ES. 2011 Observational learning of complex
motor skills. In Encyclopedia of the sciences of
learning (ed. N Seel), pp. 2491–2493, 2012 edn.
Berlin, Germany: Springer.
19. Cross ES, Kraemer DJM, Hamilton AFDC, Kelley WM,
Grafton ST. 2009 Sensitivity of the action
observation network to physical and observational
learning. Cereb. Cortex 19, 315–326. (doi:10.1093/
cercor/bhn083)
20. Bandura A. 2019 Social learning: theory. In
Encyclopedia of animal behavior (ed. J Choe),
pp. 380–386. (doi:10.1016/B978-0-12-813251-7.
00057-2)
21. Rogoff B. 1998 Cognition as a collaborative process.
In Handbook of child psychology. Vol. 2: Cognition,
language, and perception, 5th edition (eds D Kuhn,
RS Siegler, W Damon (series editor)), pp. 679–744.
New York, NY: Wiley.
22. Rogoff B, Moore L, Najafi B, Dexter A, Correa-
Chávez M, Solís J. 2007 Children’s development of
cultural repertoires through participation in
everyday routines and practices. In Handbook of
socialization (eds JE Grusec, PD Hastings),
pp. 490–515. New York, NY: Guilford.
23. Rogoff B, Turkanis CG, Bartlett L (eds). 2001
Learning together: children and adults in a school
community. New York, NY: Oxford University Press.
24. Ramsey R, Kaplan DM, Cross ES. 2021 Watch and
learn: the cognitive neuroscience of learning from
others’actions. Trends Neurosci. 44, 478–491.
(doi:10.1016/j.tins.2021.01.007)
25. Cross ES, Hamilton AFdC, Cohen NR, Grafton ST.
2017 Learning to tie the knot: the acquisition of
functional object representations by physical and
observational experience. PLoS ONE 12,1–24.
(doi:10.1371/journal.pone.0185044)
26. Renner E, White JP, Hamilton AFdC, Subiaul F.
2018 Neural responses when learning spatial
and object sequencing tasks via imitation. PLoS
ONE 13, e0201619. (doi:10.1371/journal.pone.
0201619)
27. De Felice S, Vigliocco G, Hamilton AFdC. 2021 Social
interaction is a catalyst for adult human learning in
online contexts. Curr. Biol. 31,1–7. (doi:10.1016/j.
cub.2021.08.045)
28. Matheson H, Moore C, Akhtar N. 2013 The
development of social learning in interactive and
observational contexts. J. Exp. Child Psychol. 114,
161–172. (doi:10.1016/j.jecp.2012.09.003)
29. Vygotsky LS. 1962 Thought and language.
Cambridge, MA: MIT Press. (Original work published
1934.)
30. Rohlfing KJ, Leonardi G, Nomikou I, Raczaszek-
Leonardi J, Hüllermeier E. 2020 Multimodal turn-
taking: motivations, methodological challenges, and
novel approaches. IEEE Trans. Cogn. Dev. Syst. 12,
260–271. (doi:10.1109/TCDS.2019.2892991)
31. Batial L, Shmueli K. 1997 Memorizing new words:
does teaching have anything to do with it? RELC J.
28,89–108. (doi:10.1177/003368829702800106)
32. Duff FJ, Hulme C. 2012 The role of children’s
phonological and semantic knowledge in learning
to read words. Sci. Stud. Reading 16, 504–525.
(doi:10.1080/10888438.2011.598199)
33. Jeong H, Li P, Suzuki W, Sugiura M, Kawashima R.
2021 Neural mechanisms of language learning from
social contexts. Brain Lang. 212, 104874. (doi:10.
1016/j.bandl.2020.104874)
34. Kostyrka-Allchorne K, Holland A, Cooper NR,
Ahamed W, Marrow RK, Simpson A. 2019 What
helps children learn difficult tasks: a teacher’s
presence may be worth more than a screen. Trends
Neurosci. Educ. 17, 100114. (doi:10.1016/j.tine.
2019.100114)
35. Putnam H. 1973 Reductionism and the nature of
psychology. Cognition 2, 131–146. (doi:10.1016/
0010-0277(72)90033-9)
36. Gilbert CD, Sigman M, Crist RE. 2001 The neural
basis of perceptual learning. Neuron 31, 681–697.
(doi:10.1016/S0896-6273(01)00424-X)
37. Battro AM. 2010 The teaching brain. Mind Brain
Educ. 4,28–33. (doi:10.1111/j.1751-228X.2009.
01080.x)
38. Holper L, Goldin AP, Shalóm DE, Battro AM, Wolf M,
Sigman M. 2013 The teaching and the learning
brain: a cortical hemodynamic marker of teacher–
student interactions in the Socratic dialog. Int. J.
Educ. Res. 59,1–10. (doi:10.1016/j.ijer.2013.02.
002)
39. Cartmill EA, Armstrong BF, Gleitman LR, Goldin-
Meadow S, Medina TN, Trueswell JC. 2013 Quality
of early parent input predicts child vocabulary 3
years later. Proc. Natl Acad. Sci. USA 110,11
278–11 283. (doi:10.1073/pnas.1309518110)
40. Goswami U. 2006 Neuroscience and education: from
research to practice? Nat. Rev. Neurosci. 7, 406–411.
(doi:10.1038/nrn1907)
41. Kuhl PK. 2007 Is speech learning ‘gated’by the
social brain? Dev. Sci. 10, 110–120. (doi:10.1111/j.
1467-7687.2007.00572.x)
42. Meltzoff A, Kuhl P, Movellan J, Sejnowski TJ.
2009 Foundations for a new science of learning.
Science 325, 284–288. (doi:10.1126/science.
1175626)
43. Ilyka D, Johnson MH, Lloyd-Fox S. 2021 Infant social
interactions and brain development: a systematic
review. Neurosci. Biobehav. Rev. 130, 448–469.
(doi:10.1016/j.neubiorev.2021.09.001)
44. Beckwith L, Parmelee Jr AH. 1986 EEG patterns of
preterm infants, home environment, and later IQ.
Child. Dev. 57, 777–789. (doi:10.2307/1130354)
45. Elsabbagh M et al. 2012 Infant neural sensitivity to
dynamic eye gaze is associated with later emerging
autism. Curr. Biol. 22, 338–342. (doi:10.1016/j.cub.
2011.12.056)
46. Gartstein MA, Warwick H, Campagna AX. 2020
Electroencephalogram frontal asymmetry changes
during emotion-eliciting tasks and parent–child
interaction dynamics. Soc. Dev. 30, 496–514.
(doi:10.1111/sode.12484)
47. Jones NA, McFall BA, Diego MA. 2004 Patterns of
brain electrical activity in infants of depressed
mothers who breastfeed and bottle feed: the
mediating role of infant temperament. Biol. Psychol.
67, 103–124. (doi:10.1016/j.biopsycho.2004.03.010)
48. Perone S, Gartstein MA. 2019 Relations between
dynamics of parent-infant interactions and baseline
EEG functional connectivity. Infant Behav. Dev. 57,
101344. (doi:10.1016/j.infbeh.2019.101344)
49. Pratt M, Zeev-Wolf M, Goldstein A, Feldman R. 2019
Exposure to early and persistent maternal
depression impairs the neural basis of attachment in
preadolescence. Progress Neuro-Psychopharmacol.
Biol. Psychiatry 93,21–30. (doi:10.1016/j.pnpbp.
2019.03.005)
50. Kuhl PK, Tsao FM, Liu HM. 2003 Foreign-language
experience in infancy: effects of short-term exposure
and social interaction on phonetic learning. Proc.
Natl Acad. Sci. USA 100, 9096–9101. (doi:10.1073/
pnas.1532872100)
51. Yu C, Suanda SH, Smith LB. 2017 Infant
sustained attention but not joint attention to
objects at 9 months predicts vocabulary at 12
and 15 months. Dev. Sci. 22,1–12. (doi:10.1111/
desc.12735)
52. Yu C, Smith LB. 2012 Embodied attention and word
learning by toddlers. Cognition 125, 244–262.
(doi:10.1016/j.cognition.2012.06.016)
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
9

53. Yu C, Smith LB. 2013 Joint attention without gaze
following: human infants and their parents
coordinate visual attention to objects through eye-
hand coordination. PLoS ONE 8, e79659. (doi:10.
1371/journal.pone.0079659)
54. Yu C, Smith LB. 2016 The social origins of sustained
attention in one-year-old human infants. Curr. Biol.
26, 1235–1240. (doi:10.1016/j.cub.2016.03.026)
55. Blakemore SJ. 2010 The developing social brain:
implications for education. Neuron 65, 744–747.
(doi:10.1016/j.neuron.2010.03.004)
56. Kuhl PK. 2003 Human speech and birdsong:
communication and the social brain. Proc. Natl
Acad. Sci. USA 100, 9645–9646. (doi:10.1073/pnas.
1733998100)
57. Happé FGE. 1995 Understanding minds and
metaphors: insights from the study of figurative
language in autism. Metaphor Symb. Activity 10,
275–295. (doi:10.1207/s15327868ms1004_3)
58. Whitehouse AJO, Barry JG, Bishop DJM. 2007 The
broader language phenotype of autism: a
comparison with specific language impairment.
J. Child Psychol. Psychiatry Allied Disciplines 48,
822–830. (doi:10.1111/j.1469-7610.2007.01765.x)
59. Binney RJ, Ramsey R. 2020 Social semantics: the
role of conceptual knowledge and cognitive control
in a neurobiological model of the social brain.
Neurosci. Biobehav. Rev. 112(January), 28–38.
(doi:10.1016/j.neubiorev.2020.01.030)
60. Lauricella AR, Gola AA, Calvert SL. 2011 Toddlers’
learning from socially meaningful video characters.
Media Psychol. 14, 216–232. (doi:10.1080/
15213269.2011.573465)
61. Lauricella AR, Gola AAH, Calvert SL. 2011 Toddlers’
learning from socially meaningful video characters.
Media Psychol. 14, 216–232. (doi:10.1080/
15213269.2011.573465)
62. Sauppé A, Mutlu B. 2014 How social cues shape
task coordination and communication. Proceedings
of the ACM Conference on Computer Supported
Cooperative Work, CSCW,97–108. (doi:10.1145/
2531602.2531610)
63. Hamilton AFdC, Lind F. 2016 Audience effects: what
can they tell us about social neuroscience, theory of
mind and autism? Cult. Brain 4, 159–177. (doi:10.
1007/s40167-016-0044-5)
64. Marsh L, Pearson A, Ropar D, Hamilton A.
2013 Children with autism do not overimitate.
Curr. Biol 23, R266–R268. (doi:10.1016/j.cub.2013.
02.036)
65. Clay Z, Over H, Tennie G. 2018 What drives young
children to over-imitate? Investigating the effects of
age, context, action type, and transitivity. J. Exp.
Child Psychol. 166, 520–534. (doi:10.1016/j.jecp.
2017.09.008)
66. Verga L, Kotz SA. 2013 How relevant is social
interaction in second language learning? Front.
Hum. Neurosci. 7,1–7. (doi:10.3389/fnhum.2013.
00550)
67. Jeong H, Sugiura M, Sassa Y, Wakusawa K, Horie K,
Sato S, Kawashima R. 2010 Learning second
language vocabulary: neural dissociation of
situation-based learning and text-based learning.
Neuroimage 50, 802–809. (doi:10.1016/j.
neuroimage.2009.12.038)
68. Li P, Jeong H. 2020 The social brain of language:
grounding second language learning in social
interaction. npj Sci. Learn. 5, 8. (doi:10.1038/
s41539-020-0068-7)
69. Brockfeld T, Müller B, de Laffolie J. 2018 Video
versus live lecture courses: a comparative evaluation
of lecture types and results. Med. Educ. Online 23,
1555434. (doi:10.1080/10872981.2018.1555434)
70. Davis J, Crabb S, Rogers E, Zamora J, Khan K. 2008
Computer-based teaching is as good as face to face
lecture-based teaching of evidence based medicine:
a randomized controlled trial. Med. Teach. 30,
302–307. (doi:10.1080/01421590701784349)
71. Phillips JA. 2015 Replacing traditional live lectures
with online learning modules: effects on learning
and student perceptions. Curr. Pharmacy Teach.
Learn. 7, 738–744. (doi:10.1016/j.cptl.2015.08.009)
72. Schreiber BE, Junaid F, Fabiana G. 2010 Live lecture
versus video podcast in undergraduate medical
education. Med. Educ. 10,1–6. (doi:10.1186/1472-
6920-10-68)
73. Solomon DJ, Ferenchick GS, Laird-Fick HS,
Kavanaugh K. 2004 A randomized trial comparing
digital and live lecture formats ISRCTN40455708.
BMC Med. Educ. 4,1–6. (doi:10.1186/1472-6920-4-
27)
74. Vaccani JP, Javidnia H, Humphrey-Murto S. 2016
The effectiveness of webcast compared to live
lectures as a teaching tool in medical school. Med.
Teach. 38,59–63. (doi:10.3109/0142159X.2014.
970990)
75. John JR, Priya H, Hemasundar A, Rajendran G. 2016
A collation of two forms of demonstration: in-
person versus video demonstration which gains the
upper hand? Res. Rev. J. Dent. Sci. 4,23–29.
76. Ramlogan S, Raman V, Sweet J. 2014 A comparison
of two forms of teaching instruction: video vs.
live lecture for education in clinical periodontology.
Eur. J. Dent. Educ. 18,31–38. (doi:10.1111/
eje.12053)
77. Babiloni F, Astolfi L. 2014 Social neuroscience and
hyperscanning techniques: past, present and future.
Neurosci. Biobehav. Rev. 44,76–93. (doi:10.1016/j.
neubiorev.2012.07.006)
78. Czeszumski A, Eustergerling S, Lang A, Menrath D,
Gerstenberger M, Schuberth S, Schreiber F, Rendon
ZZ, König P. 2020 Hyperscanning: a valid method to
study neural inter-brain underpinnings of social
interaction. Front. Hum. Neurosci. 14,1–17. (doi:10.
3389/fnhum.2020.00039)
79. Dumas G, Lachat F, Martinerie J, Nadel J, George N.
2011 From social behaviour to brain
synchronization: review and perspectives in
hyperscanning. Irbm 32,48–53. (doi:10.1016/j.
irbm.2011.01.002)
80. Kelsen BA, Sumich A, Kasabov N, Liang SHY,
Wang GY. 2020 What has social neuroscience
learned from hyperscanning studies of spoken
communication? A systematic review. Neurosci.
Biobehav. Rev. 132, 1249–1262. (doi:10.1016/j.
neubiorev.2020.09.008)
81. Davidesco I, Laurent E, Valk H, West T, Dikker S,
Milne C, Poeppel D. 2019 Brain-to-brain synchrony
predicts long-term memory retention more
accurately than individual brain measures. BioRxiv
644047. (doi:10.1101/644047)
82. Klimesch W, Sauseng P, Hanslmayr S. 2007 EEG
alpha oscillations: the inhibition-timing hypothesis.
Brain Res. Rev. 53,63–88. (doi:10.1016/j.
brainresrev.2006.06.003)
83. Dikker S et al. 2017 Brain-to-brain synchrony tracks
real-world dynamic group interactions in the
classroom. Curr. Biol. 27, 1375–1380. (doi:10.1016/
j.cub.2017.04.002)
84. Nissen MJ, Bullemer P. 1987 Attentional
requirements of learning: evidence from
performance measures. Cogn. Psychol. 19,1–32.
(doi:10.1016/0010-0285(87)90002-8)
85. Bevilacqua D, Davidesco I, Wan L, Chaloner K,
Rowland J, Ding M, Poeppel D, Dikker S. 2019
Brain-to-brain synchrony and learning outcomes
vary by student–teacher dynamics: evidence from a
real-world classroom electroencephalography study.
J. Cogn. Neurosci. 31, 401–411. (doi:10.1162/jocn_
a_01274)
86. Pan Y, Novembre G, Song B, Li X, Hu Y. 2018
Interpersonal synchronization of inferior frontal
cortices tracks social interactive learning of a song.
Neuroimage 183, 280–290. (doi:10.1016/j.
neuroimage.2018.08.005)
87. Pan Y, Novembre G, Song B, Zhu Y, Hu Y. 2020 Dual
brain stimulation enhances interpersonal learning
through spontaneous movement synchrony. Soc.
Cogn. Affect. Neurosci. 16, 210–221. (doi:10.1093/
scan/nsaa080)
88. End A, Gamer M. 2017 Preferential processing of
social features and their interplay with physical
saliency in complex naturalistic scenes. Front.
Psychol. 8,1–16. (doi:10.3389/fpsyg.2017.00418)
89. Berger J. 2011 Arousal increases social transmission
of information. Psychol. Sci. 22, 891–893. (doi:10.
1177/0956797611413294)
90. Evans M, Boucher AR. 2015 Optimizing the power
of choice: supporting student autonomy to foster
motivation and engagement in learning. Mind Brain
Educ. 9,87–91. (doi:10.1111/mbe.12073)
91. Ho S, Foulsham T, Kingstone A. 2015 Speaking and
listening with the eyes: gaze signaling during
dyadic interactions. PLoS ONE 10,1–18. (doi:10.
1371/journal.pone.0136905)
92. Mundy P, Newell L. 2007 Attention, joint attention
& social cognition. Soc. Cogn. 16, 269–274. (doi:10.
1111/j.1467-8721.2007.00518.x.Attention)
93. Bohn M, Tessler MH, Frank M. 2019 Integrating
common ground and informativeness in pragmatic
word learning. PsyArXiv. (doi:10.31234/osf.io/cbx46)
94. Sabbagh MA, Baldwin DA. 2001 Learning words
from knowledgeable versus ignorant speakers: links
between preschoolers’theory of mind and semantic
development. Child Dev. 72, 1054–1070. (doi:10.
1111/1467-8624.00334)
95. Kingsbury L, Huang S, Wang J, Gu K, Golshani P,
Wu YE, Hong W. 2019 Correlated neural activity and
encoding of behavior across brains of socially
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
10

interacting animals. Cell 178, 429–446.e16. (doi:10.
1016/j.cell.2019.05.022)
96. Brown EC, Brüne M. 2012 The role of prediction in
social neuroscience. Front. Hum. Neurosci. 6,1–19.
(doi:10.3389/fnhum.2012.00147)
97. Bentin S, Allison T, Puce A, Perez E, McCarthy G.
1981 Electrophysiological studies of face perception
in humans. J. Cogn. Neurosci. 8, 551–565. (doi:10.
1162/jocn.1996.8.6.551.Electrophysiological)
98. Cacioppo JT, Berntson GG, Crites SL. 1996 Social
neuroscience: principles of psychophysiological
arousal and response. In Social psychology:
handbook of basic principles (eds ET Higgins, AW
Kruglanski), pp. 72–101. New York, NY: The
Guilford Press.
99. Schertz HH, Odom SL, Baggett KM, Sideris JH. 2013
Effects of joint attention mediated learning for
toddlers with autism spectrum disorders : an initial
randomized controlled study. Early Childhood Res. Q.
28, 249–258. (doi:10.1016/j.ecresq.2012.06.006)
100. Csibra G, Gergely G. 2009 Natural pedagogy. Trends
Cogn. Neurosci. 13, 148–153. (doi:10.1016/j.tics.
2009.01.005)
101. Csibra G, Gergely G. 2011 Natural pedagogy as
evolutionary adaptation. Phil. Trans. R. Soc. B 366,
1149–1157. (doi:10.1098/rstb.2010.03191149)
102. Brand RJ, Baldwin DA, Ashburn LA. 2002 Evidence
for ‘motionese’: modifications in mothers’infant-
directed action. Dev. Sci. 5,72–83. (doi:10.1111/
1467-7687.00211)
103. Vigliocco G, Murgiano M, Motamedi Y, Wonnacott E,
Marshall CR, Milán-Maillo I, Perniss P. 2019
Onomatopoeias, gestures, actions and words: how
do caregivers use multimodal cues in their
communication to children? CogSci. 146. (doi:10.
31234/osf.io/v263k)
104. McEllin L, Sebanz N, Knoblich G. 2018 Identifying
others’informative intentions from movement
kinematics. Cognition 180, 246–258. (doi:10.1016/j.
cognition.2018.08.001)
105. Clark HH. 1996 Using language. Cambridge, UK:
Cambridge University Press.
106. Garrod S, Pickering MJ. 2004 Why is conversation so
easy? Trends Cogn. Sci. 8,8
–11. (doi:10.1016/j.tics.
2003.10.016)
107. Holler J, Levinson SC. 2019 Multimodal language
processing in human communication. Trends
Cogn. Sci. 23, 639–652. (doi:10.1016/j.tics.2019.
05.006)
108. Schmitz M. 2014 Joint attention and understanding
others. Synthesis Phil. 29, 235–251.
109. Frith CD, Frith U. 2012 Mechanisms of social
cognition. Annu. Rev. Psychol. 63, 287–313. (doi:10.
1146/annurev-psych-120710-100449)
110. Hu J, Qi S, Becker B, Luo L, Gao S, Gong Q,
Hurlemann R, Kendrick KM. 2015 Oxytocin
selectively facilitates learning with social feedback
and increases activity and functional connectivity in
emotional memory and reward processing regions.
Hum. Brain Mapp. 36, 2132–2146. (doi:10.1002/
hbm.22760)
111. De Jaegher H, Di Paolo E, Gallagher S. 2010
Can social interaction constitute social cognition?
Trends Cogn. Sci. 14, 441–447. (doi:10.1016/j.tics.
2010.06.009)
112. Kajopoulos J, Cheng G, Kise K, Müller HJ, Wykowska
A. 2020 Focusing on the face or getting distracted by
social signals? The effect of distracting gestures on
attentional focus in natural interaction. Psychol. Res.
85, 491–502. (doi:10.1007/s00426-020-01383-4)
113. Marotta A, Lupiáñez J, Martella D, Casagrande M.
2012 Eye gaze versus arrows as spatial cues: two
qualitatively different modes of attentional
selection. J. Exp. Psychol. Hum. Percept. Perform. 38,
326–335. (doi:10.1037/a0023959)
114. Alksne L. 2016 How to produce video lectures to
engage students and deliver the maximum amount
of information. In SOCIETY. INTEGRATION.
EDUCATION. Proceedings of the International
Scientific Conference,Rēzeknes Tehnoloģiju
akadēmija, Rezekne, Latvia, 27–28 May
2016, vol. 2, pp. 503–516. (doi:10.17770/
sie2016vol2.1424)
115. Bevilacqua D, Davidesco I, Wan L, Chaloner K,
Rowland J, Ding M, Poeppel D, Dikker S. 2019
Brain-to-brain synchrony and learning outcomes
vary by student–teacher dynamics: evidence from a
real-world classroom electroencephalography study.
J. Cogn Neurosci. 31, 401–411. (doi:10.1162/jocn_
a_01274)
116. Fini C, Era V, Da Rold F, Candidi M, Borghi AM. 2021
Abstract concepts in interaction: the need of others
when guessing abstract concepts smooths dyadic
motor interactions. R. Soc. Open Sci. 8, 201205.
(doi:10.1098/rsos.201205)
117. Shafto P, Goodman ND, Frank MC. 2012 Learning
from others: the consequences of psychological
reasoning for human learning. Perspect. Psychol. Sci.
7, 341–351. (doi:10.1177/1745691612448481)
118. Hamilton AFdC. 2020 Hyperscanning: beyond the
hype. Neuron 2020,8–11. (doi:10.1016/j.neuron.
2020.11.008)
119. Nishimura T, Hara A, Miyamoto H, Furukawa M,
Maeda T. 2020 Mutual prediction model for
predicting information for human motion
generation. In Proc. of the 2020 IEEE/SICE Int. Symp.
on System Integration,Honolulu, HI, 12–15 January
2020, pp. 687–692. (doi:10.1109/SII46433.2020.
9026182)
120. Tamir DI, Thornton MA. 2018 Modeling the
predictive social mind. Trends Cogn. Sci. 22,
201–212. (doi:10.1016/j.tics.2017.12.005)
121. Novembre G, Iannetti GD. 2021 Hyperscanning
alone cannot prove causality. Multibrain stimulation
can. Trends Cogn. Sci. 25,96–99. (doi:10.1016/j.tics.
2020.11.003)
122. Hasson U, Ghazanfar AA, Galantucci B, Garrod S,
Keysers C. 2012 Brain-to-brain coupling: a
mechanism for creating and sharing a social world.
Trends Cogn. Sci. 16, 114–121. (doi:10.1016/j.tics.
2011.12.007)
123. Balconi M, Pezard L, Nandrino JL, Vanutelli ME.
2017 Two is better than one: the effects of strategic
cooperation on intra- and inter-brain connectivity
by fNIRS. PLoS ONE 12,1–17. (doi:10.1371/journal.
pone.0187652)
124. Reinero DA, Dikker S, Van Bavel JJ. 2020 Inter-brain
synchrony in teams predicts collective performance.
Soc. Cogn. Affect. Neurosci. 16,43–57. (doi:10.1093/
scan/nsaa135)
125. Hoehl S, Fairhurst M, Schirmer A. 2020 Interactional
synchrony: signals, mechanisms and benefits. Soc.
Cogn. Affect. Neurosci. 16,5–18. (doi:10.1093/scan/
nsaa024)
126. Chang A, Kragness HE, Tsou W, Bosnyak DJ, Thiede
A, Trainor LJ. 2020 Body sway predicts romantic
interest in speed dating. Social Cogn. Affect.
Neurosci. 16, 185–192. (doi:10.1093/scan/nsaa093)
127. Kruppa JA, Reindl V, Gerloff C, Oberwelland Weiss E,
Prinz J, Herpertz-Dahlmann B, Konrad K, Schulte-
Rüther M. 2020 Interpersonal synchrony special
issue. Brain and motor synchrony in children and
adolescents with ASD—a fNIRS hyperscanning
study. Soc. Cogn. Affect. Neurosci. 16, 103–116.
(doi:10.1093/scan/nsaa092)
128. Heggli OA, Konvalinka I, Cabral J, Brattico E,
Kringelbach ML, Vuust P. 2020 Transient brain
networks underlying interpersonal strategies during
synchronized action. Soc. Cogn. Affect. Neurosci. 16,
19–30. (doi:10.1093/scan/nsaa056)
129. Jiang J, Zheng L, Lu C. 2020 A hierarchical model
for interpersonal verbal communication. Soc. Cogn.
Affect. Neurosci.16, 246–255. (doi:10.1093/scan/
nsaa151)
130. Nguyen M, Vanderwal T, Hasson U. 2019 Shared
understanding of narratives is correlated with
shared neural responses. Neuroimage 184,
161–170. (doi:10.1016/j.neuroimage.2018.09.010)
131. von Zimmermann J, Richardson DC. 2016 Verbal
synchrony and action dynamics in large groups.
Front. Psychol. 7,1–10. (doi:10.3389/fpsyg.2016.
02034)
132. Gordon I, Gilboa A, Cohen S, Milstein N, Haimovich
N, Pinhasi S, Siegman S. 2020 Physiological and
behavioral synchrony predict group cohesion and
performance. Sci. Rep. 10,1–12. (doi:10.1038/
s41598-020-65670-1)
133. Kragness HE, Cirelli LK. 2020 A syncing feeling:
reductions in physiological arousal in response to
observed social synchrony. Soc. Cogn. Affect.
Neurosci. 16, 177–184. (doi:10.1093/scan/nsaa116)
134. Hasson U, Frith CD. 2016 Mirroring and beyond:
coupled dynamics as a generalized framework for
modelling social interactions. Phil. Trans. R. Soc. B
371, 20150366. (doi:10.1098/rstb.2015.0366)
135. Nguyen T, Schleihauf H, Kayhan E, Matthes D,
Vrtička P, Hoehl S. 2020 Neural synchrony in
mother–child conversation: exploring the role of
conversation patterns. Soc. Cogn. Affect. Neurosci.
16,93–102. (doi:10.1093/scan/nsaa079)
136. Shamay-Tsoory SG. 2021 Brains that fire together
wire together: interbrain plasticity underlies
learning in social interactions. Neuroscientist 28,
543–551. (doi:10.1177/1073858421996682)
137. Csibra G. 2006. Social learning and social cognition:
the case for pedagogy. Atten. Perform. 21,
249–274.
138. Tanaka F, Cicourel A, Movellan JR. 2007 Socialization
between toddlers and robots at an early
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
11

childhood education center. Proc. Natl Acad. Sci.
USA 104, 17 954–17 958. (doi:10.1073/pnas.
0707769104)
139. García AM, Ibáñez A. 2014 Two-person
neuroscience and naturalistic social
communication: the role of language and linguistic
variables in brain-coupling research. Front.
Psychiatry 5(AUG), 1–6. (doi:10.3389/fpsyt.2014.
00124)
140. Tremont G, Halpert S, Javorsky DJ, Stern RA. 2000
Differential impact of executive dysfunction on
verbal list learning and story recall. Clin.
Neuropsychol. 14, 295–302. (doi:10.1076/1385-
4046(200008)14:3;1-P;FT295)
141. Murgiano M, Motamedi Y, Vigliocco G. 2021
Situating language in the real-world: authors’
reply to commentaries. J. Cogn. 4,1–4. (doi:10.
5334/JOC.181)
142. Pan X, Hamilton AFdC. 2018 Why and how to
use virtual reality to study human social interaction:
the challenges of exploring a new research
landscape. Br. J. Psychol. 109, 395–417. (doi:10.
1111/bjop.12290)
143. Spunt RP, Adolphs R. 2017 A new look at domain
specificity: insights from social neuroscience. Nat. Rev.
Neurosci. 18, 559–567. (doi:10.1038/nrn.2017.76)
144. Sigman M, Peña M, Goldin AP, Ribeiro S. 2014
Neuroscience and education: prime time to build
the bridge. Nat. Neurosci.17, 497–502. (doi:10.
1038/nn.3672)
145. Vigliocco G, Perniss P, Vinson D. 2014 Language as
a multimodal phenomenon: implications for
language learning, processing and evolution. Phil.
Trans. R. Soc. B 369, 20130292. (doi:10.1098/rstb.
2013.0292)
146. Ward JA, Richardson D, Orgs G, Hunter K, Hamilton
A. 2018 Sensing interpersonal synchrony between
actors and autistic children in theatre using wrist-
worn accelerometers. In Proc. Int. Symp. on
Wearable Computing, ISWC,Singapore, China, 8–12
October 2018, pp.148–155. (doi:10.1145/3267242.
3267263)
royalsocietypublishing.org/journal/rstb Phil. Trans. R. Soc. B 378: 20210357
12