Educational Psychologist

ISSN: 0046-1520 (Print) 1532-6985 (Online) Journal homepage: www.tandfonline.com/journals/hedp20

The Social Brain, Language, and Goal-Directed

Collective Thinking: A Social Conception of
Cognition and Its Implications for Understanding
How We Think, Teach, and Learn

Neil Mercer

To cite this article: Neil Mercer (2013) The Social Brain, Language, and Goal-Directed
Collective Thinking: A Social Conception of Cognition and Its Implications for Understanding
How We Think, Teach, and Learn, Educational Psychologist, 48:3, 148-168, DOI:
10.1080/00461520.2013.804394

To link to this article: https://doi.org/10.1080/00461520.2013.804394

Published online: 14 Jun 2013.

Submit your article to this journal

Article views: 13586

View related articles

Citing articles: 29 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=hedp20
EDUCATIONAL PSYCHOLOGIST, 48(3), 148–168, 2013
Copyright 
C Division 15, American Psychological Association
ISSN: 0046-1520 print / 1532-6985 online
DOI: 10.1080/00461520.2013.804394

The Social Brain, Language,

and Goal-Directed Collective Thinking:
A Social Conception of Cognition and Its
Implications for Understanding How We
Think, Teach, and Learn
Neil Mercer
Faculty of Education
University of Cambridge

In recent years, researchers in evolutionary psychology and anthropology have proposed that
the distinctive nature of human cognition is the product of our evolution as social beings; we
are born with “social brains” that enable us to manage complex social relationships in ways
other animals cannot. I suggest that the concept of the social brain is potentially useful for
understanding the dynamic, iterative relationship between individual and collective thinking,
and the role of language in mediating that relationship. However, I argue that its current con-
ceptualization is too narrow and individualistic; the concept should be redefined to take account
of the distinctive human capacity for thinking collectively. I suggest that Vygotskian socio-
cultural theory offers a framework for this reconceptualization, which would then enable us to
achieve a better understanding of the relationship between “intermental” (collective) activity
and “intramental” (individual) intellectual activity and development. I use this theoretical base
to propose three explanations for the observed effects of collaborative learning on individual
learning and development.

In 2010 the United Kingdom’s Royal Society of Arts (RSA) the series did not allow any opportunity for the development
organized a seminar entitled The Social Brain and the Cur- of these ideas. For example, there were implications for how
riculum, which brought together researchers from evolution- we could integrate different fields of study, and for how such
ary psychology, neuroscience, computer-related studies, and a unified approach might address educational issues. One of
educational research to share their views on this matter. One my motivations for writing this article is to try to take these
of the key issues that emerged was “The brain’s sociality: issues further.
The brain’s constant orientation to others and the creation of The concept of the “social brain” was introduced by the
meaning through brains interacting, rather than through the evolutionary anthropologist Dunbar (1998) and has since
operation of individual internal cognition” (RSA, 2010, p. 2). generated some interesting and imaginative discussion about
In that seminar, the neuroscientist Frith took an evolutionary the relationship between individual mental capacities and
psychology stance to argue that the human brain is designed social interaction. It represents the view that human intelli-
to enable people to adjust sensitively to one another’s per- gence has an intrinsically social quality, in that evolution has
spectives and emotions, so as to enable cooperative activity equipped us with brains that enable us to operate effectively
from which a whole community can benefit (see Wolpert & in complex social networks. By linking cortical functioning,
Frith, 2004). As a contributor to the seminar series, I was individual thinking, and social interaction, it seems to me
heartened to discover that neuroscientists and evolutionary that this concept could usefully bring together research in
psychologists were interested in social aspects of cognition neuroscience, evolutionary psychology, developmental psy-
and in addressing educationally relevant questions. However, chology, social psychology, and educational psychology, as
well as connecting with other fields such as sociolinguistics
Correspondence should be addressed to Neil Mercer, Faculty of Edu-
and linguistic philosophy. However, I argue that if it is to
cation, University of Cambridge, 184 Hills Road, Cambridge, CB2 8PQ, fulfill that role, the social brain concept needs further devel-
United Kingdom. E-mail: nmm31@cam.ac.uk opment, because the social nature of human cognition has
 THE SOCIAL BRAIN 149

not been properly recognized in its development so far. The have shaped, and been shaped by, the struggle of our ances-
main focus of interest in the social brain has been on how tors to survive and eventually become the dominant species.
individuals cope with the informational and emotional com- In early stages of its development, evolutionary psychology
plexity of social life, so as to maximize the achievement of invoked a strongly individualistic notion of human cognition,
their personal goals. The concept has been strongly linked to explaining the origins and nature of human thinking in terms
evolutionary theory, but only through a narrow focus on the of the selective advantage individuals would gain over other
survival value for an individual of being able to understand individuals through thinking and behaving in certain ways,
the behavior and motives of others. No account has been which would give them and their offspring a competitive
taken of the potential survival value of a cognitive capacity advantage over others. The focus thus tended to be on com-
for collaboration. Yet one reason why people engage socially, petition between humans, with the success of the species as a
which can also be linked to human evolutionary success, is whole being explained through the proportionally greater sur-
so that they can think collectively in order to pursue com- vival of the offspring of more successful individuals. Thus
mon goals. I suggest that our brains are “social” in that they Miller (1999) argued that young men produce more “dis-
have been designed, through evolution, to enable us to reason plays” of art, music, and other cultural products than women
together and get things done. or older men because they are most highly motivated to com-
Our evolved capacity for collective thinking also enables pete for mates. At its crudest, as sometimes represented in
each new generation to benefit from the past experience of the press or in media discussions, this approach might seem
their community, which has given our species survival ad- to be based on doubtful, analogical comparisons between the
vantages over competitors. It is widely recognized by psy- ways males of other species aggressively compete and the
chologists working in developmental and educational fields social behavior of men in modern society; but this would do
of enquiry that children learn to make sense of society and a great disservice to the scholarly work in this field. How-
their environment by being engaged in dialogue by their ever, evolutionary psychology has had a strong association
carers, drawn into collective activities, and guided in ways with the kind of evolutionary account offered in Dawkins’s
of reasoning about experience. The potential capabilities of (1976) book The Selfish Gene, whereby altruistic and collab-
each child’s social brain are developed through social in- orative behavior is expected to be limited to closely related
teraction. However, this recognition does not seem to have individuals and interaction with nonrelatives is perceived as
permeated discussions of the social brain within cognitive basically competitive. Thus one evolutionary psychologist
and evolutionary psychology. The crucial role of language writes, “It is not unreasonable to hypothesise that humans
in such processes has also not yet been properly accom- have evolved xenophobic fears of strangers and outgroup
modated within those discussions. I propose a way that an members” (Buss, 2001, p. 966). There have been philosoph-
evolutionary account of human origins can be linked to our ical criticisms of such highly individualistic, competitive ap-
understanding of the psychological mechanisms that under- plications of Darwinism (e.g., Midgeley, 2010), and recently
pin these aspects of human life, and in particular to the social within evolutionary psychology (and the closely related field
and cognitive processes involved in education. To do so, a of evolutionary anthropology) a recognition of the intrinsi-
theoretical framework is needed for dealing with the rela- cally complex nature of human social life, and the cogni-
tionship between collective thinking and the development tive demands this makes of people, has begun to influence
of individual cognition—between the “intermental” and the accounts of how and why the human brain has evolved in
“intramental”—and I suggest this could best be achieved the way that it has. As I explain, some evolutionary scholars
by taking a sociocultural perspective based upon Vygotsky’s have begun to describe the distinctive nature of human cogni-
work (Vygotsky, 1962, 1978). On this basis, I offer three tion in terms of our ability to operate within complex social
possible explanations of how intermental activity might in- relationships. This links with other fields of psychological
fluence intramental development and then relate these to find- investigation, and not only with social psychology. For ex-
ings from several fields of research. A broader conception of ample, researchers in neuroscience claim to have identified
the social brain emerges from this discussion, which has im- features of neural function which show that we are sensi-
plications for future research in education and other fields, as tive, sometimes unconsciously, to very subtle “social signals”
I explain toward the end of the article. But first I consider how that enable us to respond to the intentional actions of oth-
the concept of the social brain has emerged from research in ers. This evidence has been used by some evolutionary psy-
evolutionary studies and neuroscience. chologists (and anthropologists) to support their claims that
the design of the human brain supports an inherently social
function.
EVOLUTIONARY PSYCHOLOGY, Dunbar (1998), a leading scholar in the field of evolution-
NEUROSCIENCE, AND THE SOCIAL BRAIN ary psychology/anthropology, has been prominent in setting
out the case for the brain’s “prosociality.” He commented that
Researchers within the relatively new field of evolutionary the conventional wisdom over the past 160 years in cogni-
psychology have striven to explain how the ways we think tive psychology and neuroscience has been that the prime
150 MERCER

function of the human brain is to enable individuals to pro- rons having social-cognitive functions, which include under-
cess factual information about the world as effectively as standing and predicting the actions of others, and language
possible. The unstated implication has been that the evolu- processing.
tionary struggle for survival would favor individuals with the Even if mirror neurons are not hard-wired, it seems that
best sensory, information-processing, and memorizing abil- our brains have been evolutionarily designed for living in a
ities. It is only relatively recently that it has been suggested complex society. As Grist (2009) explained,
that the nature and size of the human brain might also re-
flect the survival advantages of a more subtle kind of mental
capacity, that of being able to make sense of complex so- We become aware of others because our brains can apply
cial relationships. This has led evolutionary psychologists to “theory of mind”—this is the cognitive endeavor of attribut-
propose that our brains have evolved to be able to perform ing thoughts to others. Part of theory of mind consists in
specific “cognitive feats such as calculating the status tra- thinking about what other people are thinking about other
people—“what does Jane think about Tom’s behavior to-
jectories of oneself and others in the group, and modeling
wards Pablo, given that Pablo is upset about his father’s
the consequences of the injury or death of a kin member”
illness?” This is a very complicated kind of cognition and
(Buss, 2001, p. 968). New fields of research have since been is, as far as we know, unique to humans. The social brain
defined (“social neuroscience” and even more specifically hypothesis in evolutionary anthropology contends that hu-
“developmental social cognitive neuroscience”; see Zelazo, man brains have evolved to be as big as they are so that
Chandler, & Crone, 2009) to pursue this perspective. There we can think about and manage our relationships with other
has been a particular interest in how we notice and respond people. (p. 44)
to the subtle social signals of people we interact with, even if
we are not consciously aware of doing so. Research has iden-
tified “mirror neurons,” which become active not only when The claims made by Grist and by Frith and Singer can be
a primate carries out an action but also when they observe related not only to the well-established concept of theory of
a community member carrying out the same action. Most mind (Premack & Woodruff, 1978) but also to that of so-
studies have involved monkeys, but recently neuroscientists cial cognition (as discussed, e.g., by Fiedler & Bless, 2001).
have reported identifying the activity of mirror neurons in hu- The basic claim made by evolutionary anthropologists and
mans (Mukamel, Ekstrom, Kaplan, Iacoboni, & Fried, 2010). psychologists is that advanced skills in social cognition—the
However, this “mirror” activity apparently arises only with ability to infer emotion and interpret social behavior—would
the observation of intentional actions. It had already been be differentially selected for among our early ancestors. A
well established by social psychologists that when people in- species that possessed such skills would be more able to or-
teract, they tend to reflect each other’s gestures and postures ganize larger and more complex social groups. Individuals
(e.g., Chartrand & Bargh, 1999), and so this seems to offer with the most developed skills within a community would
a neural correlate for such findings. Commenting on the sig- have advantages in achieving their goals (including repro-
nificance of this, Frith and Singer (2008) said, “Through the ductive goals) because they could make sense of such social
automatic activation of mirror systems when observing the complexity (Dunbar, 1998) to help them compete with ri-
movements of others, we tend to become aligned with them vals (Buss, 2001) and so ensure that more of their offspring
in terms of goals and actions” (p. 3875). survived.
However, doubts have been expressed about the evolu- These discussions of social cognition by evolutionary psy-
tionary account of the origins of mirror neurons in humans, chologists represent a significant shift in perspective from the
for example by the psychologist Heyes (2010). She pointed individual to the social. But they still tend to focus only on
to the lack of evidence that the human brain has single neu- individual cognition and competition, rather than joint intel-
rons that are biologically programmed to discharge when lectual activity and collaboration. They also tend to describe
an action is performed or observed, and argued that rather the relevant skills as just being “there,” inherent in the natu-
than being “hard-wired” to respond in that way, the evidence ral capabilities of the individual brain. From the perspective
reflects the effects of our associative learning, as we corre- of an educational researcher, this seems simplistic and even
late the experiences of performing and observing the same inaccurate. Theory of mind can be considered a distinctive
actions. She cited studies which show that experience modi- capacity within the broader notion of metacognition, in that
fies mirror responses (so that, e.g., pianists respond more to it involves making assessments not only of one’s own state of
observations of piano-playing finger movements than non- knowledge, intentions, and so on, but also of those of other
pianists), supporting that view. But whereas Heyes is un- people. Research in educational and developmental psychol-
convinced that hard-wired mirror neurons should be given ogy suggests that, like other metacognitive and social skills,
a major role in the evolutionary development of the social theory of mind develops through practice and guidance as
brain, she conceded that her associative learning explana- children become involved in dialogues with other people, as
tion of their origins is still compatible with mirror neu- I describe in a later section.
 THE SOCIAL BRAIN 151

INDIVIDUAL AND COLLECTIVE THINKING One does not have to be convinced that “telling and show-
ing” is a uniquely human behavior to accept that our species
Making a link with evolutionary psychology, the linguistic has developed it to unique levels of sophistication. Through
philosophers Mercier and Sperber (2011) proposed that the becoming sensitive to the limits of one another’s levels of un-
human capacity for reasoning should primarily be understood derstanding, we are especially able to engage effectively in
as a competitive social mechanism, whereby we each strive interactions whereby one person helps another to learn. We
to persuade others to comply with our preferred courses of differ from our primate relatives by having what Hermann,
action. Although this is interesting and plausible, it shares Call, Hernandez-Lloreda, Hare, and Tomasello (2007) called
the weakness of the evolutionary psychology accounts just “cultural intelligence,” by which they mean the ability to
discussed by explaining human survival only in terms of make sense of a complex, shifting social world and operate
individual people negotiating the social world and pursuing as a fully functioning member of it. Their comparative study
their own agendas. This may provide part of the explanation found that young children (1 year old) “far outstripped” chim-
for the origins and nature of the cognitive capacity of our panzees and orangutangs on tests of social cognition. But we
species, but it fails to acknowledge one of the most important have evolved the ability to do more than just make sense of
functions of our social-cognitive capabilities, which is that a complex society and interact successfully within it. To an
we are able to engage together in goal-orientated, knowledge- extent that is impossible for other primates, we are able to
building, and problem-solving activities. We do not only use not merely interact but to “interthink” (Littleton & Mercer,
reasoning as an individual weapon to resist other people’s 2013; Mercer, 2000). Through the use of language and other
agendas; we also use it in dialogues to find the best possible modes of representation, we can link our individual minds
solutions to the problems we jointly encounter. Frith and to create a powerful problem-solving tool. It is likely to be
Singer (2008) hinted at this when they said, “When joint this, rather than individualized competition, that has ensured
action requires cooperation, shared representations of task the dominance of our species. As an emergent species, our
requirements and goals are very important in order to achieve ancestors would compete well with other species because
better performance. Such sharing is referred to as common they could use language and other modes of representation
knowledge” (p. 3876). to define common problems and plan in advance how to deal
The creation of “common knowledge” is an interactive, with them. They could also reflectively discuss the success
complex, discursive process, as educational research has or otherwise of those actions and plan again accordingly.
shown (Edwards & Mercer, 1987/2012), and there is more to By creating abstract representations of experience together,
collective thinking than sharing knowledge. Theory of mind they could use them to explore the past and the future. Older
capabilities allow us not only to assess others’ emotional members of a community could induct each new generation
states or try to exert a social influence on them: They also into their complex society’s way of getting things done and
enable us to make assessments about what knowledge we share with them the insights of past collective thinking. Their
have in common with another person, and judge their levels children would learn ways of using language for represent-
of understanding or skill in relation to particular topics or ing the physical world and for reasoning about it. Collective
tasks. In ways that are just not possible for other species, we intellectual activity thus would become to have a significant
use interactions to continually refine our judgments of the influence on the development of individual cognition. This
relative states of knowledge and understanding of other peo- perspective on the relationship between the social and the in-
ple (Jeong & Chi, 2007). Both assessments of what others dividual may seem obvious to educational and developmental
think and know and metacognitive reflections on our own psychologists, but it has not yet figured in discussions of the
thought processes are involved. These allow us to jointly ac- concept of the social brain.
tivate the practical cycles of planning, acting, reflecting, and Why has there been so little recognition of the impor-
replanning by which we solve problems, share knowledge, tance, or even the existence, of collective thinking in evolu-
and construct new joint understandings. Such mental capac- tionary psychology? It may be because cognitive psychology
ity distinguishes us from other primates, as Tomasello (2008) has generally upheld the commonsense view that thinking
explained: just goes on in individuals’ heads, and topics such as prob-
lem solving, learning and remembering have been studied
accordingly. It is one of the most prestigious fields of psy-
The great apes—chimpanzees, bonobos, gorillas and chological enquiry, and so it may simply be that this per-
orangutans—communicate almost exclusively for the pur-
spective has been carried over into evolutionary psychology.
pose of getting others to do what they want. Human infants,
The study of collective thinking, which includes how people
in addition, gesture and talk in order to share information with
others–they want to be helpful. The free sharing of informa- solve problems in groups or how one person helps another to
tion also creates the possibility of pedagogy–in which adults learn, has been pursued quite separately by developmental,
impart information by telling and showing, and children trust educational, and discursive psychologists (see, e.g., Cazden,
and use this information with confidence. Our nearest primate 2001; Howe, 2010; Mercer & Littleton, 2007; Middleton &
relatives do not teach and learn in this manner. (p. 1) Edwards, 1990). Studies of how people think alone and how
152 MERCER

they think together have never been united. I suggest that this to each other, and it is vital to human life that we do so.
division is an obstacle to developing a proper understanding But as a generalization, Pinker’s claim is patently false, as
of how humans think. Cognitive achievements and activities any teacher who has instructed a class of children on what
do not, of course, take place only in social settings, and I to do will know when they all begin to do rather different
offer no arguments against the study of individual learning, things. Language’s power as a tool for creative, collective
remembering, or problem solving. But unless researchers thinking partly lies in the possibility that listeners may each
recognize how the inherently social nature of our species interpret a speaker’s words in rather different ways. Scholars
shapes our cognition and its development, and the fact that are still interpreting the words of orators, philosophers, play-
we frequently learn, reason, and solve problems collectively, wrights, and poets, hundreds of years after those words were
the accounts they generate will misrepresent how people expressed, and offering new insights into the human condition
actually think and learn in real life. as a result. As Vygotsky’s contemporary and compatriot, the
literary scholar Bakhtin (1981), put it—or rather, as I inter-
pret what he wrote—the words we hear or read do not simply
THE COGNITIVE FUNCTIONS OF LANGUAGE activate a mental dictionary, they generate dialogic responses
in our own minds as we use our existing knowledge to make
Dunbar (1998) argued for the crucial role of language in the sense of them. Vygotsky (1962) himself argued that language
operation of the social brain as follows: is both a cultural tool and a psychological tool, linking the
“intermental” and the “intramental” in a reciprocal relation-
For humans, one important aspect of [theory of mind] con- ship, and so is inextricably bound up with the development
cerns its relevance to language, a communication medium and application of more advanced forms of reasoning. Thus
that crucially depends on understanding interlocutors’ men- Vass and Littleton (2010) suggested that “interpsychological
tal states or intentions. The kinds of metaphorical uses of thinking is a prerequisite for intrapsychological thinking: it is
language that characterize not only our rather telegraphic ev- through speech and action with others that we learn to reason
eryday exchanges (in which “you know what I mean?” is a and gain individual consciousness” (p. 107).
common terminal clause) but also lies at the very heart of the The view that language is thoroughly integrated with other
metaphorical features of language. (p. 189)
important nonlinguistic aspects of thinking is, of course, at
odds with some influential views on language and cognition,
This is in accord with my own argument even if, once again, notably of those who remain committed to a view of language
the focus is on the assessment of emotions and intentions, as a discrete cognitive “module” or capacity (e.g., Pinker,
rather than on the collective representation of experience and 1994, 2007). But emerging findings from neuroscience sup-
the joint pursuit of solutions to problems. It could well be, port the more integrated view, as they suggest that mental
as some have argued (Levinson, 2006), that our evolving abilities associated with some nonlinguistic skills, such as the
capacity for reading the intentions of others underpinned appreciation of rhythmic patterns and structures in music, are
the emergence of language; but this does not explain the also involved in language abilities. Thus Goswami (2009, p.
significance of language as both a cognitive and cultural tool. 182), who has studied both linguistic and musical aspects of
Language is the prime means at our disposal for making a cognition, cited research (Abrams, Nicol, Zecker, & Kraus,
dynamic assessment of shared understanding and developing 2008) on children’s cortical responses to speech, which has
it, and so has a central, integrated position in enabling human shown that neural tracking of the “speech envelope” (the
cognition to be both individual and social: rhythmic, syllabic sequence of speech sounds) is accurately
monitored by the right, not the left, hemisphere. Review-
By participating in the conversations that accompany and ing evidence from neuro-imaging research, Patel (2003) ar-
grow out of the everyday activities in which he or she is
gued that syntactic features shared by music and language
involved together with other members of the culture, the child
learns to use the semiotic tool of language, which enables
are processed by the same parts of the brain, and therefore
him or her to “connect” with other people; at the same time, that neuroscientists should no longer focus on language “in
and by virtue of the mediating role that conversation plays isolation.”
in these activities, the child simultaneously “assimilates the We might also consider here research on bilingualism for
experience of humankind”, as this is encoded in the semantic what it tells us about the relationship between language and
system of that culture’s language. (Wells, 1999, pp. 19–20) general cognitive functioning. Up to the middle of the 20th
century, reviews of research tended to conclude that growing
There is more to this than humans using language to share up bilingually causes some cognitive deficits, through some
information accurately, as some cognitive/evolutionary psy- kind “overloading” of cognitive capacities (Darcy, 1963;
chologists have claimed, such as Pinker (1994), when he said, Macnamara, 1966). In contrast, more recent research has
“Simply by making noises with our mouths we can reliably shown that there are general cognitive benefits from grow-
cause precise new combinations of ideas to arise in each ing up bilingual (Diaz, 1983; Grosjean, 2010). For example,
other’s minds” (p. 15). Indeed we can pass on information it seems that bilingual children perform better in nonverbal
 THE SOCIAL BRAIN 153

problem-solving tasks, which depend on selective attention but in any case this theory’s potential for making the link be-
or inhibitory control, as these controlling abilities have been tween social processes and individual thinking is there to be
enhanced through exercising linguistic choices between dif- exploited. Sociocultural theory provides a theoretical basis
ferent languages (Bialystok & Feng, 2010). This advantage for the primacy of language as a cultural and cognitive—and
seems to continue throughout the bilingual’s lifespan. The hence educational—tool. A model of the social brain in which
current view seems to be that bilingualism may offer both language is a fully integrated component is compatible with
cognitive benefits and disadvantages. This is apparently also two Vygotskian claims that (a) the development of individual
supported by evidence from neuro-imaging research (Bia- psychological (intramental) functions is commonly preceded
lystock et al., 2005). Bialystok and Feng (2010) offered this by involvement in related social (intermental) activity, and
summary: “The picture emerging from these studies is a com- (b) the acquisition of language has a transformative effect on
plex portrait of interactions between bilingualism and skill the nature of an individual’s thinking. Sociocultural theory
acquisition in which there are sometimes benefits for bilin- does not, as some of its critics have claimed, suggest that
gual children, sometimes deficits, and sometimes no conse- “everything that can be thought can be thought in language,
quence at all” (p. 121). It is not of great significance to us here and everything that can be represented, can be represented in
as to whether the effects of bilingualism on nonverbal reason- language” (Jewitt, 2008, p. 7). There is no need to make any
ing are positive or negative: The key point is that this research such extreme claims to justify a special focus on language,
indicates that language skills are integrated with nonverbal as I hope to show.
reasoning skills, and that language experience is linked to Empirically, it has been known for some time that the
the development of those skills. As one bilingualism scholar quality of children’s language experience in the early years
wrote, “Language attributes are not separated in the cogni- is a good predictor of their subsequent educational achieve-
tive system, but transfer readily and are interactive” (Baker, ment (Hart & Risley, 1995; Wells, 1986, 2009). A good ex-
2006, pp. 168–169). Evidence thus increasingly supports the planation for this involves children’s induction into ways of
view that language is involved with both specific and non- using language for explaining and reasoning. Education is
specific aspects of brain function, and so is a fully integrated popularly understood as the transmission and acquisition of
component of human cognition. If we want to understand facts and skills, but educational success requires the abil-
the functioning of the social brain, in the broader sense of ity to justify opinions, analyses, solutions, and conclusions.
the term, and its development through childhood, I therefore Although arguments can sometimes be presented through
suggest we need to recognize the crucial role that language other communicative modes (such as the use of mathemat-
plays in enabling us to think collectively and individually. ical notation, and by physical demonstration in science or
music), language is essentially involved in all academic sub-
jects. Moreover, achieving competence in specific subjects
THE ROLE OF LANGUAGE IN DEVELOPMENT requires the use of specialized discourses, or genres, of sub-
AND LEARNING ject communities; and those genres are not mere jargon,
but cultural tools designed for pursuing collective scholar-
In recent years, one major influence on classroom research ship and enquiry (Christie & Martin, 1997; Kress, 1987;
has been the emergence of a sociocultural theory of education Swales, 1990). The linguist Martin (1993) described them
and cognitive development derived from Vygotsky’s work (as as rule-governed, goal-orientated social processes embodied
described, e.g., by Daniels, 2001, 2008; it is also known as in language. They represent ways that individual thinking is
cultural-historical activity theory: van Oers, Elbers, van der made accountable to the normative rules of collective activity
Veer, & Wardekker, 2008). Its basic premise is that human within specific communities of thinkers; and fluency in the
intelligence is essentially social and cultural, and that the appropriate genres is a requisite for full admission to those
relationship between social activity and individual thinking communities. Thus the educational researcher Pea (1993) ar-
underpins cognitive development. As Vygotsky (1978) put gued, “Expertise is defined dynamically through continuing
it, “Human learning presupposes a specific social nature and participation in the discourse of a community, not primarily
a process by which children grow into the intellectual life through the possession of a set of problem-solving skills and
of those around them” (p. 88). This has inspired empirical conceptual structures” (p. 271). As Lemke (1990) famously
examinations of how social interaction influences individual put it, science education should enable students to become
learning, problem solving, and representations of knowledge. fluent speakers of science. The same principle applies in a
One leading sociocultural psychologist suggests that “all too great many nonacademic knowledge domains (Barton, 2009;
often the focus of sociocultural research has been on inter- Goodman, 1996; Lave & Wenger, 1991). Even if one is con-
mental (social, interactional) processes per se, to the neglect cerned with the development of understanding the content
of explanations of how these intermental processes forge the of a field of knowledge, rather than the manner in which
intramental processes that sit at the heart of cognitive devel- it is expressed, language has a crucial role. In the debate
opment” (Rojas-Drummond, 2009, p. 241). My own view is about key factors in enabling students’ conceptual change
that interest in intramental processes has tended to dominate; in science education, Treagust and Duit (2008a), leading
154 MERCER

researchers in that field, have agreed with classroom re- ment of individual children (see, e.g., Howe, 2010; Slavin,
searchers such as Wells (2008) that dialogue is commonly 2009). I pursue that interest in the next section. But that is
the “dynamic motor” for such change. Their own examples not the only reason for studying collaborative learning. We
of transcribed dialogue (Treagust & Duit, 2008b) illustrate should recognize the importance of children learning to be
this well. good collaborative problem solvers. Understanding the role
A sociocultural explanation of the role of language in indi- of language is vital for analyzing more symmetrical, collab-
vidual and collective cognition can be based upon Vygotsky’s orative types of learning and problem solving. In almost all
proposed link between the intramental and the intermental. It joint problem-solving tasks it is impossible to collaborate
begins with children learning and using the functional forms properly without talking with a partner (or using the writ-
of language they hear around them. They practice these forms ten “talk” of electronic communications). As the eminent
in “pretend play,” acting out social roles such as “teacher” and classroom researcher Barnes (2010) put it,
“pupil” (Elbers, 1994). They “ventriloquate” the ways they
hear adults making sense of the world as they try to make It is worth considering what is implied when, after someone
sense of the world themselves (Maybin, 2006). Those social has asked you to make a decision, you reply, “I’d like to talk
modes of communication offer templates for individual sense it over first.” It’s not that you expect that the talk will give you
making. If learning to participate in a subject like science new information. It’s rather that you know from experience
involves learning to “speak” it, then learning to think like a that the matter in question can often appear quite differently
as a result of talking it through. The situation and its prior-
scientist must involve some internalization of the subject dis-
ities and implications can take a different shape, a different
course as a tool for reasoning about the relevant phenomena. meaning. You may see connections that were not immedi-
Thus the genres of various discourse communities provide re- ately apparent, or realise that some of the options might have
sources for organizing the process of individual thinking. One results that need to be considered. In that case, the talking
strength of Vygotsky’s model, not always well appreciated, is is contributing to your understanding by reshaping what you
that he envisaged the intermental–intramental relationship as already in a sense “know.” This provides a useful model for
enabling a kind of spiral intellectual development: Members the function of talk in learning, in that some kinds of talk
of a new generation gain cultural understanding from their contribute to understanding without necessarily adding new
elders through the forms and content of dialogue, which then material. (p. 7)
empowers them to make new, original insights, which are
then shared and so enrich the culture of the community. It is
through using the resources of specialized language genres THE INTERMENTAL AND THE INTRAMENTAL:
that subjects, disciplines, trades, and other fields of human EXPLAINING THE EFFECTS OF COLLECTIVE
endeavor can persist and grow, within and across genera- REASONING ON INDIVIDUAL REASONING
tions. In evolutionary terms, the emergent capacity to refine
the tool of language into a diversified toolkit would offer The distinctive quality of the social brain that has enabled the
adaptive advantages in dealing with changing environmental success of our species is, I have suggested, the capacity to
circumstances. link the intermental and the intramental. We have evolved to
Understanding the role of language in children’s intellec- become able not only to make sense of complex social rela-
tual development requires an analysis of how the intermental- tions as individuals, but also to design and use those relations
intramental relationship is embodied in social interaction. to create cognitive capability that transcends individual lim-
The process whereby an expert guides a novice toward new itations. Moreover, we use collective cognition to enable the
understanding and competence is one of the basic, key fea- cognitive development of each new generation. This is the
tures of human society; it is a manifestation of the powers of essence of the Vygotskian claim that social activity shapes
the social brain. It depends on the establishment and main- psychological development. But the nature of that influence,
tenance of an intersubjectivity that, at least in formal ed- regarding how elements of collective, social thinking affect
ucational settings, is normally mediated through language. subsequent individual thinking, has never been made clear
Instructing a learner, or providing useful feedback on their ef- (in Vygotsky’s work or that of any other sociocultural theo-
forts, is very difficult without some kind of spoken or written rist). This is a potentially very complex matter, and so for the
dialogue. It is interesting and useful to highlight the multi- sake of clarity I choose to approach it in the following way.
modality of classroom education, by which gestures, nonlin- First, I consider a specific aspect of cognitive development:
guistic symbols, images, artifacts, and physical settings can the development of children’s reasoning. Second, I consider
also shape learning processes (e.g., Jewitt, Kress, Ogborn, & how one kind of collective thinking, collaborative problem
Tsartsarelis, 2004), but such analysis should not obscure the solving in the classroom, might stimulate the development of
prime, central role of language. an individual child’s reasoning capabilities. As I explain later,
Educational research on collaborative learning has mainly there is evidence that such joint activity can indeed, under
been motivated by an interest in if, and how, group-based certain conditions, have stimulating, beneficial effects—but
learning activities help the learning and conceptual develop- for the moment, I am concerned only with elucidating the
 THE SOCIAL BRAIN 155

hypothetical mechanisms for any such effects. I propose that their peers. “Co-construction” would also explain find-
there are three possible explanations, which might be called ings such as those of Smith et al. (2009), who reported
appropriation, co-construction, and transformation. I believe that peer discussion among undergraduates enhanced
they are worth considering in some detail, because, although their understanding of scientific concepts even when
they are not mutually exclusive, they represent a series of rel- none of the individual participants initially had the
atively stronger theoretical claims about how language-based requisite knowledge.
collaborative learning and problem solving might shape in- 3. Transformation: The experience of group discussion
dividual learning and development. could transform the nature of subsequent individual
reasoning. If the norms of discussion required reason-
1. Appropriation: Children could learn successful ing to be made explicit, and claims to be justified, the
problem-solving strategies and explanatory accounts argumentation involved in collaborative problem solv-
from each other during joint activity. Through talk- ing might promote children’s metacognitive, critical
ing, they can share relevant knowledge effectively as awareness of how they reasoned. Engaging in ratio-
they carry out a task, and explain their strategies to nal debate could also stimulate their theory of mind
each other. They could thus acquire new, useful infor- capacities, as they became more aware of the possibil-
mation and successful strategies for solving problems ity of different points of view, of how these might be
from each other and go on to apply them in any subse- set against their own assumptions and of any contra-
quent individual situation. This represents a relatively dictions generated. This would encourage a “reflective
weak claim for the influence of joint activity on the stance” and develop a capacity for intramental “dia-
development of individual thinking, because it merely logue” (Muller-Mirza & Perret-Clermont, 2009). This
identifies other people’s explanations and demonstra- would help them to become more able to assess possi-
tions of knowledge and cognitive strategies as impor- ble problem-solving strategies in a critical way, and to
tant resources for individual learning. Language plays monitor and regulate their own problem solving when
an important role here, but only in so much as it is a subsequently doing a task on their own. A child’s think-
medium for transmitting information with a fair degree ing would thus be transformed, through intermental
of accuracy from one mind to another. activity, so that they began to reason intramentally in a
2. Co-construction: By using talk to coordinate their more “dialogic” way.
mental efforts, children could not only share ideas but
also argue productively about them, to jointly construct All three explanations are compatible with current versions
new, robust, generalizable strategies together for com- of sociocultural theory. All three acknowledge, in different
pleting a task that are better than any of them would ways, the case for the special significance of language use in
have devised alone. Their improved group performance the development of reasoning. They all invoke the capacities
would reflect what has been called the “assembly bonus of the social brain for sharing information, assessing common
effect,” whereby the performance of a group is better knowledge, and planning goal-directed activity—and for en-
than that of its best member (Laughlin, Hatch, Silver, abling useful knowledge and cognitive skills to be taken up
& Boh, 2006). Similarly, Woolley, Chabrsi, Pentland, among members of a community. As they are not mutually
Hashmi, and Malone (2010) invoked the notion of “col- exclusive, all three explanations might be invoked to account
lective intelligence” when reporting a study showing for the effects of collaborative learning on individual learning
that the success of group endeavors is not strongly cor- and achievement. But although the “appropriation” explana-
related with the average intelligence of group members tion is quite prosaic, the “co-construction” and “transfor-
but is correlated with the average social sensitivity of mation” explanations are not. “Co-construction” implies the
members and the equality in distribution of their con- exercise of some collective intelligence, which can achieve
versational turn-taking. Individuals could then go on more than individuals can alone and which depends upon the
to use the new, effective strategies and solutions that quality of interaction. The “transformation” explanation is
the group had generated when faced with similar prob- the most intriguing, because it embodies Vygotsky’s claims
lems alone. This represents a stronger claim for the about the transformative effects of social experience on psy-
influence of social (intermental) activity on individ- chological development, and of the key role of language in
ual (intramental) learning and development because it shaping individual cognition. It is also in accord with philo-
locates the genesis of (at least some) effective cog- sophical arguments offered by Wegerif (2010) and others that
nitive strategies, understanding of problems and the higher forms of human reasoning are essentially dialogic,
generation of solutions in the dynamics of the “dia- meaning that the skilled thinker is able to take and consider
logic space” (Wegerif, 2007) of collaborative activity different, even conflicting, viewpoints and debate them inter-
and not just in individual heads. Individuals in a group nally. It also could be related to the claims discussed earlier
could also gain new levels of understanding through that reasoning functions best when set in argumentative con-
co-construction, through any explanations provided by texts (Mercier & Sperber, 2011). But it links argumentation
156 MERCER

with the success of collective endeavor and with individual most advanced level, a child began to use speech for self-
cognitive development, in a way which purely competitive, regulation, which was similar in form and function to the
individualized accounts of the functions of reasoning, like speech used earlier by the adult when scaffolding the child’s
Mercier and Sperber’s, cannot. activities. He therefore provided evidence, underpinned by
We might also consider if, and how, such explanations sociocultural theory, for how external regulation provides the
might relate to less socially symmetrical educational pro- resources for developing self-regulation. We know now that
cesses, such as when an adult helps a child, individually or as this process is not confined to preschool development. Chil-
a member of a class, to learn or gain new understanding. The dren as old as 9 or 10 can be observed coregulating their
simplest explanation is again one of appropriation, whereby group activity in class by invoking the earlier, authoritative
the adult simply shares relevant information with the child, instructions of the teacher (Warwick, Mercer, Kershner, &
provides instruction or demonstrates a strategy that the child Kleine Staarman, 2010). Regarding individual development,
acquires and uses. There would normally be no expectation research indicates that children’s ability to regulate, monitor,
here that the adult would gain a better understanding of the and reflect upon their problem solving correlates strongly
phenomenon under consideration, or would improve their with their success as learners and problem solvers (Veenman
level of skill in performing a task. But the process might & Spaans, 2005; Whitebread & Pino Pasternak, 2010). On the
nevertheless involve some form of co-construction, whereby basis of meta-analyses of such research, some have claimed
the adult and the child generated a new, shared conception of that the emergence of self-regulation is the main determinant
the task in hand or the topic being studied. The adult might of effective learning (Swanson, Hoskyn, & Lee, 1999; Wang,
gain new insights into the learning task through becoming Haertel, &Walberg, 1990). The effects of dialogic experience
aware of the limits of understanding and misunderstandings on the development of the ability to self-regulate could thus
of the child, whereas the child’s understanding would con- be a crucial, transformative feature of cognitive development.
verge on that of the adult. The interaction might also generate It has been widely observed that young children quite nat-
a new, clearer, and more explicit representation of the rele- urally talk aloud as they play alone, using language to regulate
vant knowledge, which was then shared by both teacher and their solitary activities. Such “egocentric speech,” as Piaget
learner. (It is a common experience for a teacher to find that (1926) first described it, is normally no longer used when we
teaching a topic to a learner who is struggling reveals the grow older; not only because its use would be seen as a mark
limits of their own understanding of it.) The child and adult of eccentricity, but (if we follow Vygotsky rather than Piaget)
would use language and other modes to establish intersubjec- because it has become internalized as the “silent speech” or
tivity (in the sense used by Wertsch, 1979) and pursue the kind “inner speech” of more mature cognition. The “transforma-
of goal-orientated, progressive, interactive process known as tion” explanation of the effects of dialogue on reasoning
“scaffolded” learning (Bruner, 1983). Subsequently, the child invokes a similar process to this process, whereby collec-
could start to use a mental representation of the adult-led, tive reasoning amongst peers, or scaffolded learning with an
scaffolded interaction in which they had been involved as a adult, acts as a template for the self-regulating mechanism
way of self-regulating their future individual activity. Their of “inner speech,” which is used in individual reasoning. As
thinking would be transformed as a result. As I go on to show, Clark (1998) put it,
this explanation has links with accounts of the development
of children’s ability to self-regulate their learning offered by When the child, confronted by a tricky challenge, is “talked
developmental psychologists and educational researchers. through” the problem by a more experienced agent, the child
can often succeed at tasks which would otherwise prove im-
possible (think of learning to tie your shoelaces). Later on,
DIALOGUE AND THE DEVELOPMENT OF when the adult is absent, the child can conduct a similar
SELF-REGULATION AND METACOGNITION dialogue, but this time with herself. (p. 66)

Wertsch’s (1979) seminal article on social interaction and Researchers have explored the ways that parents vary in how
higher mental processes made an important contribution by they use language to scaffold learning activities (e.g., Fidalgo
illustrating and explaining how the social regulation inher- & Pereira, 2005). Some have noted how parents of young
ent in adult-child dialogues provides a model for the self- children often use elaborative questions to guide discussions
regulation of individual cognition. His transcribed examples about shared experiences—and that such discussions seem
of parent-child interaction provided the kind of empirical il- to enhance children’s abilities to recall those events (Reese,
lustration that was lacking in Vygotsky’s original account. Haden, & Fivush, 1993). Observing how mothers and chil-
Moreover, Wertsch went beyond the exposition of Vygot- dren (aged 30–42 months) discussed events they had previ-
sky’s ideas by offering an original model of four levels of ously experienced, Rudek and Haden (2005) noted that
interaction between an adult “teacher” and a child working
together on a task, based on the quality of intersubjectivity the “system” for talking about mental states seems . . . to be
attained by the participants. He showed how, at the fourth, surprisingly well developed just as children enter a period
 THE SOCIAL BRAIN 157

in which there is marked changed in their understanding of The particular form of self-explanation that we have exam-
mind. . . . The use of mental terms . . . during reminiscing may ined involves asking children to explain the reasoning of
help to focus children’s attention on mental processes such another person. In particular, children are presented with a
as thinking, believing, and knowing and encourage thought problem, they advance an answer, they are given feedback
about memory and what it takes to remember. (pp. 543–544) concerning the correct answer, and then the experimenter
asks them, “How do you think I knew that?” (Siegler, 2002,
pp. 38–39)
Reviews of research on the development of metacognitive and
self-regulatory abilities by Dignath, Buettner, and Langfeldt
Siegler then went on to describe the effects on learning of this
(2008); Hattie, Biggs, and Purdie (1996); and Whitebread
interactive, dialogic technique: “The results indicated that, as
and Pino Pasternak (2010) conclude that such abilities can
hypothesised, encouraging children to explain the reasoning
be significantly improved through adult guidance, which
underlying the experimenter’s answer resulted in their learn-
inevitably involves dialogue. Typically, those interventions
ing more than feedback alone or feedback in combination
that obtained significant improvements have involved mak-
with requests to explain their own reasoning” (Siegler, 2002,
ing metacognitive and learning strategies explicit to children
p. 40). Spoken dialogue has a crucial function in the practice
and encouraging them to reflect and talk about their learn-
of the other techniques too, with references by the researchers
ing. Several pedagogical techniques have been used to do so.
involved to children being encouraged to ask questions, give
Whitebread and Pino Pasternak summarized the main types
explanations, explain themselves to another person, and so
as follows:
on. One of the researchers responsible for the development
of the self-assessment techniques included in the previous
• “Co-operative group work” (Forman & Cazden, 1985): list has commented,
A range of techniques involving children in collaborative
activities that oblige them to articulate their own under- The core activity of assessment for learning is the involve-
standings, evaluate their own performance, and be reflec- ment of learners in formative dialogue, with their teachers,
tive about their own learning. and with one another. Only through such activity can they be-
• “Self-explanations” (Siegler, 2002): An instructional come actively engaged in their own learning, and so acquire
practice that requires children to give “how” and “why” the confidence and skill needed to become effective learners.
explanations about, for example, scientific phenomena or (Black, 2009, p. 5)
the events in a story and then asks children to give expla-
nations of their own and an adult’s reasoning. Researchers into seemingly individual activities such as
• “Self-assessment” (Black & Wiliam, 1998): A range of metacognition and self-regulation increasingly recognize that
pedagogical ideas involving children’s self-assessment of cognitive processes can be embedded in social interaction,
their own learning, including, for example, children mak- and that collective metacognitive activity enables individ-
ing their own choices about the level of difficulty of tasks ual metacognition. Terms like “socially-shared metacogni-
to be undertaken and selecting their best work for reflec- tion” (Iiskala, Vauras, Lehtinen, & Salonen, 2010) and “so-
tive portfolios. cially mediated metacognition” (Larkin, 2009) have been
• “Debriefing” (Leat & Lin, 2003): A range of techniques used by researchers to describe people reflecting together
for reflecting upon an activity or piece of learning includ- about problem-solving strategies and the outcomes of ac-
ing “encouraging pupils to ask questions,” “making pupils tions. Groups are described as using reasoned discussion to
explain themselves,” and “communicating the purpose of “co-regulate” their activities (Volet, Summers, & Thurman,
lessons” (Whitebread & Pino Pasternak, 2010, p. 686). 2009).
Moreover, research on metacognition and self-regulation
Psychologists committed to an individualistic account of provides some indirect support for the most radical, “trans-
learning might see the success of these techniques as no more formation” explanation of how spoken dialogue can enable
than evidence of effective instruction; in other words, they cognitive development. It seems that researchers consider the
might feel it necessary only to invoke the appropriation ex- development of the individual ability to self-regulate a trans-
planation for children’s learning to self-regulate. Techniques formation in the quality of a child’s thinking—a step change
like “self-explanation” and “self-assessment” sound as if they in the way they are able to learn and solve problems. Their
are just things an individual does alone. But that would ig- research also supports the view that dialogue, of the right
nore the essentially interactive quality of those techniques, kind, enables this transformation. The important role of spo-
which involve conversations between experimenters/teachers ken language in that developmental process, though not al-
and children which make demands on children’s communica- ways explicitly acknowledged in research on self-regulation,
tive and theory of mind capabilities. For example, a leading is also apparent. As I go on to explain, this provides us with
researcher on the educational value of self-explanations de- a useful perspective for understanding the results of research
scribes what is involved as follows: on collaborative learning.
158 MERCER

EDUCATING THE SOCIAL BRAIN: Schulz-Hardt, Brodbeck, Mojzisch, Kerschreiter, and Frey
COLLABORATIVE LEARNING AND (2006).
COGNITIVE DEVELOPMENT Other research has suggested that for dialogue in problem-
solving groups to be productive (in terms of assisting problem
Research on collaborative learning and problem solving has solving, learning, and the development of understanding), it
been an active field since the middle of the last century. In- should have the characteristics of what my colleagues and
terventional methods have commonly been used to try to I have called Exploratory Talk (a term originally used by
improve social relations or communications among students Barnes, 1976; see also Barnes, 2008). Exploratory Talk is
and see if this improves the quality of collaborative learning, dialogue in which
and hence learning outcomes. Most research has involved
children, but some studies have been carried out with adults. partners engage critically but constructively with each other’s
As previously mentioned, most of that research has also been ideas. Statements and suggestions are offered for joint con-
concerned with the effects of collaborative learning on indi- sideration. These may be challenged and counter-challenged,
vidual learning and development, rather than on the devel- but challenges are justified and alternative hypotheses are
offered. Partners all actively participate, and opinions are
opment of collaborative problem-solving skills. It has been
sought and considered before decisions are jointly made.
found, for example, that experience of group-based reason-
(Mercer & Littleton, 2007, p. 59)
ing activities improves subsequent individual performance
of reasoning on a task (e.g., Augustinova, 2008). School-
It represents language being used not just to distribute infor-
based studies have provided convincing evidence for the ed-
mation among people (cf. Pinker, 1994) but being employed
ucational value of collaborative learning. Roseth, Johnson,
as a social mode of reasoning. Others have used the terms
and Johnson’s (2008) meta-analytic review of 148 studies in-
“transactive dialogue” (Berkowitz, Gibbs, & Broughton,
volving students aged 11 to 15 concluded that cooperative
1980) and “accountable talk” (Keefer, Zeitz, & Resnick,
learning has positive effects on academic achievement, with
2000; Michaels & O’Connor, 2002) to describe similar ways
Slavin’s (2009) review drawing similar conclusions.
of using talk effectively for collective reasoning. As Howe
We also know that some particular features of the talk
(2010) commented, research on collaborative learning does
between collaborating partners are associated with good in-
“not merely confirm that . . . transactive dialogue, exploratory
dividual learning outcomes. Howe (2009, 2010) described a
talk, or whatever can precipitate growth; it also shows that
series of related studies on collaborative learning in science
these forms of social interaction are so powerful that they
education. For example, in pairs, 8-year-olds were asked to
can sustain cognitive activity over many weeks” (p. 80).
predict whether an empty metal box, or a solid rubber ring,
The value of collaborative, group-based activity in the
would float in a tank of water. Having talked about this and
classroom has been clearly demonstrated, in relation to the
agreed on a prediction, they would then test this with real
study of various curriculum subjects (see also Mercer &
objects. The children were pre- and posttested (immediately
Sams, 2006; Sfard, 2001; Slavin, Groff, & Lake, 2009). But
after the task, and after a substantial delay of some weeks)
research on classroom-based group work embodies a para-
on their understanding of the relevant phenomena. It was
dox: It has shown the value of collaborative learning, but it
found that significantly better results on delayed posttests
has also shown that much of the group activity which goes
of learning and understanding were obtained when groups of
on in classrooms has little educational value. The relevant
children (a) were asked to seek agreement on their predictions
research was mainly carried out some time ago (e.g., Bennett
before testing them (even if they did not achieve agreement)
& Cass, 1989; Wegerif & Scrimshaw, 1997), but no evidence
and (b) worked in a group in which contrasting opinions were
has been offered more recently to suggest that the situation
expressed. Moreover, it did not seem to matter whether agree-
has improved significantly. Moreover, a recent meta-analysis
ment was actually reached, or if contrasting views were rec-
of collaborative learning approaches for developing reading
onciled. What was important was that “seeking agreement”
skills concludes
and “contrasting opinions” were features of the discussions.
In explaining these results, Howe suggested that (a) having
that not all discussion approaches are created equal, nor are
to seek agreement encourages children to pursue their dis- they equally powerful at increasing students’ high-level com-
cussions in more depth and to more certain conclusions and prehension of text. . . . It is one thing to get students to talk to
(b) unresolved contradiction between ideas during conversa- each other during literacy instruction but quite another to en-
tion particularly primes children’s metacognition—with the sure that such engagement translates into significant learning.
result that they subsequently reflect more on what they think (Murphy, Wilkinson, Soter, Hennessy, & Alexander, 2009, p.
about the phenomenon, and on the significance of their ob- 761)
servations. This is in accord with the results of some studies
involving adult participants, which found that the generation This paradox can be resolved, fortunately, by distinguishing
of debate was a requirement for group activities to lead to between what normally happens and what could, or should,
improved performance on reasoning tasks: see, for example, happen. The ability to think collectively may be an important
 THE SOCIAL BRAIN 159

and defining characteristic of our species, but that does not (2008) concluded that “classroom work is made more enrich-
mean that children are born knowing how to do it well. To ing when discussion involves the co-construction of mathe-
make the most of collaborative learning activities, it is nec- matical knowledge through the respectful exchange of ideas”
essary for partners to use their social brains and the cultural (p. 543). Webb, Nemer, and Ing (2006) reported that differ-
and psychological tool of spoken language to best effect. ences between teachers in the extent to which they asked
Some educational research has studied the effects of training students to elaborate their problem-solving strategies cor-
children in the use of language as a tool for collective reason- responded strongly to the extent to which students did so
ing, as I go on to describe in the next section. That research during group discussions. Their general conclusion is that
will also provide a useful basis for further consideration of one of the main influences on children’s talk in groups is the
the three explanations (appropriation, co-construction, and kind of talk that their teacher uses in interactions with them;
transformation) of how interamental activity can assist in- but that teachers may not often model effective discussion
tramental development. in whole-class sessions. Yet children may not have many op-
portunities to learn how to conduct reasoned discussions in
their out-of-school lives, or may not realize that they should
INTEGRATING TEACHER-STUDENT TALK engage in them when given collaborative tasks in the class-
WITH COLLABORATIVE LEARNING room. If their teachers do not raise their awareness of how
they might talk and work together, or provide them with mod-
A study of teachers in Mexican classrooms (Rojas- els and guidance, they are unlikely to develop the relevant
Drummond, Mercer, & Dabrowski, 2001) found that those skills for collective thinking and apply them appropriately.
whose students achieved the best learning outcomes (as It is not so surprising, then, that in peer group discussions
represented by assessments of their progress in mathematics the talk is often off-topic, unproductively disputational and
and literacy) not only avoided the dominance of closed inequitable. Children may all have inherited the capabilities
questions but also organized more interchanges of ideas and of the social brain, but like most human capabilities they
mutual support amongst pupils and generally encouraged require exercise and training.
pupils to take a more active, vocal role in classroom events Some interventional research has studied the effects of
than the less effective teachers. That is, they enacted a teachers guiding students in “collaborative reasoning.” For
sociocultural, “dialogic” model of education, even though example, Retznitkaya et al.’s (2001) study involved students
they did not necessarily describe it as such. More large-scale aged 10 to 11 years, constituting three classes that partici-
studies are needed to test the value of a dialogic pedagogy pated in teacher-led collaborative reasoning discussions of
(Rezniyskaya & Gregory, 2013). But overall, the available literary texts for a period of 5 weeks. These students and
evidence supports those researchers who have argued for students from three comparable classrooms, who had not
a more dialogic classroom pedagogy in which students engaged in collaborative reasoning, were asked to write per-
have opportunities to express their understandings and suasive essays. The essays of the intervention class students
misunderstanding, think aloud, ask questions, and explore were found to contain a significantly greater number of rel-
ideas without being immediately evaluated as “wrong” or evant arguments, counterarguments, rebuttals, formal argu-
“right” by the teacher (Alexander, 2001, 2008; Dawes, 2008; ment devices, and uses of text information. Through ana-
Mercer, 1995; Nystrand, 1997; Skidmore, 2006; Scott, 2008; lyzing the talk during such interventions, in which a teacher
Wells, 1999, 2009). It seems, therefore, that classroom modeled and guided discussions of that kind about literary
education should provide opportunities for students to think texts, Chinn, Anderson, and Waggoner (2001) concluded,
collectively, co-constructing knowledge and understanding
and solving problems collectively. But if group work is to be Four cognitive processes integral to good thinking and greater
productive, teachers also need to scaffold the development of learning were found more frequently in Collaborative Rea-
students’ intramental capabilities. As the science education soning discussions than in Recitations [i.e. those based on the
researcher Black (2009) commented, traditional ‘closed question’ type of interaction]. In compar-
ison with students in Recitations, students in Collaborative
Reasoning discussions (a) made many more elaborations, (b)
By listening carefully to what others say, by giving emphasis
made many more predictions, (c) provided evidence at a rate
to reasoned understanding rather than to formulaic answers,
nearly 10 times higher than in Recitations, and (d) were much
and by trying to help the class to arrive at consensus in a
more likely to articulate alternative perspectives. (p. 398)
shared understanding rather than by imposing a conclusion
arbitrarily, a teacher can make whole-class dialogue a model
for pupils’ group discussions. In both contexts, pupils are A rather different illustration of how whole-class and group
experiencing engagement in reasoned discourse. (p. 4) work activities can be integrated is provided by research in-
volving a revision method called prescriptive tutoring (Soong
Black’s comments are supported by empirical research. In & Mercer, 2011). Its aim was to reveal secondary school stu-
their review of mathematics teaching, Walshaw and Anthony dents’ misconceptions and misunderstandings in physics by
160 MERCER

getting them to solve physics problems with a partner in the interventional classes became significantly better at rea-
their class. However, the partners collaborated only through soning, as assessed by the Raven’s Progressive Matrices test
computer-mediated communication, and their dialogue was (Raven, Court, & Raven, 1995) both collectively and alone.
saved as text. Before they began to work together online, the Before the intervention, children in both control and ex-
students were encouraged by their teacher to agree on a set of perimental classes were given the Raven’s test twice: once
“ground rules” for making their discussions suitably rational, in groups and once (using a different version of the test) on
explicit, and equitable. (The rationale for this is explained in their own, in that order. After the intervention, they all again
more detail next, in relation to the Thinking Together re- did the test in groups and as individuals. The intervention
search.) These ground rules stated that the students would children obtained better group and individual scores on the
agree to postintervention collaborative application of the test than the
control children (although the intervention children had no
• share their ideas and listen to each other; more opportunities to practice the Raven’s problems than the
• consider what their partner(s) has written or drawn; control children). It seems that they had learned not only how
• respect each other’s opinions; to “interthink” more effectively but also how to reason better
• give reasons for their ideas; on their own. These results have been replicated in Mexican
• express their ideas and workings neatly and clearly; schools (Rojas-Drummond & Mercer, 2004).
• in the case of disagreement, ask “why?” or provide reasons These results can be related to the three explanations of
for their disagreement; how intermental activity influences intramental learning and
• only work on solving the problems (e.g., no web-surfing); development I offered earlier. I consider each in turn, re-
• try to concur on a solution, prior to asking the teacher to garding how they might explain the greater success of the
check their answer. children in the experimental groups in the postintervention,
group-based, and individual Raven’s tests.
By agreement with the students, their teacher was allowed
to read their online dialogues and use them as a basis for
Appropriation. The relative success of the experimen-
subsequent whole-class discussions. Implemented and eval-
tal children in the postintervention group-based Raven’s test
uated in a public secondary school in Singapore, the results of
must depend on those children applying better problem-
prescriptive tutoring showed that students in the experimen-
solving strategies than those in control classes. It is unlikely
tal group significantly outperformed students in a matched
that more individual children in the experimental classes
control group on postintervention tests of understanding of
came along with effective strategies already in place than
physics concepts.
those in the control classes, so greater success in the group
Some researchers have tried to ensure that collaborative
test would depend on the experimental children being better
learning is more effective by encouraging teachers to use
than control children at sharing and making use of any ef-
whole-class sessions to guide and model children’s use of
fective strategies that individual members already knew, or
language for reasoning. The Thinking Together intervention
devised on the spot, in this one session. An appropriation
studies carried out by myself and colleagues have so far
explanation of the relatively greater success of the experi-
involved more than 700 children, aged 6 to 14. These stud-
mental children in the individual tests would thus depend on
ies have been described in detail elsewhere (e.g., Mercer &
a significant number of them having learned better strategies
Littleton, 2007), so I summarize here only those aspects rele-
from each other in that one group session.
vant to my argument. Essentially, each class agrees to follow
a set of “ground rules” for talking together in groups, which
early studies showed help to generate more Exploratory Talk. Co-construction. The greater success of the experi-
Students are then expected to apply these rules during all mental children in the postintervention group test could result
curriculum-related group work, and the teacher uses whole- from individual children in the experimental groups being
class sessions to model Exploratory Talk and encourage chil- more able than control class children to combine and ap-
dren’s meta-awareness of the ways they use language for ply their relevant knowledge and insights to the task, and
reasoning (Dawes, 2012). so construct effective new strategies for solving the Raven’s
Compared with control classes following their normal problems than they would have done alone. On the basis of
course of study, results have shown that children who fol- their training and practice in Exploratory Talk, they would do
low the Thinking Together intervention program begin to this through a process of reasoned argumentation, whereby
use much more Exploratory Talk and pursue group activi- any suggestions made by group members would be critically
ties more cooperatively. One study with children aged 9 and evaluated and any resultant successful strategies constructed
10 found that they also gained significantly higher scores in would be learned and subsequently applied by individual
national, curriculum-based tests of science and mathematics members when working alone. Their relative success as
(Mercer, Dawes, Wegerif, & Sams, 2004). Moreover—and individuals thus would depend crucially on them being more
crucially for the case I am making here—the children in able than children in the control classes to construct new
 THE SOCIAL BRAIN 161

strategies in the one postintervention group session and then activities. This would make them more effective at sharing
use them in the individual session. knowledge and co-constructing an explanation, as previously
We can see a group of children engaging in the kind of suggested. But that external regulation, and their practice in
co-constructive approach to solving a Raven’s problem that using Exploratory Talk as a group, would also act as a tem-
would accord with this explanation in Sequence 1, which plate for self-regulation and reasoning when they tackled
comes from an experimental class in a study carried out in problems alone in the individual tests. That is, Exploratory
Mexican schools. As the sequence begins, the children (aged Talk would act as a template for carrying out the “inner
12) are trying to decide which one of a possible set of patterns dialogue” of individual reasoning. In the postintervention in-
would logically complete an unfinished sequence of patterns dividual sessions, the relative success of the experimental
made up of features that the children refer to as dots, crosses, children would thus not necessarily depend on their reten-
and stars. tion and employment of useful strategies co-constructed in
the group session. It could also be assisted by their enhanced
Sequence 1: Georgina, Luis and Mauro doing Raven’s test ability to reason “dialogically” about any problems they were
item E5 given. They would engage in any new problem-solving task in
a more metacognitive, self-regulatory way. This would make
Georgina: Here they remove the dots and this, this
it more likely that they achieved correct solutions, as they
cross (points at a drawing)
Mauro: No but wait, it does not fit would be more able to generate and apply good strategies for
Luis: No, wait themselves and not only recall them from the previous group
Mauro: No session.
Georgina: Let’s look at the sequence. Here it is like According to the “transformation” explanation, then, the
this, they remove the cross and the dots. success of the Thinking Together intervention would depend
Here they are not there any more, here on it having enabled children not only become more able to
(points) have reasoned discussions with peers but also become more
Luis: And here they remove only the dots able to have a “reasoned discussion” on their own. This rep-
Georgina: Yes, the dots. And this part, only the star resents a very strong claim for the effects of social activity
Mauro: It would be this one, look at it (points)
on individual cognition, because it explains the development
Luis: Which they have removed
of a sophisticated, educated manner of individual thinking
Mauro: It would be this one, because, look, it goes
like this (points) through the internalization of collective reasoning. The cog-
Georgina: But how, if it doesn’t have dots? nition of a normally developing child would not only be
Luis: It doesn’t have dots. Just the cross would “social” because the human brain has inherent sensitivities
remain to the signaled intentions and emotions of others (as argued
Georgina: Yeah, because they have been removed! by evolutionary psychologists), but because it embodies a
(Georgina writes down option number 1 on the answer way of thinking that takes account of the possibility of mul-
sheet, which is correct) tiple, varied perspectives and explanations which need to be
(Adapted from Rojas-Drummond & Mercer, 2004, p. 109) compared and evaluated. We might note here that, in research
discussed earlier, Howe (2009) found that the incidence of
unresolved contradictions or disagreements during children’s
Some aspects of their reasoning may not be clear to a reader, collaborative problem-solving science tasks had a positive
because they sometimes (and appropriately enough) rely on correlation with scores on delayed posttests of children’s sci-
pointing to a pattern to support their arguments; but the dis- entific understanding but not on immediate posttests. Those
cussion nevertheless has many features of Exploratory Talk findings suggest that even if the coconstruction of a sat-
with its challenges, reasoned justifications, and equitable dis- isfactory explanation is not achieved during collaboration,
tribution of conversational turns. Together, they construct and the kind of intermental activity which takes place during
employ a mental scenario in which the originators of the test Exploratory Talk may stimulate subsequent intramental “di-
have differentially “removed” certain features from specific alogue,” which enables the individual to achieve a new level
patterns in the series, leading to the logical inevitability of of understanding.
one of the possible solutions being correct. From reading the Are there reasons for believing that the “transformation”
sequence, it would be difficult to attribute the group’s success explanation applies to the results of the Thinking Together
to the contributions of any one member. interventions: that the intervention children’s “monological”
reasoning became more “dialogic” through some kind of in-
Transformation. Their teachers’ modeling of Ex- ternalization of their Exploratory Talk? There are, though
ploratory Talk and establishment of “ground rules” for dis- only to a limited extent. We have evidence that interven-
cussion would enable the experimental children to solve prob- tion class children went on to deal more successfully as
lems more effectively in the group tests, because they would individuals than control class children with new types of
use that experience of external regulation to coregulate their reasoning problems and provided better written arguments
162 MERCER

to support their arguments in English assessments (Mercer dren in intervention classes becoming better at discussing
& Littleton, 2007, Chapter 6). “Transformation” would also problems together and coregulating their activity. They im-
account for intervention children’s subsequently improved proved significantly more than control classes on measures of
performances when dealing with new mathematics, science, conceptual understanding, quality of reasoning and metacog-
and English assignments involving reasoning which they had nition on a music task. They were also able to explain their
not practiced with their peers—they could apply to them reasoning to a researcher more explicitly when doing both
an improved general reasoning skill that involved thinking a music-related and a science-related task (concerned with
dialogically. There is also some anecdotal evidence from floating and sinking). We have also tested for any improve-
teachers participating in the Thinking Together research that ments in children’s performance on a problem-solving task
they observed that children were able to reason more ef- where they had to assemble a model railway track to a given
fectively (in speech and writing) about new topics after the template (adapted from Karmiloff-Smith, 1979), but results
intervention. However, the Thinking Together studies pre- of that are not yet available. A larger study, involving a mod-
ceded the formulation of the three explanations of effect, ified intervention program, is being planned.
and so were not designed to discriminate between them. Per-
haps, as sociocultural researcher Futoshi Hiruma (personal
communication, May 12, 2008) suggested, in any new stud- DOES COLLABORATION HELP
ies children from intervention and control classes should be TO SOLVE PROBLEMS?
asked to “think aloud” as they deal individually with new
types of problem-solving tasks, to test the hypothesis that I have offered evidence in support of the value of collective
the intervention children would offer more explicitly rea- intellectual activity for the development of children’s reason-
soned protocols that were related to the structure and content ing. But my argument early in this article for the evolution-
of the classroom dialogues in which they had been involved. ary importance of the emergence of a capacity for thinking
This might be combined with Howe’s (2009) methodology of collectively also requires evidence to show that, at least in
immediate and delayed posttests of children’s understanding some situations, collective thinking is more productive and
after collaborative work—which could include their develop- effective than solitary thinking. One source is from studies of
ing understanding of how they use talk for thinking together pairs and teams of adults working together. These range from
and/or how they reason alone. The use of sociocultural dis- quasi-experimental studies (as reviewed in Nemeth, 1995;
course analysis (Mercer, 2008, 2009), with its methods for Paulus, Dzindolet, & Kohn, 2012) to more ethnographic ac-
tracking the temporal trajectory of talk through time, and counts (Miell & Littleton, 2008) and biographical case stud-
hence the appropriation by children of linguistic items and ies (John-Steiner, 2000). I mentioned earlier the assembly
structures from their earlier experiences, would be appropri- bonus effect, discovered by some researchers, whereby the
ate here. This would tell us more about if, and how, dialogue performance of a group is better than that of its best mem-
can transform the quality of individual reasoning. ber (Laughlin et al., 2006); and that although the success of
A related possible line of inquiry would build stronger group endeavors does not correlate strongly with the aver-
links between research on dialogue and that on the develop- age intelligence of group members, it is correlated with the
ment of metacognition and self-regulation. My colleagues1 average social sensitivity of members and the equality in dis-
and I have already designed and tested an intervention pro- tribution of their conversational turn-taking (Woolley et al.,
gram on that basis, for use by Year 1 teachers with their 2010). Biographical case studies strongly encourage the view
classes. Essentially, it requires teachers to use the Thinking that some of the most significant and creative human achieve-
Together approach to raise children’s awareness and skills in ments can be explained only through the joint activity of more
working and talking together, and then engage them in a series than one talented individual, with conversational interaction
of collaborative problem-solving tasks that require collective between collaborators being seen as important. Psychologi-
reasoning. This program has so far been tested in a pilot cal research on creativity has correspondingly begun to widen
study (the Children Articulating Thinking [ChAT] Project) its focus from understanding the talents of gifted individuals
involving six primary school teachers and their classes. The to include studies of how people are collectively creative.
main hypothesis has been that developing the children’s use Several studies take a sociocultural theoretical perspective,
of language as a tool for social, joint regulation in problem in which culturally framed ways of communicating and the
solving will stimulate the development of their individual use of cultural tools are considered important (Miell & Lit-
self-regulation and metacognitive abilities—and that this will tleton, 2004; Paulus & Nijstad, 2003; Sawyer, 2012; Sawyer
be demonstrated through comparisons with children in con- & DeZutter, 2009). In a meta-review of research on “or-
trol classes when they subsequently attempt individual tasks. chestrating creativity,” Hämäläinen and Vähäsantane (2011)
The results (unpublished as yet) are encouraging, with chil- noted that within recent studies of collaborative creativity,
“the different roles of group members, including mutual
1David Whitebread, Deborah Pino-Pasternak, Christine Howe, Jane War- explaining and shared knowledge construction, have been
wick, Penny Coltman, and Usha Goswami. seen to enable new creative processes and outputs” (p. 172).
 THE SOCIAL BRAIN 163

Some researchers have begun to offer “co-construction” ex- instead that our evolution has equipped us with the distinctive
planations of creativity, even sometimes using that term human capability for engaging in goal-orientated collective
(Rojas-Drummond, Mazon, Fernandez, & Wegerif, 2006). thinking, so that we are able to achieve more together than
However, reviewing experimental studies into group- we each could do alone. I have therefore proposed that the
based work activity, one team of experienced researchers concept of the social brain should be developed beyond its
has commented, “We do not know of a study that has clearly initial definition, which is concerned with how humans in-
demonstrated . . . enhanced performance of teamwork rela- terpret and negotiate complex social relations to pursue their
tive to working as individuals” (Paulus et al., 2012, p. 348). individualistic needs. The sociality of the brain does not just
Some research has highlighted the flaws in collective thinking enable each of us to cope with the complexity of society and
processes, as when “groupthink” leads to unjustifiable con- pursue our own agendas, it also enables us to solve problems
sensus (Baron, 2005; Janis, 1982). Surprisingly, too, there together and to create and develop knowledge at the cultural
seem to have been few detailed studies of how the use of talk level. Although collective intellectual activity is not always
in noneducational working groups leads to effective creative more creative and productive than individual efforts, research
and outcomes (though see Edwards & Middleton, 1986; Mid- has shown that it has the kinds of distinctive functions and
dup, Coughlan, & Johnson, 2010; Miell & Littleton, 2008). benefits which support my argument.
Nevertheless, extrapolating from the available evidence to Research in several fields provides additional support for
connect with my theme here, it is possible to draw some an expanded conception of the social brain, as I have shown.
conclusions. The most reasonable answer to the question, The special human capacity for “theory of mind” allows us
“Do people solve problems more effectively when working to appreciate that we each may have different perspectives
together than alone?” is “In the right circumstances, yes.” It and concerns, and motivates us to assess and monitor one
depends to some extent on the nature of the task, and on the another’s states of understanding and common knowledge.
quality of the interactions within a group. As in studies of This provides a basis for educating each new generation, as
collaborative learning in educational settings, working teams intermental activity allows more experienced members of a
have been found to achieve better solutions than individuals, community to have a formative influence on the intramental
and sometimes very significantly so, but only if they commu- development of less experienced members. I have used find-
nicate effectively. For example, “groupthink” can be avoided ings from several lines of inquiry to argue that the process
by following ground rules that encourage reasonable dissent of reasoning collectively, which has been crucial for the suc-
(Nemeth, 1995). Relevant research supports the view that cess of our species, provides a template for the development
the success of working teams depends on members using, of individual reasoning. The distinctive nature of the social
when appropriate, a mode of interaction which resembles brain, then, is encapsulated in the relationship between the
Exploratory Talk (Littleton & Mercer, 2013). Yet, as in the intermental and the intramental, the social and the psycho-
classroom, problem-solving discussions in the workplace are logical. Language, which has evolved in conjunction with the
often not of this kind. As mentioned earlier, in evolution- social brain, has a special role in mediating that relationship,
ary anthropology and psychology it tends to be assumed that through its integration with cognition generally and its use
the skills of social cognition are “hard-wired” and ready to as both a cultural/social and psychological tool (or, rather,
use. But, like educational research, research in this field en- toolkit). I have offered three explanations of the effects of
courages the view that although people inherit a distinctively collaborative learning and dialogue on the development of
human capacity for thinking socially and collectively, these children’s reasoning—appropriation, co-construction, and
skills need to be developed and practiced. transformation. They stand as potentially complementary,
rather than as alternative, explanations. Individualistic ac-
counts of reasoning and its development are inadequate in
CONCLUSIONS comparison because they struggle to explain both the “as-
sembly bonus effects” of collective intellectual activity and
In this article, I have suggested that the concept of the so- the ways that educated modes of individual thinking embody
cial brain could help to integrate research in evolutionary essential features of reasoned dialogue.
psychology, developmental psychology, social psychology, Through creating links between evolutionary, social, de-
educational psychology, and other fields such as linguistics velopmental, and educational psychology, the expanded con-
and philosophy, generating a better account of the distinctive cept of the social brain could strengthen a sociocultural ac-
nature of human cognition. The origin of the concept lies count of human learning and cognitive development and
in evolutionary studies of human behavior, and that lively might also help neuroscience research and educational re-
field of inquiry offers a fresh perspective that should not be search become better integrated. The direct relevance and
ignored in more established fields of psychological and ed- value of the findings of neuroscience for improving educa-
ucational research. But I suggest that we should question its tional practice and theory have been recognized to be limited
evolutionary account of our origins, which is predicated prin- so far, even by those involved (Goswami, 2007). Most neuro-
cipally upon competition between individuals. I have argued science research has been aimed at explaining brain function,
164 MERCER

not addressing educational concerns (Willingham, 2009). of society that children are taught how to become effective
Moreover, when educational issues have been addressed by interthinkers.
neuroscientists there has been a narrow focus on how research The explanatory framework I have used, sociocultural
might serve the perceived needs of practitioners, and how to theory, does not avoid or deny recognition of the role of
make the findings of neuroscience intelligible to teachers, the individual in “making sense” but examines that role in the
rather than how neuroscientists could collaborate with psy- context of processes of collective thinking activity and the
chologists of education or other researchers in investigating creation of socially shared knowledge. In an earlier issue of
educational research questions (Mason, 2009). For example, this journal, Säljö (2009) suggested that “the point of a the-
if neuroscience could begin to identify the ways in which ory does not lie in its correspondence with the world (which
language functions are related to, or integrated with, more would be the realist perspective) but rather in its explanatory
general processes of learning or problem solving, it might power in relation to a set of issues” (p. 204). My interest in
help explain why children’s early language experience can the social brain and collective thinking has a strong practi-
so profoundly affect the course of their intellectual and aca- cal dimension. I want to know why some teachers are better
demic development. Could neuroscientists also contribute to than others at helping their students achieve higher levels
our understanding of such processes by studying how brains of attainment, how levels of attainment are related to fea-
operate in conjunction: for example, how spoken dialogue tures of children’s social and communicative experience, and
is reflected in the brain functioning of partners? Or could why and how group-work promotes learning. I also want
they investigate whether the exercise of “theory of mind” to know how best to educate children as “interthinkers.” Re-
capacities, the establishment of common knowledge, and the search has shown that the potential of the social brain is often
formulation of joint plans of action, involve “mirrored” neu- being squandered in classroom education, because it is not
ral activity in the brains of partners engaged in joint problem being used, or educated, effectively. We know that collabo-
solving? More adventurously, we might hope for studies that rative learning benefits individual learning, but only under
investigate ways that the brain activity of teachers and stu- certain conditions. We are now able to identify the peda-
dents, or collaborating learners, are correlated during joint gogic strategies that can maximize the educational impact
activity. By combining the methods of neuroscience with of classroom interactions and to develop children’s skills in
those of discourse analysis, we might even seek the correlates reasoning collectively. We need to ensure that children gain
of Exploratory Talk. And if neuroscience research provided most benefit from collaborative learning, and enable them to
more evidence that language use is indeed a wholly inte- take part productively in the collective thinking activities of
grated aspect of brain function, this might encourage more the wider world. More educational benefit would be likely
educational and psychological researchers to investigate the to come from this than from the pursuit of such popular
role of language in learning and conceptual development, “neuromyths” as “learning styles,” “right/left laterality,” and
and specifically in the development of metacognition and “whole-brain learning” (as discussed by Goswami, 2007). In
self-regulation. summary, then, I hope that I have provided good reasons for
A reviewer of this article in draft suggested that it would psychologists of education to adopt, adapt, and develop the
be better directed at an audience of evolutionary psycholo- concept of the social brain. In doing so, and making links
gists and neuroscientists, rather than psychologists of educa- with colleagues in other branches of psychology, the out-
tion who commonly take a sociocultural perspective and so come might be a better and more useful understanding of the
do not need to be persuaded of its value or of the need to distinctive nature and origins of human cognition.
study processes of collaborative learning. But the focus of
most research on collaborative learning has been on the ef-
fects of joint intellectual activity on individual learning and ACKNOWLEDGMENTS
development, rather than on understanding and improving
collaborative problem solving in its own right. This suggests Thanks to Clark Chinn, Lyn Dawes, Usha Goswami, Chris-
that researchers in that field may be, ironically, trapped in a tine Howe, Karen Littleton, Bert van Oers, Deborah Pino-
similar kind of individualistic perspective to those in evolu- Pasternak, David Whitebread, and three anonymous review-
tionary and cognitive psychology. The ability to think col- ers for very constructive comments on drafts of this article.
lectively in the pursuit of common goals is just as important
today as it was in our evolutionary past. If that ability can
be developed and enhanced, as research strongly suggests,
educational research and practice should be concerned with REFERENCES
its development per se, and not only with collaboration as a
Abrams, D. A., Nicol, T., Zecker, S., & Kraus, N. (2008). Right hemi-
pedagogic device for individual tuition. We commonly hear sphere auditory cortex is dominant for coding syllable patterns in speech.
today of the need for “good team players”—and that does Journal of Neuroscience, 28, 3958–3965. doi:10.1523/JNEUROSCI.
not just mean being amiable colleagues. It is in the interests 0187-08.2008
 THE SOCIAL BRAIN 165

Alexander, R. J. (2001). Culture and pedagogy: International comparisons Dawes, L. (2008). Encouraging students’ contribution to dialogue during
in primary education (pp. 391–528). Oxford, UK: Blackwell. science. School Science Review, 90, 101–107.
Alexander, R. (2008). Culture, dialogue and learning: Notes on an emerging Dawes, L. (2012). Talking points: Discussion activities in the primary class-
pedagogy. In N. Mercer & S. Hodgkinson (Eds.), Exploring talk in schools room. London, UK: David Fulton.
(pp. 91–114). London, UK: Sage. Dawkins, R. (1976). The selfish gene. Oxford, UK: Oxford University Press.
Augustinova, M. (2008) Falsification cueing in collective reasoning: Exam- Diaz, R. (1983). Thought and two languages: The impact of bilingualism
ple of the Wason selection task. European Journal of Social Psychology, on cognitive development. Review of Research in Education, 10, 23–54.
38, 770–785. doi:10.1002/ejsp.532 doi:10.2307/1167134
Baker, C. (2006) Foundations of bilingual education and bilingualism. Dignath, G., Buettner, G., & Langfeldt, H-P. (2008). How can primary school
Clevedon, UK: Multilingual Matters. students learn self-regulated learning strategies most effectively? A meta-
Bakhtin, M. (1981). The dialogic imagination. Austin: University of Texas analysis on self-regulation training programmes. Educational Research
Press. Review, 3, 101–129.
Barnes, D. (1976). From communication to curriculum. Harmondsworth, Dunbar, R. (1998). The social brain hypothesis. Evolutionary Anthropology,
UK: Penguin. 6, 178–189.
Barnes, D. (2008). Exploratory talk for learning. In N. Mercer & S. Hodgkin- Edwards, D., & Mercer, N. (2012). Common knowledge: The development
son (Eds.), Exploring talk in school (pp. 1–16). London, UK: Sage. of understanding in the classroom. London, UK: Methuen/Routledge.
Barnes, D. (2010). Why talk is important. English Teaching: Practice and (Original work published 1987)
Critique, 9, 7–10. Edwards, D., & Middleton, D. (1986). Joint remembering: Constructing an
Barton, D. (2009). Understanding textual practices in a changing world. In account of shared experience through conversational discourse, Discourse
M. Baynham & M. Prinsloo (Eds.), The future of literacy studies (pp. Processes, 9, 423–459. doi:10.1080/01638538609544651
38–53). Basingstoke, UK: Palgrave Macmillan. Elbers, E. (1994). Sociogenesis and children’s pretend play: A variation
Bennett, N., & Cass, A. (1989). The effects of group composition on group on Vygotskian themes. In W. de Graaf & R. Maier (Eds.), Sociogen-
interactive processes and pupil understanding. British Educational Re- esis re-examined (pp. 58–72). New York, NY: Springer. doi:10.1007/
search Journal, 15, 119–132. doi:10.1080/0141192890150102 978-1-4612-2654-3 13
Berkowitz, M., Gibbs, J., & Broughton, J. (1980). The relation of moral Fidalgo, Z., & Pereira, F. (2005). Sociocultural differences and the ad-
judgement stage disparity to developmental effects of peer dialogues. justment of mothers’ speech to their children’ cognitive and language
Merrill-Palmer Quarterly, 26, 341–357. comprehension skills. Learning and Instruction, 15, 1–21.
Bialystok, E., Craik, F., Grady, C., Chau, W., Ishii, R., Gunji, A., & Pan- Fiedler, K., & Bless, H. (2001). Social cognition. In M. Hewstone & W.
tev, C. (2005). Effect of bilingualism on cognitive control in the Si- Stroebe (Eds.), Introduction to social psychology (pp. 123–146). London,
mon task: Evidence from MEG. NeuroImage, 24, 40–49. doi:10.1016/ UK: Sage.
j.neuroimage.2004.09.044 Forman, E. A., & Cazden, C. B. (1985). Exploring Vygotskian per-
Bialystok, E., & Feng, X. (2010). Language proficiency and its implications spectives in education: the cognitive value of peer interaction. In J.
for monolingual and bilingual children. In A. Durgunoglu & C. Golden- V. Wertsch (Ed.), Culture, communication and cognition: Vygotskian
berg (Eds.). Dual language learners: The development and assessment of perspectives (pp. 323–347). Cambridge, UK: Cambridge University
oral and written language (pp. 121–138). New York, NY: Guilford. Press.
Black, P. (2009) Looking again at formative assessment. Learning and Frith, C., & Singer, T. (2008). The role of social cognition in decision mak-
Teaching Update, 30, 3–5. ing. Philosophical Transactions of the Royal Society, 363, 3875–3886.
Black, P., & Wiliam, D. (1998). Inside the black box: Raising standards doi:10.1098/rstb.2008.0156
through classroom assessment. London, UK: Kings College School of Goodman, S. (1996). Market forces speak English. In S. Goodman &
Education. D. Graddol (Eds.), Redesigning English: New texts, new identities (pp.
Bruner, J. S. (1983). Child’s talk: Learning to use language. Oxford, UK: 141–163). London, UK: Routledge.
Oxford University Press. Goswami, U. (2007). Neuroscience and education: From research to practice.
Buss, D. (2001). Human nature and culture: an evolutionary psychol- Nature Review Neuroscience, 7, 406–413. doi:10.1038/nrn1907
ogy perspective. Journal of Personality, 69, 955–978. doi:10.1177/ Goswami, U. (2009). Mind, brain, and literacy-biomarkers as usable
026565908500100113 knowledge for education. Mind, Brain, and Education, 3, 176–184.
Cazden, C. (2001). Classroom discourse: The language of teaching and doi:10.1111/j.1751-228X.2009.01068.x
learning (2nd ed.). New York, NY: Heinemann. Grist, M. (2009). Changing the subject: How new ways of thinking about
Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: The human behavior might change politics, policy and practice. London,
perception–behavior link and social interaction. Journal of Personal- UK: Royal Society of Arts. Retrieved from http://www.thersa.org/
ity and Social Psychology, 76, 893–910. doi:10.1037//0022-3514.76. data/assets/pdf file/0020/250625/Nov28th2009ChangingThe-Subject
6.893 Pamphlet.pdf
Chinn, C., Anderson, R., & Waggoner, M. (2001). Patterns of discourse Grosjean, F. (2010). Bilingual: Life and reality. Cambridge, MA: Harvard
in two kinds of literature discussion. Reading Research Quarterly, 36, University Press.
378–411. doi:10.1598/RRQ.36.4.3 Hämäläinen, R., & Vähäsantane, K. (2011). Theoretical and pedagogi-
Christie, F., & Martin, J. (1997). Genre and institutions: Social processes in cal perspectives on orchestrating creativity and collaborative learning.
the workplace and school. London, UK: Cassell. Educational Research Review, 6, 169–184. doi:10.1016/j.edurev.2011.
Clark, A. (1998). Magic words: How language augments human computa- 08.001
tion. In P. Carruthers & J. Boucher (Eds.), Language and thought: Interdis- Hart, B., & Risley, T. R. (1995). Meaningful differences in the everyday
ciplinary themes (pp. 162–183). Cambridge, UK: Cambridge University experience of young American children. New York, NY: Brookes.
Press. Hattie, J. A., Biggs, J., & Purdie, N. (1996). Effects of learning skills in-
Daniels, H. (2001). Vygotsky and pedagogy. London, UK: Routledge/ terventions on student learning: a meta-analysis. Review of Educational
Falmer. Research, 66, 99–136. doi:10.2307/1170605
Daniels, H. (2008). Vygotsky and research. Abingdon, UK: Routledge. Hermann, E., Call, J., Hernandez-Lloreda, M. V., Hare, B., & Tomasello,
Darcy, N. (1963). Bilingualism and the measurement of intelligence: Review M. (2007). Humans have evolved specialized skills of social cogni-
of a decade of research. Journal of Genetic Psychology, 103, 259–282. tion: The cultural intelligence hypothesis. Science, 317, 1360. doi:10.
doi:10.1080/00221325.1963.10532521 1126/science.1146282
166 MERCER

Heyes, C. M. (2010). Where do mirror neurons come from? Neuroscience Mercer, N. (2008). The seeds of time: Why classroom dialogue needs
and Biobehavioural Reviews, 34, 575–583. doi:10.1016/j.neubiorev. a temporal analysis. Journal of the Learning Sciences, 17, 33–59.
2009.11.007 doi:10.1080/10508400701793182
Howe, C. (2009). Collaborative group work in middle childhood: Joint con- Mercer, N. (2009). The analysis of classroom talk: Methods and
struction, unresolved contradiction and the growth of knowledge. Human methodologies. British Journal of Educational Psychology, 80, 1–14.
Development, 52, 215–219. doi:10.1348/000709909X479853
Howe, C. (2010). Peer groups and children’s development. Oxford, UK: Mercer, N., Dawes, R., Wegerif, R., & Sams, C. (2004). Reasoning
Wiley-Blackwell. doi:10.1002/9781444318098 as a scientist: Ways of helping children to use language to learn
Iiskala, T., Vauras, M., Lehtinen, E., & Salonen, P. (2011). Socially science. British Educational Research Journal, 30, 367–385. doi:10.
shared metacognition of dyads of pupils in collaborative mathemati- 1080/01411920410001689689
cal problem-solving processes. Learning and Instruction, 21, 379–393. Mercer, N., & Littleton, K. (2007). Dialogue and the development of chil-
doi:10.1016/j.learninstruc.2010.05.002 dren’s thinking. London, UK: Routledge
Janis, I. (1982). Groupthink: Psychological studies of policy decisions and Mercer, N., & Sams, C. (2006). Teaching children how to use lan-
fiascoes. New York, NY: Houghton Mifflin. guage to solve maths problems. Language and Education, 20, 507–527.
Jeong, H., & Chi, M. (2007). Knowledge convergence and collab- doi:10.2167/le678.0
orative learning. Instructional Science, 35, 287–315. doi:10.1007/ Mercier, H., & Sperber, D. (2011). Why do humans reason? Arguments
s11251-006-9008-z for an argumentative theory. Behavioral and Brain Sciences, 34, 57–74.
Jewitt, C. (2008). Technology, literacy, learning: A multimodality approach. doi:10.1017/S0140525X10000968
London, UK: Routledge. Michaels, S., & O’Connor, M. C. (2002). Accountable talk: Classroom
Jewitt, C., Kress, G., Ogborn, J., & Tsartsarelis, C. (2004). Exploring learn- conversation that works. Pittsburgh, PA: University of Pittsburgh.
ing through visual, actional and linguistic communication: The multi- Middleton, D., & Edwards, D. (1990). (Eds.). Collective remembering. Lon-
modal environment of science classroom. Educational Review, 53, 6–16. don, UK: Sage.
doi:10.1080/00131910120033600 Middup, C., Coughlan, T., & Johnson, P. (2010). How creative groups struc-
John-Steiner, V. (2000). Creative collaboration. New York, NY: Oxford ture tasks through negotiating resources. In M. Lewkowicz, P. Hassanaly,
University Press. M. Rohde, & V. Wulf (Eds.), Proceedings of COOP 2010: Computer Sup-
Karmiloff-Smith, A. (1979). Problem solving construction and representa- ported Cooperative Work (pp. 203–221). London, UK: Springer-Verlag.
tion of closed railway circuits. Archives of Psychology, 47, 37–59. doi:10.1007/978-1-84996-211-7 12
Keefer, M., Zeitz, C., & Resnick, L. (2000). Judging the quality of Midgely, M. (2010). The solitary self: Darwin and the selfish gene. London,
peer-led student dialogues. Cognition and Instruction, 18, 53–81. UK: Acumen. doi:10.1017/UPO9781844654833
doi:10.1207/S1532690XCI1801 03 Miell, D., & Littleton, K. (2004). Collaborative creativity: Contemporary
Kress, G. (1987). Genre in a social theory of language. In I. Reid (Ed.), perspectives. London, UK: Free Association Books.
The place of genre in learning (pp. 33–43). Geelong, Australia: Deakin Miell, D., & Littleton, K. (2008). Musical collaboration outside school:
University Press. doi:10.1111/j.1754-8845.1992.tb01061.x Processes of negotiation in band rehearsals. International Journal of Ed-
Larkin, S. (2009). Socially mediated metacognition and learning to write. ucational Research, 47, 41–49. doi:10.1016/j.ijer.2007.11.006
Thinking Skills and Creativity, 4, 149–159. doi:10.1016/j.tsc.2009.09.003 Miller, G. F. (1999). Sexual selection for cultural displays. In R. Dunbar,
Laughlin, P. R., Hatch, E. C., Silver, J. S., & Boh, L. (2006). Groups perform C. Knight, & C. Power (Eds.), The evolution of culture (pp. 71–91).
better than the best individuals on letters-to-numbers problems: Effects of Edinburgh, UK: Edinburgh University Press.
group size. Journal of Personality and Social Psychology, 90, 644–651. Mukamel, R., Ekstrom, A., Kaplan, J., Iacoboni, M., & Fried, I. (2010).
doi:10.1037/0022-3514.90.4.644 Single-neuron responses in humans during execution and observation
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate pe- of actions. Current Biology, 20, 750–756. doi:10.1016/j.cub.2010.02.
ripheral participation. Cambridge, UK: Cambridge University Press. 045
doi:10.1017/CBO9780511815355 Muller-Mirza, N., & Perret-Clermont, A-N. (2009). Argumentation and ed-
Leat, D., & Lin, M. (2003). Developing a pedagogy of metacognition and ucation. New York, NY: Springer. doi:10.1007/978-0-387-98125-3
transfer: some signposts for the generation and use of knowledge and the Murphy, P., Wilkinson, I., Soter, A., Hennessy, M., & Alexander, J. (2009).
creation of research partnerships. British Educational Research Journal, Examining the effects of classroom discussion on students’ comprehen-
29, 383–416. doi:10.1080/01411920301853 sion of text: A meta-analysis. Journal of Educational Psychology, 101,
Lemke, J. L. (1990). Talking science: Language, learning and values. Nor- 740–764.
wood, MA: Ablex. Nemeth, C. J. (1995). Dissent as driving cognition, attitudes, and
Levinson, S. C. (2006). On the human “interaction engine.” In N. J. Enfield, judgments. Social Cognition, 13, 273–291. doi:10.1521/soco.1995.13.
& S. C. Levinson (Eds.), Roots of human sociality: Culture, cognition 3.273
and interaction (pp. 39–69). Oxford, UK: Berg. Nystrand, M. (1997). Opening dialogue: Understanding the dynamics of
Littleton, K., & Mercer, N. (2013). Interthinking: Putting talk to work. language and learning in the English classroom. New York, NY: Teachers
Abingdon, UK: Routledge. College Press.
Macnamara, J. (1966). Bilingualism and primary education. Edinburgh, UK: Patel, A. D. (2003). Language, music, syntax and the brain. Nature Neuro-
Edinburgh University Press. science, 6, 674–681.
Martin, J. (1993). Genre and literacy: Modeling context in educational lin- Paulus, P., Dzindolet, M., & Kohn. N. (2012). Collaborative creativity-
guistics. Annual Review of Applied Linguistics, 13, 141–172. doi:10.1017/ group creativity and team innovation. In M. Mumford (Ed.), The hand-
S0267190500002440 book of organizational creativity (pp. 327–358). London, UK: Academic.
Mason, L. (2009). Bridging neuroscience and education: A two-way path is doi:10.1016/B978-0-12-374714-3.00014-8
possible. Cortex, 45, 548–549. doi:10.1016/j.cortex.2008.06.003 Paulus, P. B., & Nijstad, B. A. (2003). Group creativity: Innovation through
Maybin, J. (2006). Children’s voices: Talk, knowledge and identity. Bas- collaboration. New York, NY: University Press.
ingstoke, UK: Palgrave Macmillan. Pea, R. (1993). Learning scientific concepts through material and social
Mercer, N. (1995). The guided construction of knowledge: talk amongst activities: Conversational analysis meets conceptual change. Educational
teachers and learners. Clevedon, UK: Multilingual Matters. Psychologist, 28, 265–277. doi:10.1207/s15326985ep2803 6
Mercer, N. (2000). Words and minds: How we use language to think together. Piaget, J. (1926). The language and thought of the child. New York, NY:
London, UK: Routledge. doi:10.4324/9780203464984 Harcourt Brace Jovanovich.
 THE SOCIAL BRAIN 167

Pinker, S. (1994). The language instinct. London, UK: Penguin. Siegler, R. S. (2002). Microgenetic studies of self-explanation. In N. Gra-
Pinker, S. (2007). The stuff of thought: Language as a window into human nott & J. Parziole (Eds.), Microdevelopment: Transition processes in de-
nature. London, UK: Penguin. velopment and learning. Cambridge, UK: Cambridge University Press.
Premack, D., & Woodruff, G. (1978). Does the chimpanzee have doi:10.1017/CBO9780511489709.002
a theory of mind? Behavioral and Brain Sciences, 1, 515–526. Skidmore, D. (2006). Pedagogy and dialogue. Cambridge Journal of Edu-
doi:10.1017/S0140525X00076512 cation, 36, 503–514. doi:10.1080/03057640601048407
Raven, J., Court, J., & Raven, J. C. (1995). Manual for Raven’s Progres- Slavin, R. E. (2009). Cooperative learning. In G. McCulloch & D. Crook
sive Matrices and Vocabulary Scales. Oxford, UK: Oxford Psychologists (Eds.), International encyclopaedia of education (pp. 161–178). Abing-
Press. ton, UK: Routledge.
Reese, E., Haden, C. A., & Fivush, R. (1993). Mother–child conversations Slavin, R., Groff, C., & Lake, C. (2009). Effective programs in middle and
about the past: Relationships of style and memory over time. Cognitive high school mathematics: A best-evidence synthesis. Review of Educa-
Development, 8, 403–430. doi:10.1016/S0885-2014(05)80002-4 tional Research, 79, 839–911. doi:10.3102/0034654308330968
Reznitskaya, A., Anderson, R., McNurlen, B., Nguyen-Jahiel, K., Smith, M. K., Wood, W., Adams, W., Wieman, C., Knight, J., Guild, N.,
Archodidou, A., & Kim, S. (2001). Influence of oral discus- & Su, T. (2009). Why peer discussion improves student performance on
sion on written argument. Discourse Processes, 32, 155–175. in-class concept questions. Science, 323, 122–124. doi:10.1126/science.
doi:11207/S15326950DP3202&3 04 1165919
Reznitskaya, A., & Gregory, M. (2013). Student thought and classroom lan- Soong, B., & Mercer, N. (2011). Improving students’ revision of physics
guage: Examining the mechanisms of change in dialogic teaching. Educa- concepts through ICT-based co-construction and prescriptive tutor-
tional Psychologist, 48, 114–133. doi:10.1080/00461520.2013.7758958 ing. International Journal of Science Education, 33, 1055–1078.
Rojas-Drummond, S. (2009). Rethinking the role of peer collabo- doi:10.1080/09500693.2010.489586
ration in enhancing cognitive growth. Human Development, 52, Swales, A. (1990). Genre analysis: English in academic and research set-
240–245. tings. Cambridge, UK: Cambridge University Press.
Rojas-Drummond, S., Mazon, N., Fernandez, M., & Wegerif, R. (2006). Swanson, H. L., Hoskyn, M., & Lee, C. (1999). Intervention for students
Explicit reasoning, creativity and co-construction in primary school chil- with learning disabilities: A meta-analysis of treatment outcomes. New
dren’s collaborative activities. Thinking Skills and Creativity, 1, 84–94. York, NY: Guilford.
doi:10.1016/j.tsc.2006.06.001 Tomasello, M. (2008, May 25). How are humans unique? The New York
Rojas-Drummond, S., & Mercer, N. (2004). Scaffolding the development of Times Magazine. Retrieved from http://www.nytimes.com/
effective collaboration and learning, International Journal of Educational Treagust, D., & Duit, R. (2008a). Compatibility between cultural studies
Research, 39, 99–111. doi:10.1016/S0883-0355(03)00075-2 and conceptual change in science education: There is more to acknowl-
Rojas-Drummond, S., Mercer, N., & Dabrowski, E. (2001). Teaching- edge than to fight straw men! Cultural Studies of Science Education, 3,
learning strategies and the development of problem solving in Mexi- 387–395. doi:10.1007/s11422-008-9096-y
can classrooms. European Journal of Psychology and Education, 16, Treagust, D., & Duit, R. (2008b). Conceptual change: A discussion of the-
179–196. oretical, methodological and practical challenges for science education.
Roseth, C., Johnson, D., & Johnson, R. (2008). Promoting early adoles- Cultural Studies of Science Education, 3, 297–328. doi:10.1007/s11422-
cents’ achievement and peer relationships: the effects of cooperative, 008-9090-4
competitive and individualistic goal structures. Psychological Bulletin, van Oers, B., Elbers, E., van der Veer, R., & Wardekker, W. (Eds.). (2008).
134, 223–246. doi:10.1037/0033-2909.134.2.223 The transformation of learning: Advances in cultural-historical activity
Royal Society of Arts. (2010). Royal Society of Arts education seminars: theory. Cambridge, UK: Cambridge University Press.
Curriculum and the social brain. Retrieved from http://www.thersa.org/ Vass, E., & Littleton, K. (2010). Peer collaboration and learning in the
projects/education/education-seminars-2010/curriculum-and-the-social- classroom. In K. Littleton, C. Wood, & J. Kleine Staarman (Eds.), Inter-
brain national handbook of psychology in education (pp. 105–136). Leeds, UK:
Rudek, D., & Haden, C. (2005). Mothers’ and preschoolers’ mental state Emerald.
language during reminiscing over time. Merrill-Palmer Quarterly, 51, Veenman, M. V. J., & Spaans, M. A. (2005). Relation between intellectual and
523–549. doi:10.1037/0033-2909.134.2.223 metacognitive skills: Age and task differences. Learning and Individual
Säljö, R. (2009). Learning, theories of learning, and units of analy- Differences, 15, 159–176. doi:10.1016/j.lindif.2004.12.001
sis in research. Educational Psychologist, 44, 202–208. doi:10.1080/ Volet, S., Summers, M., & Thurman, J. (2009). High-level co-regulation in
00461520903029030 collaborative learning: How does it emerge and how is it sustained? Learn-
Sawyer, R. K. (2012). Explaining creativity: The science of human innova- ing and Instruction, 19, 128–143. doi:10.1016/j.learninstruc.2008.03.
tion (2nd ed.). New York, NY: Oxford University Press. 001
Sawyer, R. K., & DeZutter, S. (2009). Distributed creativity: How collective Vygotsky, L. S. (1962). Thought and language. Cambridge, MA: MIT
creations emerge from collaboration. Journal of Aesthetics, Creativity, Press.
and the Arts, 3, 81–92. doi:10.1037/a0013282 Vygotsky, L. S. (1978). Mind in society: The development of higher psycho-
Schulz-Hardt, S., Brodbeck, F. C., Mojzisch, A., Kerschreiter, R., & logical processes. Cambridge, MA: Harvard University Press.
Frey, D. (2006). Group decision making in hidden profile situations: Walshaw, M., & Anthony, G. (2008). The teacher’s role in classroom dis-
Dissent as a facilitator for decision quality. Journal of Personality course: A review of recent research into mathematics classrooms. Review
and Social Psychology, 91, 1080–1093. doi:10.1037/0022-3514.91.6. of Educational Research, 78, 516–551. doi:10.3102/0034654308320292
1080 Wang, M. C., Haertel, G. D., & Walberg, H. J. (1990). What influences
Scott, P. (2008). Talking a way to understanding in science classrooms. In learning? A content analysis of review literature. Journal of Educational
N. Mercer & S. Hodgkinson (Eds.), Exploring talk in school (pp. 17–36). Research, 84, 30–43.
London, UK: Sage. Warwick, P., Mercer, N., Kershner, R., & Kleine Staarman, J. (2010).
Sfard, A, (2001). Symbolizing mathematical reality into being: How mathe- In the mind and in the technology: The vicarious presence of the
matical discourse and mathematical objects create each other. In P. Cobb, teacher in pupil’s learning. Computers and Education, 55, 350–362.
K. E. Yackel, & K. McClain (Eds.), Symbolizing and communicating: doi:10.1016/j.compedu.2010.02.001
Perspectives on mathematical discourse, tools, and instructional design Webb, N., Nemer, K., & Ing, M. (2006). Small-group reflec-
(pp. 37–98). Mahwah, NJ: Erlbaum. tions: Parallels between teacher discourse and student behavior in
168 MERCER

peer-directed groups. Journal of the Learning Sciences, 15, 63–119. Wertsch, J. V. (1979). From social interaction to higher psychological pro-
doi:10.1207/s15327809jls1501 8 cesses: a clarification and application of Vygotsky’s theory. Human De-
Wegerif, R. (2007). Dialogic education and technology: Expanding the velopment, 22, 1–22. doi:10.1159/000272425
space of learning. New York, NY: Springer. Whitebread, D., & Pino Pasternak, D. (2010). Metacognition, self-regulation
Wegerif, R. (2010). Mindexpanding: Teaching for thinking and creativity & meta-knowing. In K. Littleton, C. Wood, & J. Kleine Staarman (Eds.),
in primary education. Maidenhead, UK: Open University Press/McGraw International handbook of psychology in education (pp. 615–713). Leeds,
Hill. UK: Emerald.
Wegerif, R., & Scrimshaw, P. (Eds.). (1997). Computers and talk in the Willingham, D. (2009). Three problems in the marriage of science and
primary classroom. Clevedon, UK: Multilingual Matters. education. Cortex, 45, 544–545.
Wells, G. (1986). The meaning makers: Children learning language and Wolpert, D., & Frith, C. (2004). The neuroscience of social interactions:
using language to learn. London, UK: Hodder & Stoughton. Decoding, influencing, and imitating the actions of others. Oxford, UK:
Wells, G. (1999). Dialogic inquiry: Toward a sociocultural practice and Oxford University Press.
theory of education. Cambridge, UK: Cambridge University Press. Woolley, A., Chabrsi, C., Pentland, A., Hashmi, N., & Malone, T. (2010).
doi:10.1017/CBO9780511605895 Evidence for a collective intelligence factor in the performance of human
Wells, G. (2008). Learning to use scientific concepts. Cultural Studies of groups. Science, 330, 686–688. doi:10.1126/science.1193147
Science Education, 3, 329–350. doi:10.1007/s11422-008-9100-6 Zelazo, P., Chandler, M., & Crone, E. (Eds.). (2009). Developmen-
Wells, G. (2009). The meaning makers: Learning to talk and learning to tal social cognitive neuroscience. New York, NY: Psychology Press.
learn (2nd ed.). Bristol, UK: Multilingual Matters. doi:10.1111/j.1750-8606.2012.00241.x