![Transfer reflected in averaged fNIRS activity. (a) Averaged hemodynamic responses (oxy-hemoglobin [O2Hb]) comparing the groups in which students transferred (gray) or did not transfer (black) the learned knowledge (standard error of the mean). Rest = period after dialogue, Control = control dialogue. (b) Correlation of the averaged hemodynamic responses (oxy-hemoglobin [O2Hb]) for each teacher–student pair, for the groups in which the student transferred (gray dots) the learned knowledge or not (black dots). Lines correspond to linear fits for both groups (After Holper et al., 2013).](https://www.researchgate.net/profile/Cecilia-Calero/publication/262336427/figure/fig1/AS:267559094190113@1440802467423/Transfer-reflected-in-averaged-fNIRS-activity-a-Averaged-hemodynamic-responses_Q320.jpg)
Content uploaded by Cecilia Calero
Author content
All content in this area was uploaded by Cecilia Calero on Nov 23, 2017
Content may be subject to copyright.
MIND, BRAIN, AND EDUCATION
The Cognitive Neuroscience of
the Teacher–Student Interaction
Antonio M. Battro1, Cecilia I. Calero2,AndreaP.Goldin
2, Lisa Holper3, Laura Pezzatti2, Diego E. Shal´
om2,
and Mariano Sigman2,4
ABSTRACT— Pedagogy is the science and art of teaching.
Each generation needs to explore the history, theory, and
practice of the teacher–student interaction. Here we pave
the path to develop a science that explores the cognitive
and physiological processes involved in the human capacity
to communicate knowledge through teaching. We review
examples from our previous work in this research area and
discuss a path to reveal the cognitive and cerebral mechanisms
by which we teach, unfolding a complex operation such as
teaching in its constituents and components.
THE TEACHER–STUDENT INTERACTION
The dialogue between teacher and student is the core of
pedagogy. In a sense, the history of education can be viewed
as a progression and transformation of this basic interaction
(Marrou, 1948). As recently discussed by Watanabe (2013),
teaching may be often confused with simply one means of
enhancing learning. Instead, Watanabe characterizes teaching
as a dynamic phenomenon where interpersonal interactions
occur explicitly and implicitly at multiple levels. Bonding
through coordinated interpersonal interactions occupies
a substantial portion of teaching. Moreover, as recently
emphasized by Strauss and Ziv (2012), teaching is a natural
cognitive ability in humans which develops following a well-
defined path.
Cognitive neuroscience has progressively contributed a
wealth of scientific knowledge to improve education, largely
1Academia Nacional de Educaci´
on, Buenos Aires, Argentina
2Laboratorio de Neurociencia Integrativa, Departamento de F´
ısica,
FCEyN, UBA and IFIBA, Conicet
3Biomedical Optics Research Laboratory (BORL), Division of Neonatol-
ogy, University Hospital Zurich
4Universidad Torcuato Di Tella
Address correspondence to Lisa Holper, Biomedical Optics Research
Laboratory (BORL), Division of Neonatology, University Hospital
Zurich, Frauenklinikstrasse 10, 8091 Zurich, Switzerland; e-mail:
lisa.holper@usz.ch
focused on the learner (Goldin et al., 2013; Lipina & Sigman,
2012). More generally, the neuroscience of how we learn has
been a very active and prominent field of study (Gilbert,
Sigman, & Crist, 2001). This enterprise has largely ignored
the teacher perspective and the complex interaction which
arises in the core educational dyad. We have proposed in other
publications the need to reveal the cognitive and cerebral
mechanismsbywhichweteach,unfoldingacomplexoperation
such as teaching in its constituents and components, as has
been done with other domains of cognition such as attention
(Petersen & Posner, 2012), mathematics (Dehaene, 2009),
and theory of mind (Saxe, Carey, & Kanwisher, 2004). We
have synthesized this effort, coining the term the ‘‘teaching
brain,’’ which is the title of this special series that brings
together scientists and teachers to reconceptualize how we
understand the phenomenon of teaching as an interaction. We
suggest that combined cognitive psychology, neuroscience,
and educational science should be investigated in teacher and
student dialogues (Battro, 2010; Goldin, Pezzatti, Battro, &
Sigman, 2011; Holper et al., 2013).
THE SOCRATIC DIALOGUE
We have paved the path in this direction with a first step
capitalizing in what is probably the most famous educational
dialogue described by Plato in Meno, a teacher–student
interaction between Socrates and an illiterate slave (Crane,
2000). The choice to study this famous lesson of geometry
by Socrates turned out to be of paramount importance, not
only by its symbolic and historical importance in our Western
civilization, but also because it has been recorded in extenso
by Plato in such a way that it could be standardized and tested
today in different cultures.
The Socratic dialogue is a lesson in geometry where the
student learns (or, in the words of Plato, rather ‘‘discovers’’)
how to double the area of a given square, which in essence is a
demonstration of Pythagoras’ theorem. We have standardized
this dialogue in 50 questions and recorded the behavioral
interaction in over 70 pairs of teachers and students during
the spoken dialogue.
©2013 The Authors
Volume 7—Number 3 Journal Compilation ©2013 International Mind, Brain, and Education Society and Blackwell Publishing, Inc. 177
Cognitive Neuroscience of the Teacher–Student Interaction
In the first study (Goldin et al., 2011), we discovered a
remarkable similarity of errors in the reasoning of contempo-
rary students with those committed some 2400 years ago by
the young slave of Meno. This has pertinence for basic science,
speaking about universals in human reasoning. More relevant
to educational practice, we discovered that nearly half the
students, those who followed the dialogue more tightly
without shortcutting questions with prior knowledge, failed
to generalize, that is, to transfer the knowledge to a virtually
identical situation—a square of a different size. These results
questioned the efficacy of the Socratic dialogue, and suggested
that only students with a scaffold of knowledge were the ones
for whom the dialogue was successful. This led to a provocative
hypothesis: students with no prior knowledge, those who are
more engaged during the dialogue, are the ones achieving less
generalization. We see this as a fruitful bridge between edu-
cation (success of the dialogue), cognitive psychology (effect
of prior knowledge on learning), and neuroscience (mental
effort measures revealed by activity in the frontal cortex).
THE SOCRATIC DIALOGUE ASSESSED USING OPTICAL
BRAIN IMAGING
In a second study using a hyperscanning setup based on
functional near-infrared spectroscopy (fNIRS) (Holper et al.,
2013), we examined the hemodynamic correlates of the
Socratic dialogue. The study involved 17 pairs of subjects, in
each session one representing the role of the teacher and the
other the role of the student. During performance on the same
version of Socratic dialogue used in our previous study (Goldin
et al., 2011) we recorded both the student and the teacher
simultaneously using wireless fNIRS sensors (Muehlemann,
Haensse, & Wolf, 2008) over the prefrontal cortex. Owing
to the relative high tolerance of the wireless fNIRS sensors
to motion, subjects were allowed to move their hands,
speak, and interact with each other. In addition, a control
condition was included consisting of an approximately 10-min
dialogue written on paper (Plato, 2008), which resembles the
sort of responses, answers, and timing of the experimental
dialogue. In this control dialogue, students and teachers were
instructed to simply read aloud, each one taking over a part of
this dialogue. Before and after the dialogues two rest periods
(2 min each) were included as a baseline condition.
To assess the interaction between the teacher–student pairs
during the experimental conditions as well as the transfer
of knowledge, subsequent statistical analysis performed on
the averaged prefrontal hemodynamic activity considered the
factor ‘‘Transfer’’ (Yes versus No).
Our results showed two main findings. First, we observed
that students who successfully transferred the knowledge
showed smaller prefrontal hemodynamic responses during
performance on the Socratic dialogue than those who did
Fig. 1. Transfer reflected in averaged fNIRS activity. (a) Averaged
hemodynamic responses (oxy-hemoglobin [O2Hb]) comparing the
groupsinwhichstudentstransferred(gray) or did not transfer (black)
the learned knowledge (standard error of the mean). Rest =period
after dialogue, Control =control dialogue. (b) Correlation of the
averaged hemodynamic responses (oxy-hemoglobin [O2Hb]) for
each teacher–student pair, for the groups in which the student
transferred (gray dots) the learned knowledge or not (black dots).
Lines correspond to linear fits for both groups (After Holper et al.,
2013).
not transfer. Figure 1a illustrates the differences in the
hemodynamic response, distinguishing the groups in which
students showed transfer from those who did not show
transfer. In particular, these results showed that teachers
teaching students eliciting transfer showed comparable
levels of activation to those who taught students not
eliciting transfer (Figure 1a, left). Hence, the hemodynamic
activity of the teachers was not dependent on the transfer
aspect. In contrast, students showed a remarkable different
hemodynamic response depending on whether they could
transfer or not. In particular, while students who did transfer
of knowledge elicited significant lower levels of hemodynamic
activity during the whole Socratic dialogue, students that
could not transfer showed higher levels of activity during
the entire dialogue (Figure 1a, right). In agreement with our
prediction and closing the bridge, this indicated that those
students with less prefrontal activity were the ones who
showed more learning.
Moreover, above and beyond the main effect of reduced
frontal activity, we considered the fact that students and
teachers were recorded simultaneously. In particular, we
investigated whether there was a correlation in the level
of hemodynamic activity between teachers and students. The
results demonstrated a strong positive correlation in activity
178 Volume 7—Number 3
Antonio M. Battro et al.
between the students and the teachers in efficient educational
dialogues (in which the student transferred the knowledge)
(Figure 1b). In the same group of students a similar effect was
foundduring the control dialogue, whereas during rest, activity
was uncorrelated. These findings indicated that whenever a
student showed greater activity (compared to the average of
the student population) the teacher also showed greater levels
of activity (relative to the average of the teacher population).
The opposite effect was observed for the group of students
who did not transfer, that is, a negative correlation indicating
that whenever the teacher showed greater activity, this was
accompanied by a decrease in activity from the corresponding
student in the Socratic dialogue.
Second, we explored the temporal dynamics of the
hemodynamic response during the Socratic dialogue. Figure
2 illustrates the time courses of both the averaged as
well as the correlated hemodynamic activity. The averaged
hemodynamics showed an interesting pattern. At about 60%
of the dialogue, we observed a discontinuity, revealing a small
(non-significant) effect of the students’ transfers within the
teachers’ fNIRS signal (marked by the arrow in Figure 2a).
This time point of the Socratic dialogue corresponds roughly
to the question when the crucial ‘‘diagonal argument’’—which
is in the essence of the final line of reasoning (how to double
a square)—begins to be outlined. Similarly, we investigated
whether the correlation effect was consistent throughout the
Socratic dialogue by calculating correlation coefficients for
both groups in sliding windows of 1% of the total duration of
the Socratic dialogue. The results were very consistent with
what we observed in the average hemodynamic changes. Cor-
relations were non-significant in the beginning of the dialogue
(Figure 2b). At about 60% of Socratic dialogue, in coincidence
with the discontinuity observed in the average activity and
corresponding to the critical moment of reasoning throughout
the dialogue, we observed that correlations became more
significant revealing opposite effects: positive correlations
for the group of students who did transfer and negative
correlations for those who did not transfer. These results may
indicate that a sort of cortical coupling between students and
teachers is involved in successful educational interactions.
Taken together, the fNIRS study demonstrated that
brain measures signaling relevant pedagogical variables (the
transfer) can be obtained in a—though experimental—but
realistic educational dialogue. On the basis of previous
work (Rodriguez, 2013) that proposed by the structure of
the human nervous system and its sensing, processing, and
responding components as a framework for a reconceptualized
teaching system, our results may pave the path for a program
investigating brain activity in real educational setups where
knowledge is acquired in complex processes involving the
synchronistic interaction between teachers and students to
achieve optimal teaching and learning experience.
Fig. 2. Transfer reflected in temporal dynamics of fNIRS activity.
(a) Averaged time course of the Socratic dialogue in normalized
time (0 and 1 correspond to beginning and end of all dialogues)
are shown for teachers and students separated for the groups in
which students transfer (gray) or did not transfer (black). The arrow
roughly corresponds to the normalized time of the question when
the crucial ‘‘diagonal argument’’—which is in the essence of the final
line of reasoning (how to double a square)—begins to be outlined.
(b) Socratic dialogue time course of the correlation coefficients in
normalized time (0 and 1 correspond to the beginning and end of all
dialogues). Time course of the p-values are log-transformed. Dashed
line corresponds to p=.05 (After Holper et al., 2013).
TEACHING BRAIN CONSORTIUM
The opening sentences of the Socratic dialogue, a dialogue
about virtue, are still engaging us into further questions
concerning the origins of learning and the scope of teaching.
Meno asked Socrates ‘‘whether virtue is acquired by teaching
or practice; or if neither by teaching nor practice, then whether
it comes to man by nature, or in what other way.’’ Today the
old alternative ‘‘teaching or practice’’ can be replaced by the
pedagogy of ‘‘learning by doing’’ and the option of a ‘‘natural’’
acquisition of knowledge can be understood as the result of
genetic and epigenetic processes. The Socratic dialogue was,
as such, a pedagogical procedure with strong philosophical
implications, like the theories of reminiscence that cannot
be accepted without difficulty by modern science. Moreover,
we have discovered that a classical Socratic dialogue has also
serious limitations as a pedagogical method because it cannot
ensure a stable cognitive acquisition at the end of the ‘‘lesson.’’
The student may fail to generalize the cognitive procedure
Volume 7—Number 3 179
Cognitive Neuroscience of the Teacher–Student Interaction
in a different setting (of course this failure on transfer of
knowledge is common to other pedagogical methods too). To
this observation regarding the behavior of the student we add,
for the first time, the neurobiological component that may
predict the success or failure of the whole lesson, but we need
more research to understand this important fact. In this sense,
considering all potential contributions from different research
areas, we propose the idea of a Teaching Brain Consortium.In
what follows, we discuss some important points in order to
establish such a common platform that could support further
explorations.
NEW PEDAGOGICAL ENVIRONMENTS
A key point of a Teaching Brain Consortium would be to
establish bridges between the experimental laboratory and
the classroom. It is important to develop a team work
of scientists and teachers, which establishes the school as
the place of choice for research in teaching and learning.
The accelerated progress in mobile, wearable, and low-cost
equipment of neurophysiology and brain imaging will reach
the classroom soon, as did the information and communication
technologies some two decades ago (Yano, 2013). Many
schools today, for instance, have already implemented complex
digital laboratories on robotics that may show the way to
introduce brain research in a real pedagogical environment.
Of course, this will raise some important ethical questions
regarding privacy, among other concerns that should be taken
into account (Lopez-Rosenfeld, Goldin, Lipina, Sigman, &
Fernandez Slezak, 2013).
The Socratic dialogue described above constitutes an
example of how education and neuroscience can be bridged.
It is very important to be aware, as identified in John Bruer’s
seminal paper (Bruer, 1997), whether observation of brain
activity adds substantial additional information relevant for
educational practice, or if instead (as in the majority of the
cases as demonstrated by Bruer) it merely serves to ground
ideas established by cognitive psychology which are relevant
to basic neuroscience but not to education.
NEW METHODS OF ASSESSMENT
A second point that could be addressed by a Teaching
Brain Consortium would be the development, evaluation, and
introduction of new methods of assessments in the classroom.
In common practice the fact that we can establish the
difference between learning and understanding may suffice to
evaluate the student. The important novelty, however, is that
with a careful monitoring of brain activity of both student and
teacher we could predict above chance which student will fail
or succeed in the generalization test (Holper et al., 2013). This
is totally new information based on the evidence of scientific
records. It is certainly not a kind of ‘‘lie detector’’; the student
is perfectly convinced of the truth of his or her sayings in
both cases, but the brain images offer a ‘‘plus’’ that is hidden
to everyone and not detectable by the common test. This is
similar to the information that can be obtained by analyzing
gestures that students do with their hands when they talk,
which is typically not consciously accessible and yet provides
an early sign that the speaker is ready to learn a particular task
(Broaders, Wagner Cook, Mitchell, & Goldin-Meadow, 2007).
Our example is very modest, but we think it is compelling
enough to inform about a failure or a success. We can imagine
in the future that new kinds of tests based on brain imaging
will add evidence to assess the quality of the learning process,
as we did in our case with the generalization test.
Acknowledgments—This work is funded by CONICET and
UBACYT. Mariano Sigman is sponsored by the James
McDonnell Foundation 21st Century Science Initiative in
Understanding Human Cognition—Scholar Award.
REFERENCES
Battro, A. (2010). The teaching brain. Mind, Brain, and Education,4,
28–33. doi:10.1111/j.1751-228X.2009.01080.x
Broaders, S., Wagner Cook, S., Mitchell, Z., & Goldin-Meadow, S.
(2007). Making children gesture brings out implicit knowledge
and leads to learning. Journal of Experimental Psychology: General,
136, 539–550.
Bruer, J. (1997). Education and the brain: A bridge too far. Educational
Researcher,26(8), 4–16.
Crane, G. (2000). (Ed.). PerseusDigitalLibraryProject. Tufts University,
http://www.perseus.tufts.edu
Dehaene, S. (2009). Origins of mathematical intuitions: The case of
arithmetic. Annals of the New York Academy of Science,1156, 232–259.
Gilbert, C., Sigman, M., & Crist, R. (2001). The neural basis of
perceptual learning. Neuron,31, 681–697.
Goldin, A., Pezzatti, L., Battro, A., & Sigman, S. (2011). From
ancient Greece to modern education: Universality and lack of
generalization of the Socratic dialogue. Mind, Brain, and Education,
5, 180–185.
Goldin, A., Segretin, M., Hermida, M., Paz, L., Lipina, S., & Sigman, M.
(2013). Training planning and working memory in third graders.
Mind, Brain, and Education,7, 136–146.
Holper, L., Goldin, A., Shalom, D., Battro, A., Wolf, M., & Sigman,
M. (2013). The teaching and the learning brain: A cortical
hemodynamic marker of teacher–student interactions in the
Socratic dialog. International Journal of Educational Research,59,
1–10.
Lipina, S., & Sigman, M. (2012). La pizarra de Babel. Puentes entre
neurociencia,psicolog´
ı a y educaci´
on. Buenos Aires, Argentina: Libros
del Zorzal.
Lopez-Rosenfeld, M., Goldin, A., Lipina, S., Sigman, M., & Fernandez
Slezak, D. (2013). Mate Marote: A flexible automated framework
for large-scale educational interventions. Computers &Education,
68,307–313.
180 Volume 7—Number 3
Antonio M. Battro et al.
Marrou, H. (1948). Histoire de l’´
education dans l’Antiquit´
e[A history of
education in antiquity]. Paris, France: Le Seuil. Madison, WI:
University of Wisconsin Press.
Muehlemann, T., Haensse, D., & Wolf, M. (2008). Wireless
miniaturized in-vivo near infrared imaging. Optics Express,16(14),
10323–10330.
Petersen, S., & Posner, M. (2012). The attention system of the human
brain: 20 years after. Annual Review of Neuroscience,35,73–89.
Plato. (2008). Apolog´
ıa de S
´
ocrates Men´
on. Cr´
atilo. Madrid, Spain:
Alianza Editorial.
Rodriguez, V. (2013). The human nervous system: A framework for
teaching and the teaching brain. Mind,Brain,and Education,7,2–12.
doi:10.1111/mbe.12000
Saxe, R., Carey, S., & Kanwisher, N. (2004). Understanding
other minds: Linking developmental psychology and functional
neuroimaging. Annual Review of Psychology,55,87–124.
Strauss, S., & Ziv, M. (2012). Teaching is a natural cognitive ability for
humans. Mind, Brain, and Education,6, 186–196. doi:10.1111/j.1751-
228X.2012.01156.x
Watanabe, K. (2013). Teaching as a dynamic phenomenon with
interpersonal interactions. Mind, Brain, and Education,7, 91–100.
doi:10.1111/mbe.12011
Yano, K. (2013). The science of human interaction and teaching. Mind,
Brain, and Education,7, 19–29. doi:10.1111/mbe.12003
Volume 7—Number 3 181
