John Geake’s Discussant Comments
Brain, Neuroscience and Education SIG
AERA Montreal April 2005
63.050 Invited Paper Session Developmental Neuroscience: Directions and
Implications for Educational Research
1605 – 1805 Thursday 14th April
Salle de Balle Est, Le Centre Sheraton
Chair: Dr George Hruby, Utah State University
Keynote contributors:
Dr Bill Greenough, University of Illinois
How brains acquire information in development and learning
Dr Liz Spelke, Harvard University
Core knowledge and conceptual change: Number and arithmetic
Dr Lisa Freund, National Institute of Child Health and Human Development
The neurobiology of social interaction and its affect on early learning
Dr Usha Goswami, University of Cambridge, UK
Phonological awareness and reading: Can we use ERP to predict literacy?
Discussant: Dr John Geake, Oxford Brookes University, UK
Discussant Comments:
As an educator who has recently become a cognitive neuroscientist, I found these four
papers to be of considerable interest. In Oxford I convene a Forum of cognitive
neuroscientists and educationists which is particularly concerned with implications of
neuroscientific research for educational practice. In this endeavour we strongly
believe that this should be a two-way street, i.e., that educationists should propose
questions for neuroscientific research projects as much as seek implications for
education from extant experiments which might not have had any educational
rationale in their original formulation. Consequently, in seeking educational
applications of neuroscientific work, the approach of Michael Fullan to educational
change: “Ready, Fire, Aim”, seems to have much to recommend it. For educational
neuroscience to acquire professional credibility on a par with other more established
areas of education, there seems to be virtue in taking action, both experimental and
pedagogic, lest the meetings of this SIG become limited to talk-shops.
To such an end, all four of these papers were rich in potential educational
implications. Child development is a topic of great salience for both educators and
clinicians, not to mention parents. Development has certainly been one area of the
Oxford Forum’s focus, and last July we held a conference for educational
psychologists in the UK which featured several presentations of research into early
child development of memory. Here at this SIG Invited Paper Session, all four papers
were information rich, and 20 minutes was obviously quite an inadequate presentation
time for the audience to fully benefit from the considerable knowledge of the
presenters. As teachers we engage in professional reflection on our successes and
failures, and Liz Spelke laid extensive data of children’s arithmetic development on
the table to inform such professional reflection. In a previous session of this SIG,
George Hruby articulated a range of paradigms within which educators might attempt
to interrogate cognitive neuroscience. I would argue that such paradigms entail
implicit models of brain functioning, and Usha Goswami’s paper presented explicit
data which could be used to challenge or support the robustness of these various
implicit neurological models. Needless to say there is no end point to such enquiry,
and Lisa Freund’s presentation reminded us of the essential scientific regress: the
more we discover the more we find to discover. As a most important case in point,
Bill Greenough drew attention to the relatively neglected contributions of astrocytic
processing to brain development and function. These, it might be reasonably
conjectured, will play critical roles in interpreting future data on neural connectivity,
e.g., the effective bandwidth of inter-modular information transfer. All four papers
demonstrated the centrality of neural connectivity in understanding how functional
modularity contributes to learning. We could hope that future DTI (diffusion tensor
imaging) analyses might shed some much needed light on how distally distributed
networks synchronise to produce educational outcomes.
[AT ABOUT THIS POINT THE SESSION HAD RUN OUT OF TIME AND HAD
TO CLOSE]
In addition to the salient content, which I won’t rehearse, each paper made many
points of contact with the wider research literature, and with several of the papers
presented at earlier sessions of this SIG.
Bill Greenough’s opening remark, that brain development is dynamic, i.e., nonlinear,
cannot be re-emphasised too often. Perhaps at some future AERA conference our
SIG could hold a combined session on nonlinear dynamical models of brain function
with the Chaos and Complexity in Education SIG? However, at this AERA, the
papers of the CCE SIG which I attended consisted mainly of post-modern solipsisms
(complexity in its non-scientific usage), and thus were quite bereft of any
mathematical modelling, or applications to cognition.
Another unfortunate example of research being misinterpreted outside of the science
community is seen with some of the takes on Bill Greenough’s pioneering work on
the comparative effects on brain development of enriched vs. impoverished
environments. Advocates of infant hot-housing fail to equate an enriched laboratory
environment with a normal non-lab environment for rats. The effects of a truly
impoverished human environment, e.g., Romanian orphanages, are all too tragic, but
in a normal home there is more than enough perceptual and physical stimulation for a
young infant, e.g., the contents of the kitchen cupboard, without the imposition of
flash cards, or even Toys-R-Us.
Consistently, precocious intelligence can be explained by a relative advance in neural
maturation, a point also made by Lisa Freund. This connects with the growing VBM
(voxel-based morphometry) literature (e.g., Richard Haier’s lab) on grey matter
correlates of IQ, which has implicated not just the prefrontal cortex but also other
cortical areas, as well as the cerebellum. More specifically, Tuesday’s SIG session
paper by Jodene Fine, University of Texas, reported reading deficits predicted by
differences in structure of the corpus callosum.
However, Bill Greenough’s main thrust was that it’s not just the grey matter that
matters: in fact, for development, astrocytic and capillary plasticity might be of
greater importance. Implications abound. To mention but two: the construction of
fMRI activation maps is usually dependent on the correlated BOLD (blood oxygen
level dependent) response, i.e., a neuronal-driven capillary plasticity. However, in a
clinical setting, diagnostic fMRI might be significantly compromised if the aetiology
involves pre-dilated capillaries within the region of pathology (i.e., there would be no
change in BOLD signal, and thus no fMRI activation in the very region of interest).
A similar problem could arise in educational neuroscientific research when employing
repeated fMRI to seek neural correlates of learning. By Hebbian models, learning
enhances neural efficiency, but such increased neural efficiency might actually
decrease the BOLD signal in the cortical area which is most important. Second, the
so-called brain-based intervention of brain gym assumes a direct causal link between
certain physical (somatic) activities and the stimulation of specific brain areas. In a
dynamic system with feedforward and feedback, and where the circulatory system in
the brain is fractally pervasive, such an assumption seems quite untenable.
Liz Spelke articulated a most fundamental question for educational neuroscience:
given the evolutionary trajectory of the human brain (reviewed by Jeremy Genovese,
Cleveland State University, in an earlier roundtable session), how does our brain
appropriate evolved functions to undertake symbol processing? In the case of
arithmetic, it is clear that human infants, along with other mammals and birds, have a
sense of the numerosity of small groups of objects. How then, thanks to education in
part, do we move on to a complete decimal system of computation, and beyond? And
why do some children seem not to get it as readily as others? Such individual
differences are not apparent when comparing young girls and boys, yet, as Larry
Summers recently and controversially asked, why after decades of affirmative action
programmes do we still find in engineering and physics faculties an unchanged ratio
of men to women about 12 to one?
Clearly this is a question of how development interacts with education? Liz Spelke’s
work sets a basis for such an enquiry. Her model of children’s arithmetic
development seems similar to that of the Maths Recovery programme of Bob Wright,
Southern Cross University, Australia, which has been employed in both Australia and
in the UK with significant positive effect. Moreover, there is a large research
programme at the University of Oxford on cross-modal information processing, and
Liz Spelke’s cross modal data adds to those findings that afferent information
activates cortical regions associated with other modalities en route to being abstracted
for higher-order processing. My take on this, following the dynamic workplace
model of brain function of Dehaene and Changeux, is that the popular brain-based
notion of VAK learning styles is mistaken: there might be acuity-based perceptual
preferences for visual, auditory or kinaesthetic perception, but not for learning per se,
most of which (obvious exceptions not withstanding) is independent of modality.
Lisa Freund‘s overview of the emergent field of educational neuroscience highlighted
its interdisciplinary progress (cf. previous SIG session paper by Michael Atherton), to
be continued by this SIG together with non-AERA groups such as the newly formed
International Mind, Brain and Education Society (IMBES), Harvard University, the
Centre for Neuroscience and Education, University of Cambridge, UK, and the
Oxford Cognitive Neuroscience Education Forum. Lisa Freund also drew attention to
serious challenges to such progress arising from the absence (both theoretical and
actual) of a normative database of brain development. Although we can attribute
development to changes in variables such as cortical volume increases in prefrontal
cortex, due in large part to axon thickening through mylenation, as per Bill
Greenough’s paper, resulting in faster connectivity with better signal to noise ratios,
inter-individual variance seems too high for this to become more than an in-principle
account at present. For example, there is good neuroimaging evidence for a frontalparietal circuit that underpins spatial reasoning, an attribute in which males seem to be
generally advantaged. Interestingly, probability maps of sulci variability show that
the largest variance for males is in the inferior parietal lobe, but how such variability
contributes to advantageous spatial functioning, and how this develops, are important
research questions for the future, as Lisa Freund recommended. Such research would
have particular salience for education, because without such a developmental data
base, neural-based predictions of learning difficulties, e.g., ADHD, are not possible.
Moreover, a greater understanding of brain development should inform our
understanding of the contribution of social interactions, especially parenting. For
children with relatively immature frontal cortices, and therefore limited executive
functioning, it seems critical for parents to provide the extrinsic regulation necessary
for the development of emotional responses (cf. previous SIG session paper by
Kathryn Patten and Stephen Campbell, Simon Fraser University). As an
anthropological footnote, Australian tribal aboriginal parents are held responsible for
their children’s behaviour until the children are initiated as adults, at around age 12
years.
Usha Goswami took up the issue of individual differences in the context of how the
brain appropriates evolved functions for the development of symbol system
processing, as per Liz Spelke’s paper. A pre-literate lexicon of 14,000 words would
seem to place some burden on long-term memory, not less when it needs to be
transformed into literacy. Usha Goswami demonstrated how a scientific dissection of
an educational enterprise – learning to read – could be mapped on to neurophysiology
(cf. previous roundtable paper by Deborah Jensen, Rice University), in this case,
mapping the phonemic patterns involved in reading on to auditory afferents and
consequent ERPs as patterns of brain function. This is potentially informative as such
a phonological lexicon is not explicitly employed in education. Interestingly, the risetime of the auditory signal is a good predictor of reading variance, e.g., the N100 in
the brains of dyslexic children. I wonder if this is an indication of continual novelty
of what should be familiar sounds, which in turn affects access to lexical memory,
similarly to the way in which children who are poor at arithmetic fail to see that, for
example, 19 + 3 calls on the same number facts as 9 + 3? Excitingly, Usha
Goswami’s work has potential informative links with models of culturally-specific
symbol learning in the fusiform gyrus, and research into musical and linguistic
prosody, as well as providing a step towards the establishment of that much-needed
database of children’s brain development.