Reading John Bruer’s book The Myth of the First Three Years was like watching the past few months of my life flash before my eyes. That’s because, gratifyingly, Bruer comes to similar conclusions about the basis of child development theories currently influencing public policy. My one reservation about his book is its title and the repeated references to the ‘myth’ in the text. Technically Bruer is right to use this word for the cluster of popular ideas about the first three years because myths often have some basis in fact, but I think it’s a bit misleading because the word ‘myth’ is also widely used to mean something that has no basis in fact. The myth of the first three years conforms closely to the OECD definition of a ‘neuromyth’ – a “misconception generated by a misunderstanding, a misreading or a misquoting of facts scientifically established (by brain research) to make a case for the use of brain research in education and other contexts”. Other people have other reservations about Bruer’s work. One Amazon.com reviewer complained that the book jacket reveals he’s ‘not even a doctor!’ She obviously missed the significance of Bruer’s presidency of the James McConnell Foundation that awards grants for biomedical and educational research.
Bruer’s book deals with three aspects of what I’ll call the first three years model. He describes how the model migrated from research findings to public policy and the changes it underwent en route. He then explains what the research findings actually were, and what they mean for parents. Although the book was published over a decade ago, the first three years model has, as is traditional, taken a decade to cross the Atlantic, so is now very relevant to the coalition government’s social policy in the UK.
How the first three years model evolved
Bruer cites the developmental psychologist Jerome Kagan’s view that infant determinism – the idea that early experiences have a lasting effect on development – dates from a folk belief in 18th century Europe popularised by writers such as Rousseau. I think that the origins are more likely to be as old as parenthood. Like many of my contemporaries I was brought up on Biblical wisdom and was familiar with sayings such as; “Train up a child in the way he should go: and when he is old, he will not depart from it” (Proverbs 22:6). “He that spareth his rod hateth his son: but he that loveth him chasteneth him betimes” (Proverbs 13:24). The book of Proverbs dates from around 900 BC – the origins of the proverbs are likely to be much older.
Infant determinism was a theme explored by Freud and further developed by figures such as John Bowlby, Mary Ainsworth, Harry Harlow, Rene Spitz and Benjamin Bloom. During this period, as Bruer points out, the study of the brain and the study of behaviour, despite being seen as complementary in the late 19th century, have followed largely independent paths. It’s only since the implications of brain research for human behaviour have become clearer that the two paths have begun to cross more frequently. From the 1970s onwards, fields such as education and child development began to co-opt findings from neuroscience into their own theoretical models. Bruer traces the way neuroscience has been interwoven with child development theories from, for example, the Clinical Infant Development Project that began in the 1970s and developed into Zero to Three: The National Centre for Infants, Toddlers and Families. And from the 1979 book Education and the Brain, through to the I Am Your Child campaign that began in the mid-1990s, and on into public policy.
Bruer identifies three key areas of neuroscience research that underpin the first three years model; neural connections, critical periods and the effects of enriched or complex environments. To illustrate how misconceptions about research have arisen, I’ll summarise the findings he discusses and what they might tell us about child development.
Neural connections Neurons are the cells in the brain that transmit information. They are formed – around 100 billion of them – by the second trimester of gestation, and soon after birth will have migrated to their final destinations in the brain. Neurons are nerve cells and in order to transmit information between each other, they need to be connected. The connections aren’t direct however, they are via synapses – points where the axon terminal of one neuron is close enough to the dendrite of another for chemical or electrical signals to pass from one to the other (see figure). In humans, synapse formation begins a couple of months before birth and peaks at 2-3 years after birth, when the number of synapses is many times greater than it is in adults.
At the same time, two other processes are going on – apoptosis and synaptic pruning. Apoptosis is cell death, and many neurons, like other cells in the body, do die off during normal development. During synaptic pruning the neurons don’t die, but the number of connections the neuron has with other neurons decreases. Between early childhood and puberty, the number of synapses in the brain gradually declines but remains above adult levels. Between puberty and adulthood, synaptic pruning increases rapidly.
Critical periods In early research into the functioning of the visual cortex of the cat, Torsten Hubel and David Wiesel found that if kittens were deprived of visual input from one eye during the first few weeks of life, they remained blind in that eye. It’s long been known that children with congenital cataracts usually develop normal vision if the cataract is removed within the first year after birth. If cataract removal is delayed, vision may never be normal. We know that children can learn to speak a second language like a native if they acquire that language before their teens – if not, the language can still be learned, but with a noticeable accent. So there are critical periods in child development when certain sensory input is necessary for typical development to take place.
Enriched or complex environments Charles Darwin had noted that animals reared in the wild tended to have larger brains than their domesticated offspring. Harlow’s famous experiments demonstrated that baby monkeys reared in isolation showed severe behavioural abnormalities. And experiments in the 1970s conducted by Greenough and colleagues showed that rats reared in complex environments had a greater density of synapses than those raised in solitary conditions. It was clear that the environment has a significant impact on brain development.
Applying the research findings to child development
The model of the first three years has arisen partly because research findings have been misunderstood. Several factors need to be borne in mind before applying the findings to children.
1. Many of the research findings relate to animals, notably rats and monkeys that are often used in research because they are small, easy to handle and have relatively short life-cycles. Rats and monkeys are also social, problem-solving mammals, which means they have a lot in common with humans. Although their brains are very likely to function in similar ways to human brains, there might be species-specific differences. One of the differences is that they develop at different rates; rats are sexually mature by about six weeks of age, and macaques at around five years, so extrapolating findings to humans relies on estimates for the ages at which children reach similar developmental stages.
2. Many of the research findings are estimations. It’s impossible to count the number of synapses or neurons in a brain, so proxy measures are used such as the density of synapses in a small portion of the cortex. But different techniques have been used to arrive at those measures; one group of scientists calculated density of synapses in whole human cortical tissue and found that density peaked at three years of age; others used a different method to calculate the synaptic density in monkeys and estimated that density would peak at seven months in humans. These differences were resolved, but demonstrate that research findings need to be interpreted carefully.
3. Many research findings apply to specific locations in the brain. Huttenlocher, for example, has reported the density of synapses in the human brain peaking at around three months after birth in the auditory cortex, at 8-12 months in the visual cortex but not until 3.5 years in the frontal area. Similarly, synaptic pruning in auditory and visual areas reduces synapse densities to adult levels by 11-12 years of age, but not in frontal areas until several years later.
4. It’s not always clear whether developmental changes in the brain are determined by biological factors, by environmental factors, or by both. Clearly biology would have more influence than environment on the migration of brain cells before birth, but work on synapse formation suggests it is triggered by novel learning experiences. Novel experiences tend to diminish as children get older, and familiar experiences are by definition experienced repeatedly, suggesting a cause for synaptic pruning. However, the increased rate of synaptic pruning after puberty implies that some biological control might be involved.
It’s easy to see how the first three years model came into being. There are differences between the brains of toddlers, older children and adults. It’s tempting to assume that the rapid synapse formation and rapid learning that take place during the first three years of life offer a now-or-never window of developmental opportunity. Synaptic pruning, viewed from a ‘use it or lose it’ perspective, makes it look as if this opportunity soon runs out, despite clear evidence that older children, adolescents and adults can learn a great many things that toddlers can’t. This misconception underpins the current vogue for early interventions. The idea of critical periods has resulted in a tendency for governments (at least in the US and UK) to front-load the education system because of an assumption that the earlier learning takes place the better. What the evidence indicates is that critical periods tend to involve low-level processing of sensory information and that broader ‘sensitive periods’ (e.g. for language acquistion) are not clear-cut. With regard to enriched or complex environments, the enriched environment enjoyed by Greenough’s lab rats was probably a lot less stimulating than the environment experienced by wild rats. Lab rats, however well cared for, tend not to lead a very exciting existence. In other words, children would need to be pretty deprived to experience the kind of environment that resulted in a reduced synapse density.
The transformation of a tentative conclusion based on findings from rats or monkeys into public policy that lacks a sound evidence-base appears to have begun in one of two ways;
• Journalists have accurately summarised research findings but their reports have been misinterpreted by some readers
• Journalists have misunderstood research findings and have inadvertently misled readers.
Casual readers of newspapers or magazines can be forgiven for getting hold of the wrong end of the stick; they often don’t have the time or inclination to check the original data. Policy-makers, in my view, have no such excuse. Governments have enormous resources at their disposal, and academics, however busy, are usually only too keen to explain their findings to willing listeners. There is no good reason why public policy shouldn’t be based on sound evidence.
The first three years model fits easily into our pre-existing mental models. Infant determinism has been seen as the key to solving problems such as poor educational and health outcomes, addiction, anti-social behaviour and criminal activity that preoccupy policy makers – it’s simpler to blame parents for bringing up their children badly than to trace back the complex causes of ‘problem’ behaviours and address them all. The model has also played to the anxieties of parents who understandably want to do the best for their children. The acceptance of the first three years model by parents and policy-makers has resulted in a proliferation of early interventions. So what effect have they had?
The outcomes of early interventions
Evidence from the evaluation of early interventions cited by Bruer, and from studies tracking changes in people’s lives over time suggests that;
• the best predictor of a child’s school performance is mother’s IQ or educational attainment
• good quality childcare improves school performance for children from ‘deprived’ environments (but not from less deprived ones)
• early interventions that focus on verbal skills improve children’s school performance up to about the age of eight – enough to reduce the number of children who repeat the first few school years
• early interventions that focus on verbal skills result in a marginal improvement in school performance, in the length of time spent in education, and in slightly later parenthood
• people tend to settle down as they get older
• chance events play a big part in lifetime achievement.
The pattern that emerges from these evaluations – and this speculative because I don’t have access to raw data – appears to centre on two factors; children’s language skills when they start school, and the way schools educate children. I think these factors have two implications for early interventions.
Firstly, children who are behind their peers with verbal skills when they start school are clearly at a disadvantage, but only because schools use language as the primary medium for education. Language is one of the things that distinguishes homo sapiens from other species. Spoken and written language allow us to communicate vast amounts of information to each other, so it’s obviously desirable that as many children as possible develop good skills in spoken and written language. But being skilled in spoken and written language is not synonymous with learning; children can also learn through practical experience, watching plays or listening to stories. Language skills, although an important component of the school curriculum, shouldn’t be made an obstacle to learning.
Secondly, although some early interventions have had positive outcomes, the effects have tended to be small. Politicians in favour of these programmes would doubtless argue that any positive outcomes are worth it, but given the importance of verbal skills, it might be more cost effective to improve access to speech and language therapy, especially since so many developmental disorders involve speech and language deficits.
What should parents do?
When I was expecting my first child, a friend gave me a book that advocated a ‘best odds’ approach to pregnancy – the author pointed out that nothing can guarantee that a baby will be completely healthy, but there are things that mothers can do that increase the odds in their baby’s favour. It’s a sensible and realistic approach. Bruer’s conclusion is that most parents are doing fine with their children and that we already know what children need and what’s bad for them. He advises parents to interact with their babies, fix known problems – especially visual, auditory and language ones – as soon as possible, and as neuroscientist Steve Petersen put it “Don’t raise your children in a closet, starve them or hit them in the head with a frying pan”.
Bruer, John T., (1999). The myth of the first three years, New York: The Free Press.
Green, E.J., Greenough, W.T. & Schlumpf, B.E. (1983). Effects of complex or isolated environments on cortical dendrites of middle-aged rats, Brain Research, 264,, 233–240.
Huttenlocher, P.R, de Courtena, C., Garey, L.J. & Van der Loosa, H. (1982). Synaptogenesis in human visual cortex — evidence for synapse elimination during normal development, Neuroscience Letters, 33, 247–252.
Organisation for Economic Co-operation and Development (2007) Understanding the brain: the birth of a learning science: The Birth of a New Learning Science: v. 2, OECD.
Image of synapses from Wikipedia.