apprentice without a sorcerer

Cummings’ essay Some Thoughts on Education and Political Priorities highlights his admiration for experts, notably scientists, but this doesn’t prevent him making several classic novice errors. These errors, not surprisingly, lead Cummings to some conclusions contradicted by evidence he hasn’t considered. I’ve focused on four of them.

oversimplifying systems

Cummings knows that systems operate differently at different levels, and although all systems, as part of the physical world involve maths and physics, you can’t reduce all systems to maths and physics (p.18). But his preoccupation with maths and physics, and lack of attention to the higher levels of systems suggest he can’t resist doing just that. In his essay maths is mentioned 473 times (almost 2 mentions per page) and physics 179 times. Science gets 507 references and quantum 238. In contrast, the arts get 8 mentions and humanities 16. Ironically, given his emphasis on complex systems, Cummings seems determined to view complex knowledge domains like education, politics, the humanities and arts, only through the lenses of maths, physics and linear scales.

Cummings’ first degree is in history, but he knows a lot of scientific facts. How deep his understanding goes is another matter. He opens the section on a scientific approach to teaching practice with the famous ‘Cargo Cult’ speech in which Richard Feynman accused educational and psychological studies of mimicking the surface features of science but not applying the deep structure of the scientific method (p.70). Cumming’s criticism is well-founded; evidence has always influenced educational practice in the UK, but the level of rigour involved has varied considerably. Ironically, Cummings’ appeal to scientific evidence then itself sets off down the cargo-cult route.

misunderstanding key concepts: chunking vs schemata

Cummings claims “experts do better because they ‘chunk’ together lots of individual things in higher level concepts – networks of abstractions – which have a lot of compressed information and allow them to make sense of new information (experts can also use their networks to piece together things they have forgotten)” (p.71).

‘Chunking’ occurs when several distinct items of information are perceived and processed as one item. The research e.g. Miller (1956), De Groot (1965) and Anderson (1996), shows it happens automatically after groups of low-level (simple) items with strongly similar features have been encountered very frequently, e.g. Morse code, words, faces, chess positions. I’ve not seen any research that shows the same phenomenon happening with information that’s associated but complex and dissimilar. And Cummings doesn’t cite any.

Information that’s complex and dissimilar but frequently encountered together (e.g. Periodic Table, biological taxonomy, battle of Hastings) forms strong associations cognitively that are configured into a schema. What Cummings describes isn’t chunking; it’s the formation of a high level schema. Chunks are schemata, but not all schemata are chunks.

Cummings is right that experts abstract information to form high level schemata, but the information isn’t compressed as he claims. The abstractions are key features of aspects of the schema e.g. key features of transition metals, birds or invasions.  I can just about hold all the key features of birds in my working memory at once, but not at the same time as exceptions (e.g ostrich, penguin) or features of different bird species. The prototypical features make it easier to retrieve associated information, but it isn’t retrieved all at once. If I think about the key features of birds, many facts about birds and their features spring to mind, but they do so sequentially, not at the same time. The limitations of working memory still apply.

The distinction between chunking and schema formation is important because schemata play a big part in expertise e.g. Schank & Abelson (1977) and Rumelhart (1980). Despite their importance, Cummings refers to schemata only once, when he’s describing how his essay is structured (p.7). The omission is a significant one with implications for Cumming’s model of how experts structure their knowledge.

experts vs novices

Experts in a particular field derive their expertise from a body of knowledge that’s been found to be valid and reliable. They construct that knowledge into schemata, or mental models. New knowledge can then be incorporated into the schemata, which might then need to be configured differently. Sometimes experts disagree strongly, not about the content of their schemata, but about how the content is configured.

The ensuing debates can go on for decades. A classic example is the debate between those who think correlations between intelligence test scores indicate that intelligence is a ‘something’ that ‘really exists’, and those who think the assumption that there’s a ‘something’ called intelligence, shapes the choice of items in intelligence tests, so correlations should come as no surprise (see previous post). Another long-standing debate involves those who think universal patterns in the structure of language mean that language is hard-wired in the brain, versus others who think the patterns emerge from the way networks of neurons compute information.

Acquiring key information about an unfamiliar knowledge domain takes time and effort, and Cummings has obviously put in the hours. What’s more challenging is finding out how domain experts configure their knowledge – experts often take their schemata for granted and don’t make them explicit. Sometimes you need to ask directly (or be told) why knowledge is organized in a certain way, and if there are any crucial differences of opinion in the field.

Cummings doesn’t seem to have asked how experts structure their knowledge. Instead, he appears to have squeezed knowledge new to him (e.g. chunking) into his own pre-existing schema without checking whether his schema is right or wrong. Or, he’s adopted the first schema he’s agreed with (e.g. genes and IQ). He admits to basing his genes/IQ model largely on Robert Plomin’s Behavioural Genetics and talks by Stephen Hsu. He dismisses the controversies and takes Plomin and Hsu’s models for granted.

evaluating evidence

There are references to the scientific method in Cummings’ essay but they’re about data analysis, not the scientific method as such. A crucial step in the scientific method is evaluating evidence – analysing data for sure, but also testing hypotheses by weighing up the evidence for and against. This process isn’t about ‘balance’ – it’s about finding flaws in methods and reasoning in order to avoid confirmation bias.

But Cummings repeatedly accepts evidence in support of one thing or against another, without questioning it. I’d suggest he can’t question much of it because he doesn’t know enough about the field. Some that caught my eye are:

  • Assuming hunter-gatherers’ knowledge is “based on superstition (almost total ignorance of complex systems)” (p.1). Anthropology that might claim otherwise, is like other social sciences, summarily dismissed by Cummings.
  • Unsubstantiated claims such as “Aeronautics was confined to qualitative stories (like Icarus) until the 1880s when people started making careful observations and experiments about the principles of flight” (p.21). Da Vinci, Bacon, Montgolfiers, Caley? No mention.
  • Attributing European economic development between 14th and 19th centuries to ‘markets and science’ and omitting the role of the Reformation, French Revolution, or Enclosure Acts (p.108).
  • Uncritical acceptance of Smith’s and Hayek’s speculative claims about the benefits of markets (p.106).
  • Overlooking systems constraints on growth – in corn yields, computing power etc. (pp.46, 231-2). No mention of the ubiquitous sigmoid curve.
  • Overlooking the Club of Rome’s Limits to Growth when discussing shortage and innovation (p.112).
  • Emphasising the importance of complex systems with no mention of systems theory as such (e.g. Bertalanffy’s general systems theory).
  • Ignoring important debates about construct validity e.g. intelligence and personality (p.49).

not just wrong

People are often wrong about things and usually a few minor errors don’t matter. In Cummings’ case they matter a great deal, partly because he’s so influential, but also because even tiny errors can have huge consequences. I chose the example of chunking because Cummings’ interpretation of it has been disproportionately influential in recent English education policy.

Daisy Christodoulou in Seven Myths about Education (2014) takes the assumption about chunking a step further. She’s right that chunking low-level associations such as times tables allows us to ‘cheat’ the limitations of working memory, but wrong to assume (like Cummings) high-level schemata do the same. And flat-out wrong to claim “we can summon up the information from long-term memory to working memory without imposing a cognitive load.” (Christodoulou p.19, my emphasis). Her own example (23,322 x 42) contradicts her claim.

Christodoulou’s claim is based on Kirschner, Sweller & Clark’s 2006 paper ‘Why minimal guidance during instruction does not work’. The authors say; “The limitations of working memory only apply to new, yet to be learned information that has not been stored in long-term memory. New information such as new combinations of numbers or letters can only be stored for brief periods with severe limitations on the amount of such information that can be dealt with. In contrast, when dealing with previously learned information stored in long-term memory, these limitations disappear.” (Kirschner et al p.77).  The only evidence they cite is a 1995 review paper proposing an additional cognitive mechanism “long-term working memory”.

I have yet to read a proponent of Kirschner, Sweller & Clarke’s model discuss the well-known limitations of long-term memory, summarised here. Greg Ashman for example, following on from a useful summary of schemata, says;

One way of thinking about the role of long-term memory in solving problems or dealing with new information is that entire schema can be brought readily into working memory and manipulated as a single element alongside any new elements that we need to process. The normal limits imposed on working memory fall away almost entirely when dealing with schemas retrieved from long-term memory – a key idea of cognitive load theory. This illustrates both the power of having robust schemas in long-term memory and the effortlessness of deploying them; an effortlessness that fools so many of us into neglecting the critical role long-term memory plays in learning”.

Many with expertise as varied as English, history, physics or politics, have enthusiastically embraced findings from cognitive science that could improve the effectiveness of teaching. Or more accurately, they’ve embraced Kirschner, Sweller and Clarke’s model of memory and learning.  Some of the ‘cog sci’ enthusiasts have gone further. They’ve taken a handful of facts out of context, squeezed them into their own pre-existing schemata, and drawn conclusions that are at odds with the research. They’ve also assumed that if an expert in ‘cog sci’ makes a plausible claim it must be true, but haven’t evaluated the evidence cited by the expert – because they don’t have the relevant expertise; cognitive science is a knowledge domain unfamiliar to them.

Nevertheless objections to the Kirschner, Sweller and Clarke model are often dismissed as originating either in ideology or ignorance. Ironic, as despite emphasising the importance of knowledge, evidence and expertise, many of the proponents of ‘cog sci’ are patently novices selecting evidence to support a model that doesn’t stand up to scrutiny. Murray Gell-Man is right that we need people who can take a crude look at the whole of knowledge (p.5), but the crude look should be one informed by a good grasp of the domains in question.

In 1797, Goethe published a poem entitled Der Zauberlehrling (Sorcerer’s Apprentice). It was a popular work, and became even more popular in 1940 when animated as part of Disney’s Fantasia, with Mickey Mouse playing the part of the apprentice who started something he couldn’t stop. The moral of the story is that a little knowledge can be a dangerous thing. Cummings has been portrayed as a brilliant eccentric and/or an evil genius. I think he’s an apprentice without a sorcerer.

references

Anderson, J (1996) ACT: A simple theory of complex cognition, American Psychologist, 51, 355-365.

Christodoulou, D (2014).  Seven Myths about Education.  Routledge.

de Groot, A D (1965).  Thought and Choice in Chess.  Mouton.

Kirschner, PA, Sweller, J & Clark, RE (2006). Why Minimal Guidance During Instruction Does Not Work: An Analysis of the Failure of Constructivist, Discovery, Problem-Based, Experiential, and Inquiry-Based Teaching Educational Psychologist, 41, 75-86.

Miller, G (1956). The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information, Psychological Review, 63, 81-97.

Rumelhart, DE (1980). Schemata: the building blocks of cognition. In R.J. Spiro et al. (eds) Theoretical Issues in Reading Comprehension.  Lawrence Erlbaum: Hillsdale, NJ.

Schank, RC & Abelson, RP (1977). Scripts, Plans, Goals and Understanding: an Inquiry into Human Knowledge Structures.  Lawrence Erlbaum: Hillsdale, NJ.

 

 

 

not all in the genes

Dominic Cummings’ 2013 essay Some Thoughts on Education and Political Priorities reveals his keen interest in the implications of intelligence research for education. His Endnote “Intelligence, IQ, genetics, and extreme abilities” (p.194) runs to 17 pages.

General Intelligence

If I’ve understood Cummings’ model of intelligence correctly, it goes like this: General Intelligence (‘g’) is a trait that’s largely genetically determined and can be measured as IQ. If we could identify the genes involved, we could spot those with high cognitive ability who are needed to find the solutions to the complex problems facing us.

There’s certainly robust evidence that cognitive ability is largely genetically determined (by multiple genes), remains stable, and is a good predictor of lifetime achievement (p.197). We do need people with high IQs to work on solutions to world problems. And children with high IQs need an appropriate education. I share Cummings’ frustration that DfE officials prioritised their notion of equality over the need to develop talent (p.64). But his model is also flawed at several levels. It includes three key components that are worth examining in more detail;

  • A hypothetical human trait – general intelligence
  • The correlation between factors within intelligence tests
  • IQ

Intelligence

Towards the end of the 19th century, researchers got very interested in measuring human characteristics. Some, such as height and weight, were easy to measure, but others – like ‘physiognomy’ or ‘eventuality’- were trickier because it wasn’t obvious what the features of ‘physiognomy’ or ‘eventuality’ were.

PhrenologyPix

You can of course measure any human characteristic you fancy. You decide what the features of ‘adhesiveness’ or ‘ideality’ are and how to measure them, and hey presto! you’ve measured ‘adhesiveness’ or ‘ideality’. There might of course be some disagreement about the features of ‘adhesiveness’ or ‘ideality’ – or even about their very existence.

Also in the late 19th century, industrialised economies were desperate for a literate, numerate, ‘intelligent’ workforce. That requirement was one of the drivers for mass education.

In his 1904 review of measures of intellectual ability, the psychologist and statistician Charles Spearman decided intellectual ability could be measured using performance in: Classics, Common Sense, Pitch Discrimination, French, Cleverness, English, Mathematics, Pitch Discrimination among the uncultured, Music, Light Discrimination and Weight Discrimination (Spearman p.276). Essentially, he defined intelligence in terms of intellectual abilities. More recent measures such as Verbal Comprehension, Visual Spatial, Fluid Reasoning, Working Memory, and Processing Speed (Wechsler Intelligence Scale for Children – V) define intelligence in terms of cognitive abilities.

‘g’

Spearman went a step further. The positive correlations between the factors in his test convinced him “that there really exists a something that we may provisionally term ‘General Sensory Discrimination’ and similarly a ‘General Intelligence’” (Spearman p.272). And the correlations between scores in cognitive ability tests have convinced others of the existence of a ‘something’ we may provisionally term ‘general intelligence’.

I haven’t been able to find out if Spearman used ‘g’ to refer to the correlation between factors, or the hypothesized ‘something’, or both. Whichever it was, critics were quick to point out that correlation doesn’t indicate causality. A positive correlation between Spearman’s factors exists, certainly. Whether ‘general intelligence’ exists other than as a folk concept is another matter.

Critics also pointed out the circularity in Spearman’s argument. Intelligence tests were assumed to measure intelligence, but because no one knew what intelligence actually was, the tests also defined intelligence – even if they varied considerably. Spearman’s measures were very different to Binet & Simon’s , and neither bears much resemblance to the WISC, or to Raven’s Progressive Matrices. As Edwin Boring put it in 1923, “intelligence is what the tests test”.

IQ

In 1912, the German psychologist William Stern developed the concept of IQ –Intelligenzquotient. IQ (initially mental age divided by chronological age, expressed as a percentage) tells you how an individual’s test score compares to the average for the population. But the criticisms of ‘intelligence’ also apply to IQ. IQ tests undoubtedly measure aspects of cognitive ability, but we don’t know whether or not they measure a genetically determined trait we may call ‘intelligence’. Or even if such a trait exists.

Advocates for general intelligence haven’t take the criticisms lying down. Cummings quotes Robert Plomin’s dismissal of the circularity criticism: “…laypeople often read in the popular press that the assessment of intelligence is circular – intelligence is what intelligence tests assess. On the contrary, g is one of the most reliable and valid measures in the behavioral domain” (p.195).

It’s worth noting that Plomin uses g and intelligence interchangeably, even though intelligence is a hypothesized trait and he refers to g as a measure. There’s no doubt that g is reliable and valid when measuring some cognitive abilities. Whether those abilities represent a genetically determined trait we may term ‘intelligence’ is another matter – which Plomin goes on to admit: “It is less clear what g is and whether g is due to a single general process, such as executive function or speed of information processing, or whether it represents a concatenation of more specific cognitive processes…” It’s also worth noting that Plomin attributes the circularity argument to laypeople and the popular press, rather than to generations of doubting academic critics.

The implicit assumptions made by those emphasizing the importance of g and IQ, are important because they can have unwanted and unintended outcomes. One is that correlations between factors might hold true at population level, but not always at the individual level. Deidre Lovecky, who runs a resource centre in Providence Rhode Island for gifted children with learning difficulties, reports in her book Different Minds having to pick ‘n’ mix sub-tests from different assessment instruments because individual children were scoring at ceiling on some sub-tests and at floor on others. How intelligent are those children? Their IQ scores wouldn’t tell us.

Also, hunting for hypothetical snarks can waste a huge amount of time and resource. It’s taken over a century for us not to be able to find out what ‘g’ is. Given the number of genes involved ,you’d think by now people would have abandoned the search for a single causal factor. It’s a similar story for chronic fatigue syndrome (‘neurasthenia’ – 1869) and autism (‘autistic disturbances of affective contact’ – 1943); both perfectly respectable descriptive labels, but costly red herrings for researchers looking for a single cause.

Characteristics, traits, states, and behaviours

What convinces Cummings that intelligence, g and IQ are ‘somethings’ that really exist is evidence from behavioural genetics. Scientists working in this field have established beyond reasonable doubt that most of the variance in human intelligence, however you measure it, is accounted for by genetic factors. That shouldn’t be surprising. Intelligence is almost invariably defined in terms of cognitive ability, and cognitive ability emerges from characteristics such as visual and auditory discrimination, reaction time, and working memory capacity, all biological mechanisms largely determined by genes.

But not all human characteristics are the same kind of thing. Some characteristics such as height and weight are clearly physical and are easily measured. For obvious reasons genes account for most of the variance in physical characteristics.

The term trait applies to physical characteristics but also to stable dispositional characteristics. Disposition refers to people’s behavioural tendencies – how introvert or extravert they are, what they like and dislike, do and don’t do etc. The evidence from behavioural genetics suggests that genes also account for most of the variance in stable traits.

States are also dispositional characteristics, but they’re temporary and usually emerge in response to environmental factors. So Joan might be extravert and prone to angry outbursts, and Felicity might be introverted and timid, but both of them are likely to become anxious if fire breaks out in the office they share. Their reactions to the fire are largely genetically determined, but are triggered by an environmental event.

Behaviours are things people do. They are undoubtedly influenced by genetic makeup, but occur primarily in response to environmental factors, because that’s the main function of behaviour. Joan might try to extinguish the fire and Felicity might take the nearest exit, but both behaviours would be in response to specific circumstances. If we were pre-programmed automatons, the human race wouldn’t have lasted very long.

In support of his genes-determine-intelligence argument Cummings cites Stephen Hsu, a physicist turned behavioural geneticist, who claims that much of the nature/nurture debate has been settled. Hsu’s right in respect of the genetic influence on traits. But that still leaves plenty of room for the environmental influence on states and behaviours. That has significant implications for Cummings’ model of education.

Genes, intelligence and education

The principal components of Cummings’ model of education are genes, intellectual ability, effective teaching, and exam results. But in real life many other factors impact on educational outcomes. Take Ryan, Joan’s nephew, for example.

Ryan lives with his mum, a single parent. She cares for her father, disabled following a work accident, and her mother who has complex health problems. They live in a former industrial town, currently in economic decline. Ryan’s parents’ relationship broke down due to the financial and time pressures on the family.

Ryan has average intellectual ability, but episodes of glue ear when he was younger left him with a slight speech and language delay. He struggled with maths and reading and was often reprimanded for not following instructions. He loved physical activities, but the regulatory education framework required Ryan, as a child who was ‘falling behind’, to do less practical activity and more arithmetic and phonics.

Ryan soon began to disengage with school. He was referred for speech and language therapy and to the educational psychologist, but both had lengthy waiting lists. By his teens, Ryan had a low reading age, was making slow progress academically, and skipped school whenever he could. His mum couldn’t find paid work to fit around caring for her parents, and was on medication for anxiety and depression.

Genes undoubtedly account for some challenges faced by Ryan and his family; his family’s health, his intellectual ability, and quite likely his glue ear. But environment plays a significant role in the shape of income, diet, viral infections, and national economic, social, and education policy. So do life events (so commonplace their importance is often overlooked); where the family happens to live, grandfather’s accident, parents’ break-up, which school is closest to home.

Then there are specific behaviours on the part of Ryan, his parents, grandparents, teachers – and government ministers. Specific behaviours are often framed as a ‘choice’, but that choice is often highly constrained by circumstances.

Choose your metrics

Cummings measures the effectiveness of the education system by exam results (although he questions the quality of the exams). Exam results are positively correlated with IQ, and IQ is largely genetically determined. So his choice of metric means Cummings places a disproportionate emphasis on influence of genes on educational outcomes.

Of course there’s nothing wrong with IQ or exam results as metrics. If you want to find someone with good cognitive abilities, a modern intelligence test can identify them. If you want candidates with a mathematical ability of at least GCSE level, check out GCSE maths results.

But the choice of a single metric for something as complex as an education system shows an inadequate understanding of complex systems. And begs the question of what education is about. If quality of life in local communities were the key metric, the education system would look very different. By bizarre coincidence, the gene pool of large populations produces people with a wide range of abilities and aptitudes, just what those populations need in order to thrive. That wide range of abilities and aptitudes should be cultivated. Cummings’ choice of metric means the exam-results tail wagging the quality-of-life dog.

Accommodating a wide range of abilities and aptitudes doesn’t equate to having ‘low expectations’ for those with less than stellar exam results. There’s no virtue in people doing jobs they don’t enjoy and aren’t good at, and careers aren’t set in stone. An academic high flyer might become a superb potter, and a former train driver might get a PhD. If the education system doesn’t offer such opportunities, it’s to the detriment of all us.

Cummings would no doubt argue that his claims about education are evidence-based; he cites evidence for pedagogical approaches that improve exam results. But his starting point is an assumption that what the world needs is academic high flyers with high IQs and ‘extreme abilities’. He looks right past those with other abilities and aptitudes essential for communities to keep functioning. And those, who through no fault of their own, can make only a very limited contribution to their communities, but like all of us have a right to a decent quality of life.

Cummings first chooses his metric and then chooses supporting evidence, but only the evidence in support of it. Ironically history is littered with examples of academic high flyers with high IQs and ‘extreme abilities’ causing chaos for the rest of us. Cummings’ use of evidence is the subject of the next post.

reference

Spearman, C.  (1904).  ‘General Intelligence’ objectively determined and measured.  The American Journal of Psychology, 15, 201-292.

acknowledgements

Image from People’s Cyclopedia of Universal Knowledge (1883) via Wikipedia https://en.wikipedia.org/wiki/Phrenology

 

 

Dominic Cummings on education

Dominic Cummings has become a highly influential figure. He steered the UK’s education system towards a ‘knowledge curriculum’, persuaded many who voted in the 2016 referendum that they wanted the UK to leave the EU, and is now well on the way to ensuring that Brexit gets done – whatever that entails.

In 2013 Cummings published online an essay entitled Some Thoughts on Education and Political Priorities. His thoughts extend to nearly 250 pages.  I had a couple of goes at reading them at the time, but was fazed by the plethora of references to mathematicians and physicists. My rusty A level maths and even more rusty O level physics weren’t quite up to checking them out.  Following Cummings’ spectacular return to public life, I scrolled past them and found myself in more familiar territory.  This is the first of three posts, on Cummings’ views of education, intelligence, and expertise.

An Odyssean Education

Cummings isn’t happy with education systems. He complains that students aren’t taught about some fundamentally important ideas, so political leaders lack them too, which explains poor political decisions. He believes the ideas could go a long way to resolving the global crises facing us, so it’s imperative they’re taught in schools and universities. He’s particularly interested in the education of people with a high IQ.

Cummings refers to Neitzsche’s distinction between ‘Apollonian’ thinkers using logical analysis and ‘Dionysians’ who use intuition and synthesis. The physicist Murray Gell-Man suggested a third group – ‘Odysseans’ – who “combine the two predelictions”, look for connections between ideas, and take a “crude look at the whole” (p.5). As Cummings puts it “An Odyssean curriculum would give students and politicians some mathematical foundations and a map to navigate such subjects without requiring a deep specialist understanding of each element” (p.7).  He’s right about the map. Human knowledge has increased exponentially over the past century, so in-depth specialisms have become the order of the day. The best anyone could currently achieve is a ‘crude look at the whole’ but that crude look is essential if we are to understand the challenges confronting us.

Cummings structures his Odyssean curriculum as a “schema of seven big areas” (p.7) sketched out on page 2:

  1. Maths and complexity
  2. Energy and space
  3. Physics and computation
  4. Biological engineering
  5. Mind and machine
  6. The scientific method, education, training and decisions
  7. Political economy, philosophy, and avoiding catastrophes.

The essay includes 15 Endnotes on specific topics, and a reading list. In this post, I focus on education, addressed in Chapter 6.

Uniformity vs diversity

Cummings is critical of an education policy that aims for increased uniformity of achievement, based on the assumption that all students have the same potential, and would reach it if aspirations were raised and equal opportunities provided. Cummings’ model in contrast, assumes students don’t have the same potential because differences in ability are largely genetic in origin. He thinks more effective teaching will raise attainment levels for all, but will also widen the attainment gap (pp.74, 83). In my view, both models are wrong due to flaws in their implicit starting assumptions. Here’s why:

Human beings have been ‘successful’ in the evolutionary sense, in part because speech enables us to communicate complex information to each other. To survive and maintain good quality of life, everyone doesn’t need to know everything, but we each need access to the expertise of farmers, plumbers, electricians, doctors, lawyers, poets and dancers to name but few.

What enables populations to adapt to changing environments is genetic diversity. And genetic diversity produces people with the diverse abilities, aptitudes and interests that enable communities to adapt to changing circumstances. Communities thrive, not because of their uniformity, but because of their diversity. A good general education is important for everyone because we each need to know how the world works, but the last thing we need is for everyone to be the same.

The diversity does indeed mean that improving teaching would result in larger gaps in attainment – but only if you measure attainment on a linear scale such as exam results or IQ. Cummings is right that we desperately need people with high IQs who can do the maths required to model complex systems, and politicians who understand what’s being modelled. But our society couldn’t function if it consisted entirely of people who were a whiz at complex equations and/or political decision-making; we need people with a wide range of abilities, aptitudes and interests to make life sustainable and worth living.

Uniformity appeals to policy-makers because one-size-fits-all policies look like they’ll save money.  A diversity narrative is often used to make uniformity more palatable. But diversity in communities doesn’t only make life more interesting and colourful, it’s essential for our biological and economic survival and well being.

Aptitude

Genetic diversity provides communities with the wide range of abilities, aptitudes and interests they need to thrive. Ironically, the suitability of an education to aptitude (what someone is good at) has been embedded in English education law since at least 1944, but has received scant attention since the advent of the national curriculum and standardised testing.

Paying attention to aptitude doesn’t mean every student needs a personalised education programme, nor that schools should undertake vocational training. But developing the inherent qualitative variation in aptitude would mean the ensuing quantitative variation in exam scores became less important. Gaps in academic achievement matter only to societies that accord a disproportionately high status to professions requiring academic skills.

For example, doctors and lawyers are generally well paid and have high social status. The pay and social status of train drivers and electricians is generally lower. But train drivers and electricians are no less essential to a functioning community. Cummings lauds scientists, and is pretty dismissive of doctors and lawyers, but the people who maintain the complex infrastructure of the developed world don’t feature at all in his model of education, other than often being on the wrong side of the IQ bell curve.

Cummings’ proposals

To fix the problems with the education system, Cummings proposes (pp.69-83):

  1. Largely eliminate failure with the basics in primary schools
  2. Largely eliminate failure with the basics in secondary schools
  3. A scientific approach to teaching practice
  4. Maths for most 16-18
  5. Specialist schools from which all schools (and Universities) can learn
  6. Web-based curricula, MOOCs, and HE/FE
  7. Computer Science and 3D printers: bits and atoms, learning and making
  8. Teacher hiring, firing and training
  9. Prizes
  10. Simplify the appallingly complicated funding system, make data transparent and give parents a real school choice.

Most of his criticisms of the education system are valid ones, but criticism is the easy bit – it’s more challenging to come up with alternatives. Cummings generates ideas like they’re going out of fashion, but almost invariably overlooks context; notably what caused the problems, and the implications of his ideas being implemented. Here are some examples:

Maths     For Cummings ‘the basics’ are English, Maths and Science, with Maths the sine qua non because it provides the ‘language of nature’ (p.63). His proposal that 16-18 year-olds continue to study ‘some sort of Maths course’ (p.75) was implemented in 2015 in the form of students being required to re-sit Maths and English GCSEs if they got lower than a C grade. As far as I’m aware the scheme wasn’t piloted, placed a huge burden on an FE sector already pared to the bone, and many students found their career plans stalled due to an arbitrary and unnecessary requirement.

Reading     The UK’s achievement in reading is contrasted with that of Finland (p.69), but overlooks the fact that Finnish orthography is highly transparent (almost 1-1 correspondence between graphemes and phonemes) whereas English orthography is highly opaque.

Specialist schools     Cummings has high hopes for specialist schools (pp.75-77) but doesn’t mention their introduction in the 1988 Education Reform Act, or that under New Labour most state secondaries became specialist schools. Evaluations showed the consequent small improvement in exam results was as likely due to the additional funding, rather than specialist status as such. There doesn’t appear to have been a subsequent surge in superb scientists or brilliant politicians.

Teacher hiring, firing, and training     For Cummings “real talent is rare, mediocrity ubiquitous” (p.81). He would recruit academic high flyers, pay them well, get “roughly averagely talented teachers” to use Direct Instruction scripts and allow head teachers to sack the ones who still didn’t make the grade. He doesn’t mention working conditions or why teacher retention is so low.

Cummings also claims “managing schools is much easier than being a brilliant maths teacher and requires only the import of competent (not brilliant) professional managers from outside the education world” (p.83).  The transferable management skills hypothesis has been widely tested since the 1980s and been found seriously wanting.

Lectures     We’re told “students remember little from traditional lectures” (p.72). That might because traditionally, lectures formed only the framework for the students’ learning. Traditionally, students were expected to do further reading. And the ‘proven’ Oxbridge tutorial system is not as Cummings claims, limited to Oxford and Cambridge (p.78). It’s been in use in every university I’ve been involved with from the 1970s to the present. Maybe I’ve just been lucky.

Funding     The education funding system certainly needs rationalising, but costs vary across geographical areas, so who decides what a “flat per pupil amount” with “as few tweaks as possible” (p.81) means?

Parent choice     The other things described above … could be done even if one disagrees with the idea of a decentralised system driven by parent choice and prefers the old hierarchical system run by MPs, Unions, and civil servants” (p.83). Cummings appears completely unaware that the ‘old hierarchical’ system was decentralized and run by local authorities, school governors (including parents) and head teachers. And would probably have stayed that way if it hadn’t been deliberately centralized relatively recently by the Thatcher and subsequent governments.

Data transparency     Few would want to “define success according to flawed league table systems based on flawed GCSEs” but if “private schools have defined success according to getting pupils into elite universities” (p.82) where does that leave the bulk of the population? We’re not all going to get into elite universities – if we did, they wouldn’t, by definition, be elite.

Scientific evidence     Cummings is right that an evidence-based approach to education is vital, but has a touching faith in randomized controlled trials (RCTs) (p.64). The medical community’s objections to RCTs was not, as Cummings claims, because their expertise would be challenged by data, but because individual patients don’t always share the features of a large population. The same is true for school pupils.

Cummings follows Feynman in accusing educational researchers of ‘Cargo Cult’ science – mimicking the surface features of scientific research but not applying its deep structure (p.70). Regrettably, deep structure is noticeable by its absence from the hotch-potch of findings about cognition, lectures, tutorials, testing, genetics and IQ that he proposes as an alternative.

Sub-system optimization

Cummings repeatedly does what systems theorists call subsystem optimization at the expense of system optimization. A bit of a tongue twister, but it’s a simple and common phenomenon. The components of systems, by definition, are linked to each other, so tweaking one part is likely to result in changes to another. And improving part of the system can sometimes have the effect of making things worse overall. If the components of a system are loosely coupled (weakly connected), the impact might be negligible. If they’re tightly coupled (strongly connected) the impact can be substantial.

Cummings should know this because he devotes an entire section to the features of complex systems (pp.17-21), but appears have filed complex systems under ‘mathematical modelling’ rather than ‘public policy’ in his mental directory. He doesn’t apply systems theory to his own proposals, even though he recognizes many poor political decisions are made because politicians don’t understand how complex systems work.  A similar criticism can be applied to his thoughts on genetics and IQ, the subject of the next post.