biologically primary and secondary knowledge?

David Geary is an evolutionary psychologist who developed the concept of biologically primary and biologically secondary knowledge, popular with some teachers. I’ve previously critiqued Geary’s ideas as he set them out in a chapter entitled Educating the Evolved Mind. One teacher responded by suggesting I read Geary’s The Origin of Mind because it explained his ideas in more detail. So I did.

Geary’s theory

If I’ve understood correctly, Geary’s argument goes like this:

The human body and brain have evolved over time in response to environmental pressures ranging from climate and diet through to social interaction. For Geary, social interaction is a key driver of evolved brain structures because social interactions can increase the resources available to individuals.

Environmental pressures have resulted in the evolution of brain ‘modules’ specialising in processing certain types of information, such as language or facial features. Information is processed by the modules rapidly, automatically and implicitly, resulting in heuristics (rules of thumb) characteristic of the ‘folk’ psychology, biology and physics that form the default patterns for the way we think. But we are also capable of flexible thought that overrides those default patterns. The flexibility is due to the highly plastic frontal areas of our brain responsible for intelligence. Geary refers to the thinking using the evolved modules as biologically primary, and that involving the plastic frontal areas as biologically secondary.

Chapters 2 & 3 of The Origin of Mind offer a clear, coherent account of Darwinian and hominid evolution respectively. They’d make a great resource for teachers. But when Geary moves on to cognition his model begins to get a little shaky – because it rests on several assumptions.

Theories about evolution of the brain are inevitably speculative because brain tissue decomposes and the fossil record is incomplete. Theories about brain function also involve speculation because our knowledge about how brains work is incomplete. There’s broad agreement on the general principles, but some hypotheses have generated what Geary calls ‘hot debate’. Despite acknowledging the debates, Geary’s model is built on assumptions about which side of the debate is correct. The assumptions involve the modularity of the brain, folk systems, intelligence, and motivation-to-control.

modularity

The general principle of modularity – that there are specific areas of the brain dedicated to processing specific types of information – is not in question. What is less clear is how specialised the modules are. For example, the fusiform face area (FFA) specialises in processing information about faces. But not just faces. It has also been shown to process information about cars, birds, butterflies, chess pieces, Digimon, and novel items called greebles. This raises the question of whether the FFA evolved to process information about faces as such (the Face Specific Hypothesis), or to process information about objects requiring fine-grained discrimination (the Expertise Hypothesis). Geary comes down on the Faces side of the debate on the grounds that the FFA does not “generally respond to other types of objects … that do not have facelike features, except in individuals with inherent sociocognitive deficits, such as autism” (p.141). Geary is entitled to his view, but that’s not the only hotly debated interpretation of the evidence.

folk systems

The general principle of ‘folk’ systems – evolved forms of thought that result from information being processed rapidly, automatically and implicitly – is also not in question. Geary admits it’s unclear whether the research is “best understood in terms of inherent modular constraints, or as the result of general learning mechanisms” but comes down on the side of children’s thinking being the result of “inherent modular systems”.  I couldn’t find a reference to Eleanor Rosch’s prototype theory developed in the 1970s, which explains folk categories in terms of general learning mechanisms. And it’s regrettable that Rakison & Oakes’ 2008 review of research into how children form categories (that also lends weight to the general learning mechanisms hypothesis) wasn’t published until three years after The Origin of Mind. I don’t know whether either would have prompted Geary to amend his theory.

intelligence

In 1904 Charles Spearman published a review of attempts to measure intellectual ability. He concluded that the correlations between various specific abilities indicated “that there really exists a something that we may provisionally term “General Sensory Discrimination” and similarly a “General Intelligence”” (Spearman p.272).

It’s worth looking at what the specific abilities included. Spearman ranks (p. 276) in order of their correlation with ‘General Intelligence’, performance in: Classics, Common Sense, Pitch Discrimination, French, Cleverness, English, Mathematics, Pitch Discrimination among the uncultured, Music, Light Discrimination and Weight Discrimination.

So, measures of school performance turned out to be good predictors of school performance. The measures of school performance correlated strongly with ‘General Intelligence’ – a construct derived from… the measures of school performance. This tautology wasn’t lost on other psychologists and Spearman’s conclusions received considerable criticism. As Edwin Boring pointed out in 1923, ‘intelligence’ is defined by the content of ‘intelligence’ tests. The correlations between specific abilities and the predictive power of intelligence tests are well-established. What’s contentious is whether they indicate the existence of an underlying ‘general mental ability’.

Geary says the idea that children’s intellectual functioning can be improved is ‘hotly debated’ (p.295). But he appears to look right past the even hotter debate that’s raged since Spearman’s work was published, about whether the construct general intellectual ability (g) actually represents ‘a something’ that ‘really exists’. Geary assumes it does, and also accepts Cattell’s later constructs crystallised and fluid intelligence without question.

Clearly some people are more ‘intelligent’ than others, so the idea of g initially appears valid. But ‘intelligence’ is, ironically, a ‘folk’ construct. It’s a label we apply to a set of loosely defined characteristics – a useful shorthand descriptive term. It doesn’t follow that ‘intelligence’ is a biologically determined ‘something’ that ‘really exists’.

motivation-to-control

The motivation to control relationships, events and resources is a key part of Geary’s theory. He argues that motivation-to-control is an evolved disposition (inherent in the way people think) that manifests itself most clearly in the behaviour of despots – who seek to maximise their control of resources. Curiously, in referring to despots, Geary cites a paper by Herb Simon (Simon, 1990) on altruism (a notoriously knotty problem for evolution researchers). Geary describes an equally successful alternative strategy to despotism, not as altruism but as “adherence to [social] laws and mores”, even though the evidence suggests altruism is an evolved disposition, not merely a behaviour.

Altruism calls into question the control part of the motivation-to-control hypothesis. Many people have a tendency to behave in ways that increase their control of resources, but many tend to collaborate and co-operate instead, strategies that increase individual access to resources, despite reducing individual control over them. The altruism debate is another that’s been going on for decades, but you wouldn’t know that to read Geary.

Then there’s the motivation part. Like ‘intelligence’, ‘motivation’ is a label for a loosely defined bunch of factors that provide incentives for behaviour. ‘Motivation’ is a useful label. But again it doesn’t follow that ‘motivation’ is ‘a something’ that ‘really exists’. The biological mechanisms involved in the motivation to eat or drink are unlikely to be the same as those involved in wanting to marry the boss’s daughter or improve on our personal best for the half-marathon. The first two examples are likely to increase our access to resources; whether they increase our control over them will depend on the circumstances. Geary doesn’t explain the biological mechanism involved.

biologically primary and secondary knowledge

In The Origin of Mind, Geary touches on the idea of biologically primary and secondary competencies and abilities but doesn’t go into detail about their implications for education. Instead, he illustrates the principle by referring to the controlled problem solving used by Charles Darwin and Alfred Wallace in tackling the problem of how different species had arisen.

Geary says that such problem solving requires the inhibition of ‘heuristic-based folk systems’ (p.197), and repeatedly proposes (pp.188, 311, 331, 332) that the prior knowledge of scientific pioneers such as Linnaeus, Darwin and Wallace “arose from evolved folk biological systems…as elaborated by associated academic learning” (p.188). He cites as evidence the assumptions resulting from religious belief made by anatomist and palaeontologist Richard Owen (p.187), and Wallace’s reference to an ‘Overruling Intelligence’ being behind natural selection (p.83). But this proposal is problematic, for three reasons:

The first problem is that some ‘evolved’ folk knowledge is explicit, not implicit. Belief in a deity is undoubtedly folk knowledge; societies all over the world have come up with variations on the concept. But the folk knowledge about religious beliefs is usually culturally transmitted to children, rather than generated by them spontaneously.

Another difficulty is that thinkers such as Linnaeus, Darwin and Wallace had a tendency to be born into scholarly families, so their starting point, even as young children, would not have been merely ‘folk biological systems’. And each of the above had the advantage of previous researchers having already reduced the problem space.

A third challenge is that heuristics aren’t exclusively biologically primary; they can be learned, as Geary points out, via biologically secondary knowledge (p.185).

So if biologically primary knowledge sometimes involves explicit instruction, and biologically secondary knowledge can result in the development of fast, automatic, implicit heuristics, how can we tell which type of knowledge is which?

use of evidence

Geary accepts contentious constructs such as motivation, intelligence and personality (p.319) without question. And he appears to have a rather unique take on concepts such as bounded rationality (p.172), satisficing (p.173) and schemata (p.186).

In addition, Geary’s evidence is not always contentious; sometimes it’s his conclusions that are tenuous. For example, he predicts that if social competition were a driving force during evolution, “a burning desire to master algebra or Newtonian physics will not be universal or even common. Surveys of the attitudes and preferences of American schoolchildren support this prediction and indicate that they value achievement in sports … much more than achievement in any academic area” (pp.334-5), citing a 1993 paper by Eccles et al. The ‘surveys’ were two studies, the ‘American schoolchildren’ 865 elementary school students, the ‘attitudes and preferences’ competence beliefs and task values, and the ‘academic areas’ math, reading and music. Responses show some statistically significant differences. Geary appears to overegg the evidential pudding somewhat, and to completely look past the possibility that there might be culturally transmitted factors involved.

conclusion

I find Geary’s model perplexing. Most of the key links in it – brain evolution, brain modularity, the heuristics and biases that result in ‘folk’ thinking, motivation and intelligence – involve highly contentious hypotheses.  Geary mentions the ‘hot debates’ but doesn’t go into detail. He simply comes down on one side of the debate and builds his model on the assumption that that side is correct.

He appears to have developed an overarching model of cognition and learning and squeezed the evidence into it, rather than building the model according to the evidence. The problem with the second approach of course, is that if the evidence is inconclusive, you can’t develop an overarching model of cognition and learning without it being highly speculative.

What also perplexes me about Geary’s model is its purpose. Teachers have been aware of the difference between implicit and explicit learning (even if they didn’t call it that) for centuries. It’s useful for them to know about brain evolution and modularity and the heuristics and biases that result in ‘folk’ thinking etc. But teachers can usually spot whether children are learning something apparently effortlessly (implicitly) or whether they need step-by-step (explicit) instruction. That’s essentially why teachers exist. Why do they need yet another speculative educational model?

references

Eccles, J., Wigfield, A., Harold, R.D.,  & Blumenfeld, P. (1993). Age and gender differences in children’s self‐and task perceptions during elementary school, Child Development, 64, 830-847.

Gauthier, I., Tarr, M.J., Anderson, A.W., Skudlarski, P. & Gore, J.C.  (1999). Activation of the middle fusiform ‘face area’ increases with expertise in recognizing novel objects, Nature Neuroscience, 2, 568-573.

Rakison, D.H.  & Oakes L.M. (eds) (2008). Early Category and Concept Development.  Oxford University Press.

Simon, H.A. (1990). A mechanism for social selection and successful altruism. Science, 250, 1665-1668.

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

 

 

Advertisements

Toby and the social constructionists: intelligence, race & gender

Toby Young, journalist and co-founder of the West London Free School, and appointed to the brand new Office for Students (OfS), recently delivered the 2017 Constance Holden* Memorial Address. His title was “Liberal Creationism”. The text of the lecture has been published as an opinion paper in the Elsevier Journal Intelligence. Neither Toby nor the topic of intelligence are strangers to controversy, so I was curious about to what he had to say.

the fate of intelligence researchers

For the first couple of pages I was nodding along in agreement. Toby opens with a robust defence of intelligence researchers who have faced serious consequences for suggesting that biological factors contribute to intelligence – careers ending abruptly, ignominy, and worse. His description of their opponents as a ‘neo-Marxist intersectionality cult’, ‘Social Justice Warriors’, ‘Liberal Creationists’ and ‘anti-hereditarians’ had a ring of truth about it.

The intelligence researchers’ findings have generally supported the idea that human nature is in part biologically determined.  Historically this idea is tainted by its association with eugenics, and it’s that association, Toby suggests, that’s at the root of the opposition to the research. He claims the ‘anti-hereditarians’ are wrong to think of the biological aspects of intelligence as “inextricably bound up with these toxic political movements and fundamentally incompatible with liberal values” (p.3).  I think he’s right.

human nature and individual differences

After a brief tour of some anthropological controversies to demonstrate that all human beings have some psychological traits in common, Young moves on to “what divides us” beginning with individual differences. This is the point where I felt his argument weakened considerably.

Young says “when a progressive liberal listens to a behavioural geneticist talk about the biological basis of IQ and the positive correlation between IQ and socio-economic status, what they think they are hearing is a Social Darwinist argument in favour of the current distribution of wealth and power” (p.3). He argues that at the heart of the opposition to claims about the biological basis of human nature, whether IQ, gender or race, is social constructionism, which holds the view that “anyone who believes that human differences are rooted even in part in biology rather than socially constructed is the enemy” (p.2). What Young appears to have failed to realize is that whatever the contribution of biology and however wrong the social constructionists’ reasoning, ‘intelligence’ (IQ), ‘race’ and ‘gender’ are nonetheless constructs.

intelligence, race and gender as constructs

There’s no doubt that biological factors contribute to what we call intelligence, race and gender. Some congenital medical conditions result in low intelligence, people with ancestry in different parts of the world have different physical and physiological characteristics, women are anatomically different to men, and so on.

The concepts ‘intelligence’, ‘race’ and ‘gender’ entail biological characteristics (ability to carry our particular tasks, physical features, anatomical differences), but that doesn’t mean ‘intelligence’, ‘race’ or ‘gender’ map directly on to discrete biological entities.  Although the biological characteristics are real things in the natural world, it doesn’t follow that the concepts ‘intelligence’, ‘race’ and ‘gender’ must be real things in the natural world. As Gilbert Ryle would have pointed out, intelligence, race and gender are different kinds of things to the brains, skin colour and sexual characteristics we associate with them.

The biological characteristics mean we can operationalize those constructs in biological terms for research purposes. A researcher could decide that, for the purposes of their study, intelligence consists of the ability to perform particular tasks, measured using a particular test. A particular racial grouping could be operationalized in terms of particular physical characteristics. Gender, likewise.

It doesn’t follow that intelligence, race or gender have a biological existence independent of the concept of intelligence, race or gender. In marked contrast to brains, skin or genitals.

I think Toby is right that some more vociferous opponents to the claims about intelligence have framed the debate entirely in terms of social constructs and in doing so have completely discounted biological factors. But that’s not true of all who have questioned the claims about intelligence, race or gender.

The existence of brains, skin or genitals are rarely disputed.  They are physical entities and there’s general agreement about their characteristics. But intelligence, race and gender have been the subject of controversy more or less since they came into being because they are constructs. The question looming over all of them is ‘what do you mean by….?’

what do you mean by…?

As the psychologist Edwin Boring pointed out way back in the 1920s, intelligence is what[ever] intelligence tests measure.   Somebody, somewhere makes a decision about what a particular intelligence test measures. Intelligence tests undoubtedly measure something. Whether it’s intelligence or not depends on what you think intelligence is.

Race is controversial because the characteristics that are supposed to be typical of different racial groups are, because of the way genetic expression works, on a multi-dimensional continuum.  Even if the features considered to characterise a particular racial group are very clearly defined, it’s often difficult to decide whether or not an individual belongs to that group.

As for gender… The ‘social constructionists’ so disparage by Young make an important distinction between someone’s sex (reproductive anatomy and secondary sexual characteristics) and their gender – their social role based on their sex. The distinction is a helpful one, but two terms are often, unhelpfully, used interchangeably.

A further complication is that biology, also unhelpfully, doesn’t adhere to a neat binary male/female distribution of physical sexual characteristics. One estimate§ puts the frequency of intersex characteristics as high as 1.7% of the population.

The difficulty in determining what constitutes intelligence, race or gender calls into question the validity of all correlations found between intelligence and any demographic group. That’s because someone has to decide what constitutes intelligence and what constitutes the demographic group; the validity of intelligence as a construct has been questioned ever since Spearman came up with the idea of g.

you say this, I say that

Young’s argument appears to be essentially this: Science shows that shared human characteristics and individual differences are to large extent biologically determined.  And that anti-hereditarian opposition to that finding originates in the ‘fanatical egalitarianism’ of the hard left (p.6). According to Young, the egalitarianism the hard left want can be brought about by a totalitarian dictatorship.

Interestingly, in his opening comments Young dismisses the fanatical egalitarianism embraced by the far-right (also enforced via a totalitarian dictatorship), as ‘toxic baggage’ that could be discarded. And complains “It’s not much fun to be branded a ‘Nazi’ or ‘white supremacist’ on Twitter or anywhere else” (p.1).

Totalitarian regimes of the hard left, far-right and various other political and religious persuasions, inflicted catastrophic damage on huge numbers of people during the 20th century and look set to continue doing so in the 21st. To justify their actions, some regimes used the idea that characteristics are inherited. Others have used ideology. None have taken much notice of what science has to say.

Toby has taken note of the science but has managed to fundamentally misunderstand both it and social constructionism. He’s lumped together anyone else who’s misunderstood the science, objected to spurious scientific claims, been upset by unwise off-the-cuff remarks, with left-wing ideologues.   And has used his misunderstanding to launch an attack on the hard left who he equates with Stalinists.

His argument is essentially constructed in opposition to the narrative of people he disagrees with, whether they are a ‘neo-Marxist intersectionality cult’, ‘Social Justice Warriors’, ‘Liberal Creationists’, ‘anti-hereditarians’ or vote Democrat (p.7).

If the social constructionists think intelligence, race and gender are social constructs, Young doesn’t consider the possibility they might actually be no more than social constructs, despite scientists arguing over their construct validity for decades. If sceptics get upset about scientists making questionable claims, Young assumes the scientists must be right. If any group he disagrees with marginalises the role of biological factors in individual differences, Young marginalises the role of environmental factors. If all the parties Young disagrees with fight like ferrets in a sack about the differences between their views, Young lumps them all together as if they form a homogeneous group.

That doesn’t tell us that Young is wrong and the social constructionists/neo-Marxist intersectionalists/Social Justice Warriors/Liberal Creationists/anti-hereditarians are right. What it does tell us is that misunderstandings of science can be used to justify anything.

notes

*Constance Holden wrote for the news section of Science for 40 years.

§Blackless et al (2000). How sexually dimorphic are we?  http://www.aissg.org/PDFs/Blackless-How-Dimorphic-2000.pdf