At Our Wits' End, page 17
68 Lynn, R. (2011) Dysgenics, London: Ulster Institute for Social Research.
69 Nyborg, H. (2011) The decay of Western civilization: Double relaxed Darwinian selection, Personality & Individual Differences, 53, pp. 118–125.
70 Lynn, R. & Vanhanen, T. (2012) Intelligence: A Unifying Construct for the Social Sciences, London: Ulster Institute for Social Research.
Eight
But Aren’t We Getting Smarter?
Originally from Washington DC, James Flynn emigrated to New Zealand in 1963, when he was 29 years old. He began his academic career as a political scientist, his first major paper, in 1964, being on the peace movement in the USA. This was followed by such studies as American Politics: A Radical View in 1967, Humanism and Ideology: An Aristotlean View in 1973, and assorted highly philosophical works including the article ‘Kant and the Price of Justification’ in 1979.[1]
In 1980, however, Flynn’s research profile began to change. He became fascinated by intelligence and IQ. As far as he could see, comparing old IQ tests administered to representative cohorts to modern ones, IQ scores had actually been increasing over time. He wrote to as many psychologists as he could in an attempt to obtain as much data as he could and the result was a ground-breaking paper published in 1984 in the journal Psychological Bulletin: ‘The Mean IQ of Americans: Massive Gains, 1932 to 1978.’[2] This study, and subsequent studies confirming it, was so influential that Flynn’s discovery became known as the ‘Flynn Effect’. In fact, as early as 1937, psychologists were commenting on evidence of rising IQ scores in the USA[3] and Richard Lynn had also rediscovered the effect in a 1982 paper in the journal Nature, which has led to some people terming it the ‘Lynn-Flynn Effect’.[4] But, irrespective of whoever really was the first to uncover it, the effect has now generally become known as the ‘Flynn Effect’. The Flynn Effect is the phenomenon whereby average IQ scores have increased throughout the 20th century. Later cohorts have, on average, higher scores than earlier cohorts.
When IQ tests are taken by each new cohort the average score is set at 100. The Flynn Effect has meant that with each new cohort the new average score is significantly above 100, which was, by definition, the average score of the previous cohort. As such, it is clear that the average IQ score must have been increasing over time. Between the 1930s and the 1970s, at least, we appeared to have been getting more intelligent or, at the very least, better at IQ tests. The extent of these gains in IQ was really quite staggering. IQ scores were found to be going up by between 3 and 5 IQ points per decade or about a third of an IQ point per year. If this was really a reflection of intelligence increasing then the average person today would have to be considered some kind of genius in the 1930s. Flynn Effects have been reported across Western Europe, as well as in other Western or highly developed countries such as the US, Canada, Australia, New Zealand, South Korea, and Japan.
More recently, Flynn Effects have been documented in developing countries as well. The Kenyan IQ—on the Coloured Progressive Matrices—increased by almost 14 points between 1984 and 1998. Between 1977 and 2010, the IQ of the Saudis went up by 11 points. The same phenomenon has been observed in Sudan, Turkey, Dominica, Brazil, India, Israel, Argentina, white South Africa, and China. It has also been observed in ex-Soviet countries, such as Estonia from 1935 to 1978. Over the course of the 20th century IQ scores increased in highly developed countries and the same process then occurred in less developed countries by the end of the 20th century.[5]
What Is Going On?
Superficially, the Flynn Effect makes absolutely no sense. Surely, if intelligence is strongly heritable and intelligence has been negatively associated with fertility for a long time, then IQ scores should be decreasing. How on earth can they be increasing and increasing so dramatically?
A closer look at the process allows us to make sense of it. The most important point is that the Flynn Effect has not been found to be occurring equally on all of the different parts of the IQ test. So, a representative Dutch male sample made gains of 0.6 IQ points per decade between 1952 and 1982. They made these gains on the Raven’s test. This test measures fluid intelligence; that is solving problems purely via abstract reasoning without any prior knowledge. The Wechsler test is far broader. It measures both fluid intelligence and crystallised intelligence, where you do have to draw upon prior knowledge. The only Wechsler sub-test which showed similar gains to Ravens was the ‘similarities’ test. This is the test where you have to classify things. For example, you might be given a list of animals and have to select the one that is not a mammal. So, in fact, IQ scores were not rising in general but only on very specific kinds of ability.[6]
This was backed up by a number of studies which found that the Flynn Effect is not occurring on ‘general intelligence’.[7] We have already discussed the concept of general intelligence or g. As we have seen, people who perform well on one kind of IQ test—for example, one that measures mathematical intelligence—also do well on all the others. As such, we conclude that there is a general factor underpinning ability in all of these different intelligence tests and we call this ‘general intelligence’. However, analyses of the Flynn Effect in many countries show that it is not occurring on the highly g-loaded sub-tests. In fact, the Flynn Effect is occurring most strongly on the least g-loaded parts of the test. In other words, the Flynn Effect is not a rise in general intelligence. This is a point which Jim Flynn has emphasised in his book on the Flynn Effect, Are We Getting Smarter? The Flynn Effect involves a rise in very specific abilities that are weakly associated with general intelligence. These seem to be specialised abstract reasoning skills that relate to the ability to use classification, in other words thinking in a scientific and analytical way.[8]
But, if that is the case, how can this possibly lead to massive IQ gains? The answer is quite simple. If people became dramatically better at an ability that is only weakly associated with intelligence then this would be reflected in their overall IQ score, causing their IQ score to increase. Therefore, if they were sufficiently strong in this ability (or a small set of abilities) then it could be more than enough to lead to them achieving a very high IQ score despite the fact that there had been no increase in their general intelligence. Indeed, their general intelligence might have decreased, but the massive increase in specific abilities could be enough to not only hide this decrease but show, overall, a huge increase. This would be entirely in line with the fact that the increases are only on certain specific parts of the IQ test and that these are typically the least g-loaded parts of the test.[9]
A comparable example can be seen in how well people do in their school leaving exams. In general, these exams—in various different school subjects—can be understood as a test of intelligence and, overall, the more intelligent children will do better in them than the less intelligent children. But they are very far from a perfect measure of general intelligence. They also measure other factors, including Conscientiousness and simply the degree to which people are prepared for the exams. If their parents are wealthy and have hired them a private tutor, for example, this is likely to have at least some positive impact on their result. They will have practised more, learnt more, thought about the subject matter more, been taught techniques and skills to better answer the questions, and so on. This being the case, it is quite possible—if the degree of Conscientiousness or the extent of social advantage is great enough—that a less intelligent child could do better in their school exams than a more intelligent child. In much the same way, the imperfect nature of the IQ test means that a less intelligent cohort could perform better on the test than a more intelligent cohort, as long as the test was in part measuring something which the less intelligent cohort scored so much higher in that it compensated for their lower intelligence.
In fact, we can highlight a very specific limit to the usefulness of both the IQ test and school exams. They are both excellent instruments for measuring general intelligence differences within a cohort. But they are not good instruments at measuring it between cohorts. This is because they do not control for the possibility that the infrastructure of the society is changing in such a way as to ensure that people practise these tests more and more and think in a way that is relevant to these tests more and more. Tabloid newspapers often bemoan so-called ‘grade inflation’ in the UK. The percentage of students getting top grades in their school leaving exams seems to increase every year. But then every year students are getting more experienced in taking exams, more experienced in thinking in the ways that help them to do well in these particular exams, and have greater access to more and more information. Accordingly, it becomes very problematic to assert that a person that got an A in A-Level Maths in 2015 is better at Maths, let alone cleverer, than a person who got a C in it in 1960.[10] So, to measure general intelligence changes over time it is a bad idea to use IQ tests. More objective proxies, that are less sensitive to these sorts of cultural changes, would be better and these are what we will use.
Scientific Spectacles
Even if general intelligence is decreasing, there are a variety of ways in which the modern world is going to cover this up. Most obviously, better nutrition and a more intellectually stimulating environment—such as in relation to compulsory education and widespread literacy—are likely to mean that pretty much everybody achieves their maximum genotypically allowable intelligence, in a way that would not have been as easily possible before the Industrial Revolution, when there was widespread malnutrition and illiteracy. Also, increased living standards are likely to mean that people are decreasingly stressed and decreasingly sick and this should also have some positive effect on IQ test scores, via the effect of these factors on boosting brain development and promoting specialised abilities.
In addition to these factors, Flynn himself has proposed that the modern world makes us think in a different way. It causes us to use what Flynn has called ‘scientific spectacles’.[11] In other words, the Industrial Revolution has created an environment which encourages us to think in a scientific and analytical fashion. We do this from a very young age and, therefore, become increasingly more adept at it over time. Flynn gives the example of Russian peasants who were interviewed in the 1920s. With no education at all, they were simply unable to think in the kind of scientific, analytical way that we now take for granted.
‘Q: There are no camels in Germany. The city of B is in Germany. Are there camels there or not?
A: I don’t know. I’ve never seen German villages. If B is a large city there should be camels there.
Q: But what if there aren’t any in Germany at all.
A: If B is a village there is probably no room for camels.’
The Russian peasant who was interviewed lived in concrete reality, not a world of symbols. He almost certainly couldn’t read and we probably learn more new information in the average day than he might have done in ten years. He simply found it impossible to think in an abstract way. He thought in a concrete way.[12]
With the Industrial Revolution, we are forced to think in an abstract way. If we learn to do formal mathematics, we are entering a world of symbols, which compel us to think in an abstract way. The same is true if we learn to read and write or if we learn a foreign language. The Industrial Revolution increases levels of education. It leads to a more specialised society in which more people must be literate and numerate, compelling society to make ever more schooling mandatory. It leads to a wealthier and more stable society in which there is more money to educate people and more reason to invest in the future. And it leads to easy access to information through the cheaper production of books and newspapers and the development of television and computers. To keep up in the industrial world, with its ever-increasing mechanisation, we require some understanding of science, which also compels us to think in a more abstract way. We must understand how things work and why they work as they do. Accordingly, it can be argued that the more educated society becomes the more prone it will be to thinking in an abstract way. It will also become more literate, leading to a larger vocabulary. Words can be seen as ‘thinking tools’, meaning that a larger vocabulary will permit more technically subtle thinking. Western societies have become highly educated, with everybody in full-time education until around the age of eighteen. And the nature of this education has also become more ‘scientific’ over the years. History at school, for example, has moved away from the rote-learning of important facts to attempting to understand why a period of history unfolded as it did. This will be compounded by hobbies that reflect a greater ability to think abstractly and increase our ability to do so. This will be true even of reading low-brow novels or playing computer games. Access to the internet means that we are reading and stimulating our minds and doing so almost all of the time. One of the most popular internet hobbies, and one we mentioned earlier, is genealogy. Here you must play amateur sleuth: transcribe old hand-writing, learn Early Modern English, learn basic Latin, and weigh up the significance of different clues. Genealogy, as just one example of something popularised by the internet, clearly makes you think in a more abstract way. It is hard to get our heads around how different life would be for a farm labourer in Russia in 1920. We think in a more abstract way and abstract thinking is precisely what is tested by Raven’s and by the ‘Similarities’ component of IQ tests.
So, a narrow intelligence sub-ability—the ability to think abstractly —has been increasing.[13] Indeed, it has been increasing to such an extent that it shows up on the IQ tests as a year on year increase in IQ scores. Michael A. Woodley of Menie has argued not just that the Flynn Effect is occurring on the less g-loaded parts of IQ tests—on the less g-loaded abilities—but, also, that it is occurring on the less heritable abilities. Just as we have noted that general intelligence is significantly heritable, so are the different sub-abilities, but they are heritable to different degrees. The Flynn Effect is occurring mainly on the less heritable abilities, further demonstrating that it is mainly an environmental effect. Woodley of Menie argues that, at the same time—as we will see later—there is good evidence that general intelligence is actually declining. As such, he calls this the ‘Co-occurrence Model’—because an environmentally caused rise in specialised cognitive abilities and a genetically caused fall in general intelligence have ‘co-occurred’.[14]
There is, in fact, nothing especially unusual about these sorts of co-occurrences in nature. We have already looked at the parallels between IQ and height. Both are normally distributed on a bell curve, with the largest percentage in the centre of the curve and smaller and smaller percentages of the population positioned further and further away from it. However, there are other parallels between IQ and height. Like IQ, height has been rising in Western countries throughout the 20th century. The average height of a British adult male in 1900 was about 5 foot 6. By 1971, this average height had risen to 5 foot 10.[15] As with IQ, these height improvements have been concentrated on the more environmentally influenced components of height. Some body measurements, such as wrist circumference, are strongly genetic while others, such as neck circumference, are heavily environmentally influenced.[16] Similarly, there are different measurements that can make you tall. You could be tall, for example, due to having a very long neck. You could be tall, even if you had short legs, by having a very long torso. But the driving force behind the growth in height across the 20th century has in fact been leg length.[17] Our legs—but not a lot else—got longer suggesting that leg length may be highly environmentally influenced.
As with IQ, the secular improvements in height have also slowed down towards the end of the 20th century and in some places stopped, implying that we have reached our genotypic maximum height.[18] Studies from the US have shown that there are no genetic-selection effects on height, when genetic variants that predict height are used to predict fertility outcomes in lieu of phenotypic measures of height.[19] Hence, we would expect the more heritable components of height (such as neck and torso length) to have remained unchanged over time. This lack of genetic change will have been swamped by the contribution made to the change in height over generations by the increasing length of our legs, however. Thus with height we see another example of the co-occurrence model. The 20th century has seen a Flynn Effect on height: an increase in height driven exclusively by changes occurring on the more environmentally sensitive components of height. We are all used to being told that numerous human traits—intelligence, height, physical build, and the development of certain illnesses—are a product of a combination of genes and environment; nature and nurture. But with the co-occurrence model we can be far more specific than this. We can identify which aspects of the trait are primarily caused by environment and which aspects of the trait are primarily caused by genetics. This is a far less question-begging model which allows us to explain a lot more. If we want to understand the rise in obesity, for example, we can look at the different dimensions of obesity and work out how heritable each is and therefore how sensitive each may be to environmental change. We can therefore move beyond the rather trite assertion that ‘it’s a combination of environment and genes.’
The Reversal of the Flynn Effect
If this model is accurate we would expect two further phenomena. Firstly, proxy measures for general intelligence should show evidence that this is in decline and we will see later that this is indeed what is happening. Secondly, there should be a limit to and then a reversal of the Flynn Effect. This is because the specialised abilities that are being stimulated by changes in the environment will have a phenotypic limit. We will eventually reach the absolute maximum level of this ability that the environment can produce within the bounds of our genetic limitations. Once this happens the Flynn Effect will cease and the underlying decline in general intelligence—which has been hidden by the Flynn Effect—will start to show up even on the IQ tests.
