IQ is rising in many parts of the world. What's behind the change and does it really mean people are cleverer than their grandparents?
It is not unusual for parents to comment that their children are brainier than they are. In doing so, they hide a boastful remark about their offspring behind a self-deprecating one about themselves. But a new study, published in the journal Intelligence, provides fresh evidence that in many cases this may actually be true.
The researchers - Peera Wongupparaj, Veena Kumari and Robin Morris at Kings College London - did not themselves ask anyone to sit an IQ test, but they analysed data from 405 previous studies. Altogether, they harvested IQ test data from more than 200,000 participants, captured over 64 years and from 48 countries.
Focusing on one part of the IQ test, the Raven's Progressive Matrices, they found that on average intelligence has risen the equivalent of 20 IQ points since 1950. IQ tests are designed to ensure that the average result is always 100, so this is a significant jump.
The gains have not been evenly spread. IQ has generally increased more rapidly in developing countries, with the biggest leaps seen in China and India. Progress in the developed world has been chequered - the data seem to indicate steady increases in the US, for example, but a decline in the UK.
The new research is further confirmation of a trend that scientists have been aware of for some time. In 1982, James Flynn, a philosopher and psychologist based at the University of Otago in New Zealand, was looking through old American test manuals for IQ tests. He noticed that when tests were revised every 25 years or so, the test-setters would get a panel to sit both the old test and the new one.
"And I noticed in all the test manuals, in every instance, those who took the old test got a higher score than they did on the new test," says Flynn. In other words, the tests were becoming harder.
This became known as the Flynn Effect, though Flynn stresses he was not the first to notice the pattern, and did not come up with the name.
But if the tests were getting harder, and the average score was steady at 100, people must have been getting better at them. It would seem they were getting more intelligent.
If Americans today took the tests from a century ago, Flynn says, they would have an extraordinarily high average IQ of 130. And if the Americans of 100 years ago took today's tests, they would have an average IQ of 70 - the recognised cut-off for people with intellectual disabilities. To put it another way, IQ has been rising at roughly three points per decade.
This is a puzzle not just for the US, but for all countries demonstrating the Flynn Effect. "Does it make sense," Flynn wrote in one paper, "to assume that at one time almost 40% of Dutch men lacked the capacity to understand soccer, their most favoured national sport?"
So what is going on? "There are lots of theories, none of which is particularly proven," says Robin Morris.
One possible explanation has to do with changes in education.
In most of the developed world, more people are now in school for longer, and teaching methods have evolved, moving away from the simple memorising of names, dates and facts. It seems like a reasonable assumption that education is training people to think better.
But in fact, the evidence is mixed. There has been no clear correlation between the rising IQ scores and US school performance - in SAT tests, for example.
But school prepares children for sitting IQ tests in other ways - what the psychologist Arthur Jensen has called "test wiseness". Over time, students become used to the pressure of tests and they pick up examination-room tactics that improve their performance.
A vivid demonstration of this emerges from a study of raw IQ data from Estonia. When psychologists Olev and Aasa Must laid examination papers from the Estonian National Intelligence Test from the 1930s alongside papers from 2006, they found an increase in correct answers - and also incorrect ones. The more recent students knew that they would not be penalised for guessing and getting something wrong.
James Flynn believes test wiseness may have been a factor in IQ gains in the US in the first half of the 20th Century. However, since then the amount of IQ testing taking place has waned - and IQ increases have remained steady.
Flynn puts this continued progress down to profound shifts in society as well as education over the last century, which have led people to think in a more abstract, scientific way - the kind of intelligence measured by IQ tests.
He cites the work of Russian neuroscientist Alexander Luria, who studied indigenous people in the Soviet Union. "He found that they were very pragmatic and concrete in their thinking," says Flynn, "and they weren't capable of using logical abstractions or taking hypotheticals seriously." Luria put the following problem to the head man of one tribe in Siberia: Where there's always snow, the bears are white; there's always snow at the North Pole - what colour are the bears there?
The head man replied that he had never seen bears that were any colour other than brown, but if a wise or truthful man came from the North Pole and told him that bears there were white, he might believe him. The scientific methods of hypothesising, classifying and making logical deductions were alien to him.
"Now virtually all formal schooling, when you get past the sixth grade into high school and college, means that you take hypotheses seriously," says Flynn. "This is what science is all about. And you're using logic on abstract categories."
And this kind of thinking doesn't only occur in school.
As Flynn pointed out in his Ted Talk on the Flynn Effect, in 1900 only 3% of Americans performed "cognitively demanding" jobs - now the figure is 35%, and the work itself is far more intellectually demanding than it was a century ago. Families are also smaller, so children are exposed to more adult conversation at the dinner table than in the past. "Hothouse parenting" - pushing your kids to achieve goals from an early age - may also be a factor. And when it comes to older people, a lower disease burden may have an effect on their performance in tests.
Such effects have diminishing returns after countries become fully industrialised, Flynn says, which may explain why in some North European countries, including France and Scandinavia, IQs have flatlined or diminished slightly. He admits that the pattern in Europe is a little baffling, but he has an idea why IQ scores continue to rise in the US. "I think America is a society where economic and environmental differences are much greater than they are in Scandinavia. And for example black Americans have terrible schools, and they have had terrible conditions to live under."
A few other possible causes for the Flynn Effect have been put forward, some of them very intriguing.
One, proposed by Arthur Jensen but yet to be investigated, points to the spread of electric lighting. The thought is that light from bulbs, TV screens and the like may have contributed to cognitive development in a similar way that artificial light stimulates growth in chickens.
Then there is the theory that today's world is more visual than the world of 100 years ago. The Raven's Progressive Matrices - the subject of the recent international study into the Flynn Effect by Wongupparaj, Kumari and Morris - requires people to pick out patterns from an array of stripes and squiggles. This particular test has seen the biggest IQ increases of all. Perhaps television, video games, advertisements and the proliferation of symbols in the workplace have made it easier for us to decode pictorial cues and identify patterns?
There is also a debate surrounding nutrition. In a 2008 article in Intelligence, Richard Lynn notes that measures of infants' mental development increased in the UK and US at rates correlated to the increasing IQs of slightly older children. It's difficult to see how Flynn's theories are enough to explain this. "Are infants thinking more scientifically today?" he asks rhetorically.
Lynn argues that pre-natal nutrition is a determinant of birth weight, which is in turn correlated to higher IQs. A shortage of one particular nutrient - iodine - is known to stunt intellectual development in growing children. A 2005 paper examining iodine deficiency in China found that children's IQ scores were higher in areas where there was no iodine deficiency, and it increased after a programme of supplements started.
So explanations of the Flynn Effect abound - but what precisely does it signify? Do these steadily improving results indicate that the IQ test is not, after all, measuring intelligence? Or are people really cleverer than their forefathers?
"I don't think smarter has anything to do with it," says Flynn.
"Today we have a wider range of cognitive problems we can solve than people in 1900. That's only because society asks us to solve a wider range of cognitive problems. People in 1900 had minds that were perfectly adequate for remembering first cousins once removed, they were perfectly adequate for ploughing a farm, they were perfectly adequate for making change in a store. No-one asked them to do tertiary education.
"It's like a weightlifter and swimmer. They may have the same muscles when they were fertilised in the womb, but they would have different muscles at autopsy, wouldn't they? So today at autopsy, certain portions of our brain, for example those which use logic and abstraction, would have been exercised more and look differently. Other portions of the brain would have shrivelled a bit."
It may be, then, that certain abilities - problem-solving or reasoning ability, say - have improved but a general, underlying cognitive ability has not changed. This general ability is fundamental to the way many scientists view intelligence. Although little is actually known about it, there is supposed to be a general, hereditary quality that makes an individual who is good at giving fine speeches more likely to be good at Sudoku too. The problem is that this general cognitive ability is exactly what IQ tests are supposed to measure - in fact, of all the components of the IQ test, the Ravens test was supposed to be the truest measure of it. If people aren't becoming fundamentally more intelligent, IQ tests aren't doing what they're supposed to do.
But Robin Morris is prepared to entertain the possibility that there may, over time, have been a real increase in general cognitive ability.
"It seems to me that it's reasonable to think that intellectual functioning could increase over time in more developed societies," says Robin Morris.
But do we actually notice in our midst a higher proportion of geniuses than there were in earlier generations?
"That's the baffling aspect," Morris admits. "How could it go up so much but there aren't all these very very smart people floating around? And that is a bit of a puzzle. But then, people have started to say, 'Maybe there are more bright people floating around and they're kind of hidden away because of the way science has become very specialist. They're working in their own particular field and they're doing amazing things - they're acting as geniuses - but they're not necessarily identified as such.'"
It's an odd thought. There are more and more geniuses out there, if this theory is correct - but many of them are unrecognised.
Subscribe to the BBC News Magazine's email newsletter to get articles sent to your inbox.