An online study open to anyone, that ended up involving over 100,000 people of all ages from around the world, put participants through 12 cognitive tests, as well as questioning them about their background and lifestyle habits. This, together with a small brain-scan data set, provided an immense data set to investigate the long-running issue: is there such a thing as ‘g’ — i.e. is intelligence accounted for by just a single general factor; is it supported by just one brain network? — or are there multiple systems involved?
Brain scans of 16 healthy young adults who underwent the 12 cognitive tests revealed two main brain networks, with all the tasks that needed to be actively maintained in working memory (e.g., Spatial Working Memory, Digit Span, Visuospatial Working Memory) loading heavily on one, and tasks in which information had to transformed according to logical rules (e.g., Deductive Reasoning, Grammatical Reasoning, Spatial Rotation, Color-Word Remapping) loading heavily on the other.
The first of these networks involved the insula/frontal operculum, the superior frontal sulcus, and the ventral part of the anterior cingulate cortex/pre-supplementary motor area. The second involved the inferior frontal sulcus, inferior parietal lobule, and the dorsal part of the ACC/pre-SMA.
Just a reminder of individual differences, however — when analyzed by individual, this pattern was observed in 13 of the 16 participants (who are not a very heterogeneous bunch — I strongly suspect they are college students).
Still, it seems reasonable to conclude, as the researchers do, that at least two functional networks are involved in ‘intelligence’, with all 12 cognitive tasks using both networks but to highly variable extents.
Behavioral data from some 60,000 participants in the internet study who completed all tasks and questionnaires revealed that there was no positive correlation between performance on the working memory tasks and the reasoning tasks. In other words, these two factors are largely independent.
Analysis of this data revealed three, rather than two, broad components to overall cognitive performance: working memory; reasoning; and verbal processing. Re-analysis of the imaging data in search of the substrate underlying this verbal component revealed that the left inferior frontal gyrus and temporal lobes were significantly more active on tasks that loaded on the verbal component.
These three components could also be distinguished when looking at other factors. For example, while age was the most significant predictor of cognitive performance, its effect on the verbal component was much later and milder than it was for the other two components. Level of education was more important for the verbal component than the other two, while the playing of computer games had an effect on working memory and reasoning but not verbal. Chronic anxiety affected working memory but not reasoning or verbal. Smoking affected working memory more than the others. Unsurprisingly, geographical location affected verbal more than the other two components.
A further test, involving 35 healthy young adults, compared performance on the 12 tasks and score on the Cattell Culture Fair test (a classic pen and paper IQ test). The working memory component correlated most with the Cattell score, followed by the reasoning component, with the Verbal component (unsurprisingly, given that this is designed to be a ‘culture-fair’ test) showing the smallest correlation.
All of this is to say that this is decided evidence that what is generally considered ‘intelligence’ is based on the functioning of multiple brain networks rather than a single ‘g’, and that these networks are largely independent. Thus, the need to focus on and maintain task-relevant information maps onto one particular brain network, and is one strand. Another network specializes in transforming information, regardless of source or type. These, it would seem, are the main processes involved in fluid intelligence, while the Verbal component most likely reflects crystallized intelligence. There are also likely to be other networks which are not perhaps typically included in ‘general intelligence’, but are nevertheless critical for task performance (the researchers suggest the ability to adapt plans based on outcomes might be one such function).
The obvious corollary of all this is that similar IQ scores can reflect different abilities for these strands — e.g., even if your working memory capacity is not brilliant, you can develop your reasoning and verbal abilities. All this is consistent with the growing evidence that, although fundamental WMC might be fixed (and I use the word ‘fundamental’ deliberately, because WMC can be measured in a number of different ways, and I do think you can, at the least, effectively increase your WMC), intelligence (because some of its components are trainable) is not.
If you want to participate in this research, a new version of the tests is available at http://www.cambridgebrainsciences.com/theIQchallenge
(2012). Fractionating Human Intelligence.
Neuron. 76(6), 1225 - 1237.
Shrinking of the frontal lobe has been associated with age-related cognitive decline for some time. But other brain regions support the work of the frontal lobe. One in particular is the cerebellum. A study involving 228 participants in the Aberdeen Longitudinal Study of Cognitive Ageing (mean age 68.7) has revealed that there is a significant relationship between grey matter volume in the cerebellum and general intelligence in men, but not women.
Additionally, a number of other brain regions showed an association between gray matter and intelligence, in particular Brodmann Area 47, the anterior cingulate, and the superior temporal gyrus. Atrophy in the anterior cingulate has been implicated as an early marker of Alzheimer’s, as has the superior temporal gyrus.
The gender difference was not completely unexpected — previous research has indicated that the cerebellum shrinks proportionally more with age in men than women. More surprising was the fact that there was no significant association between white memory volume and general intelligence. This contrasts with the finding of a study involving older adults aged 79-80. It is speculated that this association may not develop until greater brain atrophy has occurred.
It is also interesting that the study found no significant relationship between frontal lobe volume and general intelligence — although the effect of cerebellar volume is assumed to occur via its role in supporting the frontal lobe.
The cerebellum is thought to play a vital role in three relevant areas: speed of information processing; variability of information processing; development of automaticity through practice.
(2011). Cerebellar brain volume accounts for variance in cognitive performance in older adults.
Cortex. 47(4), 441 - 450.
Using a large data set of 241 brain-lesion patients, researchers have mapped the location of each patient's lesion and correlated that with each patient's IQ score to produce a map of the brain regions that influence intelligence. Consistent with other recent findings, and with the theory that general intelligence depends on the brain's ability to integrate several different kinds of processing, they found general intelligence was determined by a distributed network in the frontal and parietal cortex, critically including white matter association tracts and frontopolar cortex. They suggest that general intelligence draws on connections between regions that integrate verbal, visuospatial, working memory, and executive processes.
(2010). Distributed neural system for general intelligence revealed by lesion mapping.
Proceedings of the National Academy of Sciences. 107(10), 4705 - 4709.
Older news items (pre-2010) brought over from the old website
Genes more important for IQ as children get older
Data from six studies carried out in the US, the UK, Australia and the Netherlands, involving a total of 11,000 pairs of twins, has revealed that genes become more important for intelligence as we get older. The researchers calculated that genes accounted for some 41% of the variation in intelligence in 9 year olds, rising to 55% in 12 year olds, and 66% in 17 year olds. It was suggested that as they get older, children get better at controlling (or perhaps are allowed to have more control over) their environment, which they do in a way that accentuates their ‘natural’ abilities — bright children feed their abilities; less bright children choose activities and friends that are less challenging.
Haworth, C.M.A. et al. 2009. The heritability of general cognitive ability increases linearly from childhood to young adulthood. Molecular Psychiatry, advance online publication 2 June 2009; doi: 10.1038/mp.2009.55
A gene that influences intelligence
A study involving more than 2000 people from 200 families has found a link between the gene CHRM2, that activates multiple signaling pathways in the brain involved in learning, memory and other higher brain functions, and performance IQ. Researchers found that several variations within the CHRM2 gene (which is on chromosome 7) could be correlated with slight differences in performance IQ scores, which measure a person's visual-motor coordination, logical and sequential reasoning, spatial perception and abstract problem solving skills, and when people had more than one positive variation in the gene, the improvements in performance IQ were cumulative. Intelligence is a complex attribute that results from a combination of many genetic and environmental factors, so don’t interpret this finding to mean we’ve found a gene for intelligence.
(2007). Association of CHRM2 with IQ: Converging Evidence for a Gene Influencing Intelligence.
Behavior Genetics. 37(2), 265 - 272.
Common gene version optimizes thinking but carries a risk
On the same subject, another study has found that the most common version of DARPP-32, a gene that shapes and controls a circuit between the striatum and prefrontal cortex, optimizes information filtering by the prefrontal cortex, thus improving working memory capacity and executive control (and thus, intelligence). However, the same version was also more prevalent among people who developed schizophrenia, suggesting that a beneficial gene variant may translate into a disadvantage if the prefrontal cortex is impaired. In other words, one of the things that make humans more intelligent as a species may also make us more vulnerable to schizophrenia.
(2007). Genetic evidence implicating DARPP-32 in human frontostriatal structure, function, and cognition.
Journal of Clinical Investigation. 117(3), 672 - 682.
Closing in on the genes involved in human intelligence
A genetic study claims to have identified two regions of the human genome that appear to explain variation in IQ. Previous research has suggested that between 40% and 80% of variation in human intelligence (as measured by IQ tests) can be attributed to genetic factors, but research has so far failed to identify these genes. The new study has identified specific locations on Chromosomes 2 and 6 as being highly influential in determining IQ, using data from 634 sibling pairs. The region on Chromosome 2 that shows significant links to performance IQ overlaps a region associated with autism. The region on Chromosome 6 that showed strong links with both full-scale and verbal IQ marginally overlapped a region implicated in reading disability and dyslexia.
(2005). A Genomewide Scan for Intelligence Identifies Quantitative Trait Loci on 2q and 6p.
The American Journal of Human Genetics. 77(2), 318 - 326.
Damaged brains show regions involved in intelligence
Comparison of brain scans of 241 patients with differing degrees of cognitive impairment from events such as strokes, tumor resection, and traumatic brain injury, has correlated the location of brain injuries with scores on each of the four indices in the Wechsler Adult Intelligence Scale (WAIS), the most widely used intelligence test in the world. It was found that lesions in the left frontal cortex were associated with lower scores on the verbal comprehension index; lesions in the left frontal and parietal cortex were associated with lower scores on the working memory index; and lesions in the right parietal cortex were associated with lower scores on the perceptual organization index. A surprisingly large amount of overlap in the brain regions responsible for verbal comprehension and working memory may suggest that these two measures of cognitive ability may actually represent the same type of intelligence.
(2009). Lesion Mapping of Cognitive Abilities Linked to Intelligence.
Neuron. 61(5), 681 - 691.
When it comes to intelligence, size matters
The NIH MRI Study of Normal Brain Development now contains data from more than 500 children and adolescents from newborns to 18-year-olds, who had brain scans multiple times over a period of years as well as various cognitive tests. A sample of 216 healthy 6 to 18 year old brains from the dataset reveal that there is a positive link between cortical thickness and cognitive ability in many areas of the frontal, parietal, temporal and occipital lobes. The regions with the greatest relationship were the 'multi-modal association' areas, where information converges from various regions of the brain for processing. The finding supports a distributed model of intelligence.
(Submitted). Positive association between cognitive ability and cortical thickness in a representative US sample of healthy 6 to 18 year-olds.
Intelligence. 37(2), 145 - 155.
Processing speed component of intelligence is largely inherited
A new kind of scanner used on the brains of 23 sets of identical twins and 23 sets of fraternal twins has revealed that myelin quality is under strong genetic control in the frontal, parietal, and left occipital lobes, and that myelin quality (in the cingulum, optic radiations, superior fronto-occipital fasciculus, internal capsule, callosal isthmus, and corona radiata) was correlated with intelligence scores. Myelin governs the speed with which signals can travel along the axons of neurons, that is, how fast we can process information. The researchers are now working on finding the genes that may influence myelin growth.
(2009). Genetics of Brain Fiber Architecture and Intellectual Performance.
J. Neurosci.. 29(7), 2212 - 2224.
Intelligence and rhythmic accuracy go hand in hand
And in another perspective on the nature of intelligence, a new study has demonstrated a correlation between general intelligence and the ability to tap out a simple regular rhythm. The correlation between high intelligence and a good ability to keep time, was also linked to a high volume of white matter in the parts of the frontal lobes involved in problem solving, planning and managing time. The finding suggests that the long-established correlation of general intelligence with the mean and variability of reaction time in elementary cognitive tasks, as well as with performance on temporal judgment and discrimination tasks, is a bottom-up connection, stemming from connectivity in the prefrontal regions.
(2008). Intelligence and Variability in a Simple Timing Task Share Neural Substrates in the Prefrontal White Matter.
J. Neurosci.. 28(16), 4238 - 4243.
Brain network related to intelligence identified
A review of 37 imaging studies may have finally answered an age-old question: where is intelligence. Following on from recent evidence suggesting that intelligence is related to how well information travels throughout the brain, the researchers believe they have identified the stations along the routes intelligent information processing takes. These stations primarily involve areas in the frontal and the parietal lobes, many of which are involved in attention and memory, and more complex functions such as language. Basically, the researchers theorize that your level of intelligence is a function of how well these areas communicate with each other. It’s particularly interesting to note that these various imaging studies had remarkably consistent results despite the different definitions of intelligence used in them.
(2007). The Parieto-Frontal Integration Theory (P-FIT) of Intelligence: Converging Neuroimaging Evidence.
Behavioral and Brain Sciences. 30(02), 135 - 154.
Intelligence based on the volume of gray matter in certain brain regions
Confirming earlier suggestions, the most comprehensive structural brain-scan study of intelligence to date supports an association between general intelligence and the volume of gray matter tissue in certain regions of the brain. Because these regions are located throughout the brain, a single "intelligence center" is unlikely. It is likely that a person's mental strengths and weaknesses depend in large part on the individual pattern of gray matter across his or her brain. Although gray matter amounts are vital to intelligence levels, only about 6% of the brain’s gray matter appears related to IQ — intelligence seems related to an efficient use of relatively few structures. The structures that are important for intelligence are the same ones implicated in memory, attention and language. There are also age differences: in middle age, more of the frontal and parietal lobes are related to IQ; less frontal and more temporal areas are related to IQ in the younger adults. Previous research has shown the regional distribution of gray matter in humans is highly heritable. The findings also challenge the recent view that intelligence may be a reflection of more subtle characteristics of the brain, such as the speed at which nerve impulses travel in the brain, or the number of neuronal connections present. It may of course be that all of these are factors.
(2004). Structural brain variation and general intelligence.
NeuroImage. 23(1), 425 - 433.
Brain size does matter, but differently for men and women
A study involving the intelligence testing of 100 neurologically normal, terminally ill volunteers, who agreed that their brains be measured after death, found that a bigger brain size is correlated with higher intelligence in certain areas, but there are differences between women and men. Verbal intelligence was clearly correlated with brain size, accounting for 36% of the verbal IQ score, for women and right-handed men — but not for left-handed men. Spatial intelligence was also correlated with brain size in women, but much less strongly, while it was not related at all to brain size in men. It may be that the size or structure of specific brain regions is related to spatial intelligence in men. Brain size decreased with age in men over the age span of 25 to 80 years, suggesting that the well-documented decline in visuospatial intelligence with age is related, at least in right-handed men, to the decrease in cerebral volume with age. However age hardly affected brain size in women.
(2006). Intelligence and brain size in 100 postmortem brains: sex, lateralization and age factors.
Brain: A Journal of Neurology. 129(Pt 2), 386 - 398.
Correlation between brain volume and intelligence
An analysis of 26 previous international studies involving brain volume and intelligence has found that, on average, intelligence (as measured by standardized intelligence tests) increases with increasing brain volume. The correlation was higher for females than males, and for adults compared to children.
(Submitted). Big-brained people are smarter: A meta-analysis of the relationship between in vivo brain volume and intelligence.
Intelligence. 33(4), 337 - 346.
A copy of the study is available at http://www.vcu.edu/uns/Releases/2005/june/McDaniel-Big%20Brain.pdf
IQ-related brain areas may differ in men and women
An imaging study of 48 men and women between 18 and 84 years old found that, although men and women performed equally on the IQ tests, the brain structures involved in intelligence appeared distinct. Compared with women, men had more than six times the amount of intelligence-related gray matter, while women had about nine times more white matter involved in intelligence than men did. Women also had a large proportion of their IQ-related brain matter (86% of white and 84% of gray) concentrated in the frontal lobes, while men had 90% of their IQ-related gray matter distributed equally between the frontal lobes and the parietal lobes, and 82% of their IQ-related white matter in the temporal lobes. The implications of all this are not clear, but it is worth noting that the volume of gray matter can increase with learning, and is thus a product of environment as well as genes. The findings also demonstrate that no single neuroanatomical structure determines general intelligence and that different types of brain designs are capable of producing equivalent intellectual performance.
(2005). The neuroanatomy of general intelligence: sex matters.
NeuroImage. 25(1), 320 - 327.
Individual primates display variation in general intelligence
Research into cognition of non-human animals has been concerned almost entirely with the abilities of the species, not with individual variation within a species. Now a study of 22 cotton-top tamarins has revealed that these monkeys, like humans, also display substantial individual variation on tests of broad cognitive ability, although the degree of variation does seem significantly smaller than it is among humans (individual variability accounted for some 20% of the monkey’s performance, while it accounts for some 40-60% of human’s performance on IQ tasks). It may be that greater variability has been an important factor in human brain evolution.
(2009). General Intelligence in Another Primate: Individual Differences across Cognitive Task Performance in a New World Monkey (Saguinus oedipus).
PLoS ONE. 4(6), e5883 - e5883.
Full text available at http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0005883
Bigger is smarter: brain size predicts intelligence in different species
Animals with larger body sizes generally have larger brains, and it has generally been assumed that larger animals require larger nervous systems to coordinate their larger bodies. Consequently, comparison of brain size across different animal species, as an indirect measurement of intelligence, have controlled for body size. New research however suggests that, although some correction is probably needed, completely controlling for body size is almost certainly a mistake. Both overall brain size and overall neocortex size proved to be good predictors of intelligence in different primate species.
(2007). Overall Brain Size, and Not Encephalization Quotient, Best Predicts Cognitive Ability across Non-Human Primates.
Brain, Behavior and Evolution. 70(2), 115 - 124.
Size of brain areas does matter — but bigger isn't necessarily better
In a fascinating mouse study that overturns our simplistic notion that, when it comes to the brain, bigger is better, researchers have found that there is an optimal size for regions within the brain. The study found that if areas of the cortex involved in body sensations and motor control are either smaller or larger than normal, mice couldn’t run an obstacle course, keep from falling off a rotating rod, or perform other tactile and motor behaviors that require balance and coordination as well as mice with normal-sized areas could. It now seems that the best size in one that is best tuned to the context of the neural system within which that area functions — which is not really so surprising when you consider that every brain region acts as part of a network, in conjunction with other regions. This study builds upon a previous discovery by the same researchers, that a gene controls how the cortex in mice is divided during embryonic development into its functionally specialized areas. Different levels of the protein expressed by this gene changes the size of the sensorimotor areas of the cortex. It is known that significant variability in cortical area size exists in humans, and this may explain at least in part variability in human performance.
(2007). Cortical area size dictates performance at modality-specific behaviors.
Proceedings of the National Academy of Sciences. 104(10), 4153 - 4158.
Full text is available at http://tinyurl.com/2tpyhe
Bigger brains associated with domain-general intelligence
Analysis of hundreds of studies testing the cognitive abilities of non-human primates provides support for a general intelligence, and confirms that the great apes are more intelligent than monkeys and prosimians. Individual studies have always been criticized for not clearly ensuring that one species wasn’t out-performing another simply because the particular testing situation was more suited to them. However, by looking at so many varied tests, the researchers have overcome this criticism. Although there were a few cases where one species performed better than another one in one task and reversed places in a different task, overall, some species truly outperformed others. The smartest species were clearly the great apes — orangutans, chimpanzees, and gorillas. Moreover, there was no evidence that any species performed especially well within a particular paradigm, contradicting the theory that species differences in intelligence only exist for narrow, specialized skills. Instead, the results argue that some species possess a broad, domain-general type of intelligence that allows them to succeed in a variety of situations.
Deaner, R.O., van Schaik, C.P. & Johnson, V. 2006. Do some taxa have better domain-general cognition than others? A meta-analysis of nonhuman primate studies. Evolutionary Psychology, 4, 149-196.
Full-text available at http://human-nature.com/ep/downloads/ep04149196.pdf