Skip to content

Conversation with Bob Williams on David Piffer, Latent Variable Analysis, High Correlations with the g Factor, Executive Function, Leonardo da Vinci, DMN, Booze, Promiscuity, and Charles Murray: Retired Nuclear Physicist (4)


Author(s): Scott Douglas Jacobsen

Publication (Outlet/Website): In-Sight: Independent Interview-Based Journal

Publication Date (yyyy/mm/dd): 2021/09/15


Bob Williams is a Member of the Triple Nine Society, Mensa International, and the International Society for Philosophical Enquiry. He discusses: the more evidenced theories of creativity similar to g or general intelligence as the majority position of researchers in the field of general intelligence; theories of genius; the main figures in these areas of creativity and genius connected to the research on g; personality differences between scientists and artists; conscientiousness; the ability to think; the expected probability of genius at higher and higher cognitive rarities; Howard Gardner; Robert Sternberg; the works of Arthur Jensen building on Charles Spearman; and the questions remaining about genius.

Keywords: alcoholism, Arthur Jensen, Bob Williams, Booze, Camilla Persson Benbow, Charles Murray, creativity, David Becker, David Lubinski, David Piffer, Dean Keith Simonton, Default Mode Network, Executive Function, Flynn Effect, g Factor, genius, Hans Eysenck, Ian Deary, Latent Variable Analysis, Leonardo da Vinci, Linda Gottfredson, Michael Woodley, Nyborg, Promiscuity, Richard Haier, Richard Lynn, Richard Sternberg, sex drive, The Wechsler Adult Intelligence Scale.

Conversation with Bob Williams on David Piffer, Latent Variable Analysis, High Correlations with the g Factor, Executive Function, Leonardo da Vinci, DMN, Booze, Promiscuity, and Charles Murray: Retired Nuclear Physicist (4)

*Please see the footnotes, bibliography, and citation style listing after the interview.*

Scott Douglas Jacobsen: After a hiatus, round four, what would make a general test of creativity valid? Has David Piffer proposed anything? 

Bob Williams[1],[2]*: Piffer has done a good bit of work related to creativity and published several papers on it. To avoid congestion with my answer, I will append references to some of these papers. One of his particularly interesting observations: “There is some evidence that schizotypal triats and temperament are associated with creativity. Schizotypal traits as measured by the O-LIFE questionnaire were related to creative thinking styles and a subscale (but not the other three scales) ImpNon (Impulsive Noncomformity) was positively correlated to Divergent Thinking tasks in a sample of British students.“ 

Among the things he mentions in his papers are that Openness and low Conscientiousness are predictors of creativity. This has high face value and indirectly links creativity to intelligence (via Openness). He found a correlation of 0.54 between scientific and artistic creativity that was 70% genetic. Piffer suggested that the best measure of creativity is the impact of a work on its creative field. I like that definition more than the usual one of something novel and useful

From my perspective, measuring creativity is difficult. It is not like intelligence in that we don’t have a positive manifold and we don’t have good ways to check the measurement instruments. One of the problems I see is the lack of importance in creativity below the level that we see in great composers, directors, writers, etc. If a person has a very low level of creativity, or even no realistically detectable level, he will not suffer in the way that the same low standing would cause problems relative to intelligence. Piffer referred to two kinds of creativity: Big-C (as in true genius) and Pro-C (someone at a level where he can work professionally in a creative discipline). If we add one more category, Little C, we have a group where there is a range of creativity, but where it has little impact. 

People actually try to measure creativity over a full range. I’m not sure why or whether they have paid much attention to how the Little-C people are affected by their level of creativity. 

Tests have a construct validity and an external validity. The construct (internal) validity is simply an indication that the test is measuring the thing it is supposed to measure–in this case, creativity. The treatment of construct validity is less rigorous than a test of external (predictive) validity. One way it 

is done is by comparison to tests or other means of making the measurement. If it matches conventional expectations, it is showing internal validity. In the case of intelligence, the usual method is to factor analyze the test and compare the resulting factors to those found in other tests that are believed to show construct validity. 

If we consider validity to mean accuracy, the question is one of how well the test predicts creative output. If we have people at two significantly different levels of creativity, can we use their output to validate the measure, as we do in intelligence testing? I don’t know the answer; I see the whole approach to creativity measurement as fuzzy, even when compared to other life sciences. 

The more important validity is external or predictive validity, which tells us that the test is measuring things that can be predicted and verified. If the test shows that someone is in the 90th percentile of creativity, we expect that the person will display high levels of creativity in his job and life. For example, he may be a successful screenwriter or composer. Predictive validity is central to the whole

notion of being able to meaningfully test for creativity. If we are measuring things that actually predict real world outcomes, the test is useful. If it fails this, the test is of questionable value. 

Jacobsen: Why is the reliance on latent variable analysis important for the study of creativity? 

Williams: Latent traits are found in multifaceted constructs, including creativity. The use of latent traits allows the researcher to show how multiple variables interact and form a structure. Remote association tests are used in creativity research with good results. The difficulty level of making specific connections (item level in the test) can be determined using latent trait models. This is similar to Item Response Theory as used in intelligence tests. 

Jacobsen: Why is the reliance on latent variable analysis important for the study of intelligence? 

Williams: The often displayed hierarchical structure of intelligence is a representation of latent tra its. These identify narrow and broad abilities and g. All of these are latent traits and are essential to the understanding of intelligence. It is difficult to overstate the importance of g in the study of intelligence. It translates directly to the study of the brain, is remarkably stable over lifespan, and explains life outcomes better than any other single parameter. 

Jacobsen: What five items or tasks in formal intelligence tests have the highest correlation with the g factor? 

Williams: The g loadings of various factors are test dependent. For example, vocabulary is a well known factor that usually shows a very high g loading. But its specific loading depends on the structure of the test and the number of test items that correspond to each factor. If you add more test items, it tends to skew the loadings upwards. Some tests are designed to use only a single category of test items. The best known of these is the Raven’s Progressive Matrices. It can be factor analyzed to show that it has factors other than g, but those factors are usually ignored because they are not the traditional ones seen in comprehensive tests, such as the WAIS. 

The WISC-IV has only 5 Stratum II factors. Here are the g loadings for those: 

 Comprehension-Knowledge (Gc) __ .80 

 Fluid reasoning (Gf) __________ .95 

 Short-Term Memory (Gsm) _______ .62 

 Visual Processing (Gv) ________ .67 

 Processing Speed (Gs) _________ .27  

Timothy Salthouse created a factor structure from 33 of his studies (about 7,000 people, ages 18 through 95) and also found 5 Stratum II factors. The g loadings he found: 

Reasoning _____________________ .95 

Spatial ability _______________ .91 

Memory _______________________ .66 

Processing speed ______________ .60 

Vocabulary ____________________ .73 

Johnson and Bouchard found a natural structure of intelligence by using the 15 test Hawaii Battery, the

Comprehensive Ability Battery, and the The Wechsler Adult Intelligence Scale. They eliminated some subtests to avoid duplication. When they factor analyzed the massive test, a four stratum structure emerged. I consider this to be the best fully analyzed study of the structure of intelligence. The top 5 g loadings: 

Verbal ________________________ .96 Stratum III factor Perceptual ____________________ .99 Stratum III factor Image rotation ________________ .97 Stratum III factor Scholastic ____________________ .88 Stratum II factor 

Fluency _______________________ .83 Stratum II factor 

The point of presenting these different results is to show how different tests cause different factors and different loadings. The very high loadings, in the last set, are the result of the large number of diverse test items used. This causes most non-g factors to cancel out. 

Jacobsen: What do these five tasks or sub-tests tell us about the structure of general intelligence and the human brain? 

Williams: If you look at the three sets of factors, you see that they are similar. Tests are generally designed to either fit the three stratum Cattell-Horn-Carroll model, or are forced to produce another three stratum structure. All tests show one general factor, that may appear at stratum II, III, or IV. Ergo, we have accepted and repeatedly confirmed Spearman’s early findings. I am always amazed by how much he reported over a century ago and how dead-on accurate his findings were. 

Richard Haier formulated the Efficiency Hypothesis based on positron emission tomography studies he did, starting in 1988. These showed high glucose uptake in low IQ cohorts and lower glucose uptake in high IQ testees. It meant that, when trying to resolve the same mental task, the low IQ group required high mental effort, while the bright group required less mental effort. Some MRI work was available with Jensen wrote The g Factor (1998), but it has only been in the 21st century that we have had large MRI based studies. It has only been possible to look for g in the brain by using advanced imaging technologies. Among the most important are structural MRI, functional MRI, and diffusion tensor imaging. The latter two have provided the ability to study white matter and brain networks. 

The above comments are a necessary introduction to what has been learned about the general factor within the brain. We already knew that g was unitary at the psychometric level. Now we know that it is not unitary at the neurological level. Richard Haier and Rex Jung found 14 Brodmann Areas that are strongly related to intelligence and problem solving. They created a model known as P-FIT (parieto-frontal integration theory) [described in detail in Haier, R. J. (2017); The Neuroscience of Intelligence, Cambridge University Press]. The model involves a sequential transfer of information between the cognitive centers, ending in the frontal lobes where the integrated information is evaluated. 

The distributed nature of g within the brain has been confirmed by various studies, including focal lesion studies (using the Vietnam Head Injury Study). An important finding from this and other studies of networks is that damage to critical white matter areas causes lowered g. These areas are concentrated networks that link the P-FIT regions. Since the important cognitive centers work by information exchange, we have to think of g in the brain as the areas that are being linked as well as the efficiency of the connecting networks. 

Most of the P-FIT Brodmann Areas (BA) share their associations with g and other non-g traits. BA-10,

however, is only associated with g. This area appears to function as a control mechanism that is critical to the distributed processing nature of g. 

Jacobsen: What do current tests of general intelligence miss? 

Williams: As you would expect, different tests miss different things. While researchers today recommend comprehensive tests (WAIS and Woodcock-Johnson, etc.) other tests that are not diverse still work well for most purposes. This is because of Spearman’s indifference of the indicator. We are ultimately trying to measure g and can do that by a variety of seemingly unrelated tests. Each of the different tests (consider vocabulary and block design) is g loaded and is measuring the same g. 

But, we know from the structure of intelligence that there are factors, particularly at the broad abilities level (Stratum II) that are particularly important to some tasks. Arguably the most important of these is spatial ability. In this paper: [Spatial Ability for STEM Domains: Aligning Over 50 Years of Cumulative Psychological Knowledge; Jonathan Wai, David Lubinski, and Camilla P. Benbow; 2009, Journal of Educational Psychology Vol. 101, No. 4, 817–835.] the authors show that spatial ability is high in people who pursue engineering and sciences and its magnitude increases as the degrees held go from Bachelors, to Masters, to PhD. These fields are heavily dominated by males. At least part of the reason is that there is a sex difference in spatial ability favoring males. Some tests do not have any spatial ability test items, so they would certainly miss this ability. We know that various test designers try to force their tests to show invariance by sex, which may be why they do not include spatial ability test items. 

Jacobsen: How much can an individual train and change the degree of executive function in adult life? Is it a trainable skill or something more innate as with the g factor? 

Williams: I haven’t seen any research showing that the executive function can be enhanced by training. It seems, however, that some people can increase things such as Attention and the inhibitory function (both are components of the executive function) when needed and decrease them when that is appropriate. When we see people focused to a degree that blocks out virtually everything around them, they are using the executive function in conjunction with the inhibitory function to stay on task and to block external stimuli. All of this is strongly related to working memory. High WMC enhances the executive function and other factors such as rate of learning, the formation of long term memories, and fluid intelligence. 

Jacobsen: With someone like Leonardo Da Vinci, what would the structure of such a creative genius mind look like in real-time at peak performance? 

Williams: I don’t think we have any data that relates directly to brain imaging of true genius. If we did have it, I would expect that those in different fields (art versus science) would show behaviors that are similar to their colleagues and quite different from those in other disciplines.  

The issue of artistic and scientific creativity is interesting to me; I see it as unresolved. I once asked Rex Jung if the two forms were the same and he said that they were. Jensen, on the other hand, expressed a belief that intelligence was a larger factor in scientific creativity as compared to artistic creativity. To me, this has more face value. I think that Jung was considering how tests of creativity work over a wide range of ability and was not focused on the rare true genius brains.

Neurologists have done measurements of some people while doing a creative task, such as music improvisation. Their findings are certainly related to real-time creativity, but I do not see this as relating to the brains of Leonardo or Beethoven. The task of learning what is going on in their brains is so difficult that I think it will not be resolved for a long time. The starting problem is to find people who are actually at that level of creativity. Then we have to be able to make meaningful measurements at the moment they are inspired to create. I think that director David Lynch is at that level of creative genius, but I doubt that we can monitor him constantly and figure out when and how his brain comes up with the huge number of elements that go into the finished film. My guess is that it is a series of creative flashes, spaced by tasks that require either different kinds of thought or those that do not demand creativity. 

I would also expect that if we were lucky enough to be able to examine several creative geniuses, we would find different approaches. Some would probably go into long, deep, creative sessions and some would have multiple sudden insights that they combine to produce their works. And we might find some who do both over the course of a project. 

In the specific case of Leonardo we have the most extreme example of a polymath I can imagine. His brain would be a neurological treasure today, now that we finally have the technology to really study it. In such extreme cases of genius it is difficult to imagine what biological factors were combined to 

produce the end results that were so profound. One would have to assume that his brain was an extreme case of factors that simply do not exist together in others. From the little we were able to learn about Einstein’s brain, we know that his too was bizarrely different. 

Jacobsen: What is the DMN, default mode network? 

Williams: The DMN is the network that we use during mind-wandering, spontaneous cognition, imagination and divergent thinking. It is detectable by the presence of increased alpha-power. As is always true, things are messy. While the DMN is clearly linked to these things, the production of novel ideas seems to arise from the interaction of the DMN and various other networks. When the brain stops mind-wandering and focuses on a specific task, the DMN disengages and switches to other networks. We now know that the brain doesn’t lock in on a specific network for a prolonged period; it switches between networks. One of the things that emerged from the focal brain injury studies was the identification of the regulatory role of Brodmann Area 10 as I previously mentioned. I am unsure if this includes network switching, but I think it is likely. 

I once asked Richard Haier if it was known whether solutions to problems (the kind that happen after study and then hit us unexpectedly as we are doing something unrelated) are actually made in real-time while we are in the DMN or if the answers were made subconsciously and then revealed using the DMN as a vehicle. He said we don’t know yet. 

Jacobsen: Odd question, incoming: How would a universal definition of genius expand into other species? So, we see certain traits consistent across species with some conscious cognitive capacity, so as to consider them – exceptional minds in individual species – geniuses. This would seem an enlarged consideration, biologically, of genius with potential insight into the nature of human genius, so the quality of genius itself. 

Williams: The only definitions I believe are appropriate to true human genius are those that relate to a constellation of traits, expressed at a high level. In the case of animal studies, it is difficult to measure as many behavioral traits as we see in humans. For example, researchers have found a general factor of intelligence in some animals, but that factor is based on a rather small group of different categories of problem solving. It may be possible to measure factors such as zeal and persistence in animals, but we have to see that these things are actually productive. For example, I recall a study of wolves and dogs in which there was a barrier between them and food. The wolves continued to repeat the same efforts to go directly to the food. The dogs figured out that they needed help and tried to get it from humans. The point here is that, while persistence tends to be a genius trait, it is so because the genius does not repeat the same failed effort endlessly. We have seen a lot more animal studies in recent years and they are becoming more sophisticated. It is likely that they will eventually have a wider spectrum of tests and measures of animal behavior and that may lead us to identify exceptional individuals. Related to this, much of the animal kingdom is organized around male physical strength and fighting over mates, which creates a situation where the things we see as genius in humans may not show up at all. 

Jacobsen: Why do creative people tend to drink so much? 

Williams: In the book The Cambridge Handbook of the Neuroscience of Creativity (2018) Rex E. Jung (Editor), Oshin Vartanian (Editor), there is a mention of creative professions showing twice the rate of alcoholism as found in the general population. Some of the people in these professions have creativity expectations associated with the use of alcohol. In general they seem to be right, at least for the insight stage of creativity, but as the amount of alcohol they consume increases, their creative output declines. As we know, at least from modern history, creative people tend to use other drugs as well. 

Jacobsen: Why are actors the biggest drinkers? 

Williams: The book cited above confirms that actors (60%) use alcohol at a level beyond the norm for the general public. It mentions writers as being particularly likely to have serious problems with booze. This makes sense in that writers have to constantly create new material and their “writers’ block” is often mentioned in various media. 

Jacobsen: Could those without high levels of executive function, but latent creativity, help themselves with exogenous agents such as alcohol to perform creative functions? However, this leads to the deleterious lives exhibited in high-performing creatives who have to rely on alcohol and other substances to accomplish incredible creative feats. 

Williams: I haven’t seen studies that directly address this situation. It falls into a category of research that is likely to be regarded as too dangerous unless conducted from natural data. I believe that it is a case of “a little helps, but too much hurts.” It follows the distribution that is sometimes called the inverted U curve. We see this in psychosis and neurosis, both contributing positively to genius results, but only when the level is “elevated” and not substantial. Various substances, that are used to enhance creativity, appear to work this way. The problem is that the use of the substances can become a drive and cause the user to not moderate his intake. 

Jacobsen: Is there a correlation between sex drive/promiscuity and genius? 

Williams: I can only guess, as I haven’t seen a specific study relating to it. What we often see in true genius is isolation and often no children. But I expect we can find rather extreme cases of sexual behavior, depending on the specific personalities and possibly on the category of work they do. One discussion

that relates to this: Who are the “Clever Sillies”? The intelligence, personality, and motives of clever silly originators and those who follow them; Edward Dutton, Dimitri van der Linden; Intelligence 49 (2015) 57–65. The title of the paper is somewhat misleading. From the paper: “… creative, original, uncooperative, and impulsive risk-takers. These kinds of characteristics permit them, like artists, to conceive of an original idea, thus showcasing their intelligence and creativity, and take the risks 

necessary – short term ostracism – to achieve their long term goal of high socioeconomic status. The fact that some of those whom we have assessed achieved high social status but not high economic status can thus be seen as the risks only partially paying off. In addition, the lack of sexual success among some of these figures is congruous with many geniuses not having children. But their actions can be interpreted as advantageous at the group level.” 

Jacobsen: If taking one moral perspective on it, is there a correlation between perversion or various forms and genius? 

Williams: That one falls outside of my knowledge base. I can imagine that there may be various forms of perversion, but I haven’t seen anything that explores the relationship. 

Jacobsen: When does conscientiousness become a negative trait? What contexts? I do not mean simply statements on specific professions. 

Williams: Low conscientiousness is found in artistic people and high conscientiousness is characteristic of people more likely to be found in STEM. Conscientiousness is less likely among people who use drugs (per our discussion) and who have random life patterns, consisting of no schedule or traditional jobs. The extent of problems relating to low conscientiousness is probably related to specific professions. There are lots of stories of actors who were difficult to work with, inclined to walk out, or get drunk. The most extreme cases of near-zero conscientiousness are those from the world of rock music, where performers have written the book on bad behavior and short lifespans. The “27 Club” was the subject of a documentary [27: Gone Too Soon] of at least 6 high level performers, but the total toll for young deaths is much larger. Low conscientiousness was one of many things that are obvious in the world of idol worshiped musicians. 

Jacobsen: Following from the previous question, when does conscientiousness become a positive trait? What contexts? 

Williams: In most employment situations, where a person has responsibilities that relate to an entire group, conscientiousness is valuable. You want to have the person who, when given a job, can be counted on to get it done, even if there is a tight deadline. The performers we discussed would not be a good choice for this kind of business. 

Jacobsen: How much is productivity a measure of genius? 

Williams: The magnitude of output of true geniuses is high. We see massive quantities of output from composers, painters, and writers, even from those who died very young. Part of this may be related to the speed with which some art is created. I once watched a documentary of Picasso, showing him painting. He was fast and changed the painting frequently by painting over parts of the painting repeatedly. I seriously doubt that a sculptor could chisel his way through a piece of marble quickly. The task is at least partly related to productivity, in the sense of output rate.

Jacobsen: Are there any substances that temporarily or artificially increase tissue functionality? Or, more generally, what about substances going in either direction of high FA and low FA temporarily due to their intake? What would be the expected effects and productive outputs from such intake, when heading into artificial high FA and artificial low FA? Perhaps, the wording isn’t sufficiently precise in the questions, but, I think, the curiosity for the idea is there. 

Williams: That is a thought provoking question. For the benefit of readers who are not familiar with FA, in this context, it means fractional anisotropy. This is a measure of diffusivity. If FA is zero, the medium is isotropic; if it is at the other extreme, 1, it means that the diffusion is along one axis and there is no loss to radial diffusion. In brain imaging, we see high FA as desirable; this means high tissue integrity. 

In the cases I have seen reported, FA is discussed as a tissue property that does not fluctuate. If it goes down, it stays down. But there may be studies showing that there are agents that can reduce FA temporarily and that it would return to normal when the agent is no longer present. Alcohol or drugs associated with hallucination might have some impact on FA (guessing). During the past week, we had the annual conference of the International Society for Intelligence Research. One factor that was discussed during an open session was the impact of anesthesia on the brain. I was unaware that it is believed to be damaging to intelligence. Unfortunately, the discussion was in the context of a one-way trip down. 

The reason this could relate to creativity (assuming that it happens) is that low FA can result in the brain following longer paths to join information. This presumably causes brain regions that are not related to the task at hand to be activated and may result in the formation of remote associations of the type associated with creativity. This would happen if a network has broken connections, thereby causing the brain to follow longer paths to complete tasks that recruit information from different parts of the brain. 

Jacobsen: For Mensa International, Intertel, the Triple Nine Society, the Prometheus Society, and the Mega Society, you observed a trend or pattern – non-absolute – of individuals who may not succeed in “education, profession, and personal relationships.” They seem more prone to becoming a part of them. Jensen mentioned in the Mega Press interview the dilettantish nature of the interactions and a void in deep, critical evaluation. Yet, the qualifications of the societies ground themselves in higher, sometimes abnormally, higher than normal IQs. Which leads to an associated, but somewhat distant, question, what is IQ missing regarding critical intelligence if that’s the case? The stereotype with some truth to it: A genius level IQ without a sense of the mechanics of the social and professional world, or the right question to probe an intellectual problem appropriately. 

Williams: It is certainly true and easily observable that these groups are statistically more attractive to people who have failed to establish meaningful careers, despite having high intelligence. Jensen mentioned that he was personally able to form satisfying relationships with his work colleagues and that, while all were bright, none belonged to Mensa (the only example he mentioned). Part of the answer may lie in the nature of personality. Of the Big Five, only Openness is significantly correlated with intelligence. That leaves a lot of room for other factors, as well as those that only appear in other personality test batteries, to cause problems. In fact, if you look at the four other traits, all of them can be expressed in a direction that could be poisonous to careers. I would expect that two traits would be particularly damaging: low conscientiousness and high neuroticism. 

Jacobsen: Does Charles Murray account for global population growth with the 1.5 times per year number in genius emergence? In short, is this number larger in more recent history with vastly more people living at the same time compared to the past, e.g. 0.75 times per year at some point in the past and 3 times per year at a time closer to the present? This is taking into account the speculation of a decline in mean national intelligence. 

Williams: No. Murray simply identified 4,002 people of extreme eminence over the period 800 B.C. to 1950 and limited his study to arts and sciences. The problem of computing the rate of genius birth is complicated because of the decline in real intelligence that is largely driven by the negative correlation between intelligence and fertility rate. [See At Our Wits’ End: Why We’re Becoming Less Intelligent and What It Means for the Future, by E. A. Dutton & M. A. Woodley of Menie. Exeter, UK: Imprint Academic.]  Dutton and Woodley express concern that the births of geniuses will become increasingly rare, despite many births among low intelligence groups. They fear that this will or already has led to a reduction in innovation and discovery rates. 

Jacobsen: What are the difficulties in estimation of mean national intelligence? 

Williams: We approach the study of national intelligence (the comparison of mean IQs by nation) by gathering as many datasets as possible for the nation in question, then converting them to a single standard. The conversion is parallel in principle to what we would do in a national economic comparison. In the latter case, we would convert all currencies to a single reference, such as the dollar or Euro. The standard we use for intelligence is white British. This standard is sometimes called the Greenwich IQ Standard. The details of conversion are discussed in Richard Lynn & David Becker (2019). The Intelligence of Nations. Ulster Institute for Social Research, London GB ISBN 9780993000157. 

For the most part, the difficulties are simply that it takes a large amount of work to deal with the full set of nations for which we have IQ data. There are lots of studies available for developed nations and most emerging nations, but some poorly developed nations have limited data available. When The Bell Curve was written, there were only a few reports of intelligence for sub-Saharan Africa. But since 1994, we have had data pouring in from around the world. Today we have so much data for many nations that we can map intelligence within the nation by states or provinces. These data have resulted in within-nation studies that have shown patterns that seem to largely reflect migration and economic factors. A rather large number of nations exhibit a higher mean IQ in the northern regions and a decrease at lower latitudes. The opposite is seen in Britain, where the brightest region is in the south and the dullest in the north. Researchers have explained this as the result of the decline in the coal mining industry and its impact on migration. In India, Intelligence is higher in the South and in states with a coastline (indicating economic factors relating to trade). When Richard Lynn first reported the intelligence gradient for Italy (higher in the North) he explained it by noting that mean local intelligence reflects the fraction of the population that immigrated from the Near East and North Africa. In that study regional IQs predict income at r = 0.937. This resulted in papers objecting to his findings and that resulted in an exchange of published papers. It appears that Lynn was (as I would have guessed) right. [The title of the initial paper is a good summary of what was found. In Italy, north–south differences in IQ predict differences in income, education, infant mortality, stature, and literacy; Richard Lynn; Intelligence 38 (2010) 93–100.] 

Jacobsen: What is the validity of the measurements done globally now? Some areas must be more reliable than others because of the finances and expertise to do it properly. 

Williams: I haven’t seen any reports of reliability for the IQ scores used in the national level studies. When IQ and the Wealth of Nations appeared, two things were triggered. The first was that researchers began to try different curve fits and concluded that a log scale works best and that nations with IQs below 90 were either in poverty or had valuable natural resources (usually oil). Some researchers attacked Lynn as usual. They claimed that his numbers were wrong; that they were based on too few data; that the nations were he used neighboring scores to estimate means could not be true; and that his entire study was politically incorrect and could not be trusted. But, the data, as mentioned above, kept coming in from sources around the world. Now we can say that Lynn was right on every point and that even the estimated mean scores were very close to measured scores that are now available. The validity of this work is shown in the many things that national mean IQ predicts: At the national level, mean national IQ correlates positively with per capita GDP, economic growth, economic freedom, rule of law, democratization, adult literacy, savings, national test scores on science and math, enrollment in higher education, life expectancy, and negatively with HIV infection, employment, violent crime, poverty, % agricultural economy, corruption, fertility rate, polygyny, and religiosity. These are the kinds of things used to establish the predictive validity of IQ tests. Naturally, there are confounds, such as the presence of natural resources in some low IQ nations, but the statistical predictions remain powerful. 

Jacobsen: Who else, other than Gardner, are individuals qualifying as individuals who are “in a category that is highly regarded by the general public and not by many serious intelligence researchers”? 

Williams: The first who comes to mind is Robert Sternberg. His triarchic theory was shredded by Linda Gottfredson and is not something other researchers have accepted. He has been criticized for grossly over citing his own work. In general, the public has embraced such things as emotional intelligence, grit, mindset, and other tabloid worthy inventions. In his book, In the Know: 35 Myths about Human Intelligence, Russell Warne goes through his list of things that the public loves to love but which are not science. I think the single most disliked person (from the perspective of researchers) is the late Stephen Jay Gould. His book, The Mismeasure of Man was an intentional distortion of facts and is loved by the public because politically left people wanted to hear his false message. He attacked g and other factors, such as brain size, using outrageous comparisons to what researchers were doing in the distant past. It was almost as extreme as claiming that chemistry is worthless because alchemists were unsuccessful. 

Jacobsen: Who are the most serious researchers and commentators on genius, on IQ, and on the g factor? I take those as three related, but separate, questions in one. 

Williams: Genius – Jensen wrote a good piece on genius in the last chapter of Intellectual Talent: Psychometric and Social Issues by Camilla Persson Benbow & David Lubinski; The Johns Hopkins University Press (January 22, 1997). Dean Keith Simonton has written numerous articles on genius. His work impresses me as biased and inaccurate. Eysenck wrote about genius and the personalities of genius. Some of this can be found in H. Nyborg, Editor, The Scientific Study of Human Nature: a Tribute to Hans J. Eysenck at Eighty, Pergamon, Oxford (1997). Eysenck believed that true genius required elevated neuroticism and psychoticism. Overall, the material we have about genius is based on observations of various eminent historical figures. Statistical studies are not seen because there is no satisfactory way to find and test a statistically meaningful group of such rare people. 

IQ – The most prolific and brilliant commentator on intelligence was Arthur Jensen. His lifetime output of 7 books and over 400 papers is huge and remains influential. I think that Richard Haier is probably the most important living commentator. With only 1 book and one DVD lecture set, he is nonetheless a major factor in our understanding of IQ from the neurological perspective. While

Charles Murray is accurately described as an author, he is one of the most knowledgeable intelligence scholars alive. Like Jensen, he has been willing to take the heat from the left and calmly discuss the realities of IQ. Ian Deary has been a high profile researcher and department head. Two young researchers have shown themselves to be bright, competent, and broadly focused. Michael Woodley has authored or co-authored half a dozen books, covering a wide range of topics. His work has been at the forefront of new understandings of such topics as the Flynn Effect and the decline of intelligence. Like Woodley, Stuart Ritchie has rapidly become a serious contributor to the understanding of intelligence. I have read his books and find that his writing style is particularly appealing. His most recent book, Science Fictions, is a detailed account of abuses of the scientific process of doing research and reporting it. 

Psychometric g – Jensen almost single-handedly convinced researchers worldwide that intelligence is about g and that their work should be focused on g. His book The g factor: The science of mental ability is the most cited in all of intelligence research. Linda Gottfredson has been a prolific writer of g related papers and articles. She has devoted much of her energy to explaining g and its consequences to non-experts and has made her entire output available to the public on her web site. Today, intelligence research is g research, so it is fair to say that we have lots of people writing about g and studying how it relates to the neurology of the brain. 


Can creativity be measured? An attempt to clarify the notion of creativity and general directions for future research; Davide Piffer; Thinking Skills and Creativity, Volume 7, Issue 3, December 2012 

The personality and cognitive correlates of creative achievement; Davide Piffer; Open Differential Psychology April 7th, 2014. 

Shared genetic and environmental influences on self-reported creative achievement in art and science; Yoon-Mi Hur, Hoe-Uk Jeong, Davide Piffer; Personality and Individual Differences, Volume 68, October 2014, Pages 18-22.

Appendix I: Footnotes

[1] Retired Nuclear Physicist.

[2] Individual Publication Date: September 15, 2021:; Full Issue Publication Date: January 1, 2022:

*High range testing (HRT) should be taken with honest skepticism grounded in the limited empirical development of the field at present, even in spite of honest and sincere efforts. If a higher general intelligence score, then the greater the variability in, and margin of error in, the general intelligence scores because of the greater rarity in the population.


In-Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Based on a work at


© Scott Douglas Jacobsen and In-Sight Publishing 2012-Present. Unauthorized use and/or duplication of this material without express and written permission from this site’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Scott Douglas Jacobsen and In-Sight Publishing with appropriate and specific direction to the original content. All interviewees and authors co-copyright their material and may disseminate for their independent purposes.

Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: