Human Intelligence Redescribed

Introduction

There exists a standard description of human intelligence, a description that intelligence researchers know and understand well. It is almost exclusively a biological description of human intelligence. This description ties intelligence directly to the human brain, an organ that is characterized as physically hosting and producing intelligence, with the brain's workings and abilities determined primarily by its genetic underpinnings (Basten et al., 2015; Plomin & von Stumm, 2018). External influences are given some consideration within this standard description, but these external influences are never depicted as being direct or primary (Tooley et al., 2021). The physical environment, for instance, is studied for its potential impact upon the neurons in the brain and for its effect upon genetic expression, influences that can be both deleterious to neurons and to gene expression (for instance, pollution) and that can be beneficial for neurons and for gene expression (for instance, good nutrition). Ultimately, however, in the standard description, intelligence always comes back to the human brain and remains a direct function of the brain itself (Colom et al., 2010). When today's intelligence researchers wish to observe intelligence directly, they gather DNA samples and take neuroimaging pictures, they do not probe and photograph the brain's surroundings.

Let me be clear about something from the outset. The purpose of this essay is to argue strongly against the standard description of human intelligence and to provide an alternative view that I believe brings greater overall clarification to the subject of human intelligence. But that being said, it also needs to be understood that I accept unreservedly nearly every piece of data and nearly every scientific finding that has backed the standard description of human intelligence. IQ tests, the g-factor, neural activity, genetic correlations, the importance of intelligence in everyday life—all these topics will be discussed within this essay and all these topics will be incorporated, mostly as is, into the alternative view I provide. In my way of seeing things, the standard description of intelligence is not so much wrong as it is incomplete. The standard description's fatal flaw is that it purports to tell the entire story of human intelligence, when in fact it has embraced only half of that story. Intelligence researchers do not realize this, but their standard description is missing an entire aspect of human intelligence that is essential for explaining the observed patterns in IQ exam results, not to mention essential for explaining the course of human history and the development of human behavioral modernity. The purpose of this essay is to introduce and to outline that other half of the human intelligence story.

For those who are unfamiliar with the details of the standard description, I would recommend a reading of the second edition of The Science of Human Intelligence, by Richard J. Haier, Roberto Colom, and Earl Hunt (Haier et al., 2023). This book is recent and thorough, and its authors, being themselves ardent proponents of the standard description, lay out the evidence for its case with both rigor and enthusiasm. Here you will find a treasure trove of accurate information regarding the history of intelligence tests; the psychometric analyses that underlie those tests; twin, adoptive and other family studies that undergird the genetic basis for intelligence; the latest in neuroscience techniques and their multitude of findings; the statistical associations of intelligence exam performance to academic achievement, to job success, and to various other important life circumstances; and finally a fair-minded and fact-based discussion regarding sex, age, individual, and group intelligence differences. Some briefer, albeit less complete, alternatives to The Science of Human Intelligence can be found in The Neuroscience of Intelligence, by Richard J. Haier (Haier, 2016), and in the second edition of Intelligence: A Very Short Introduction, by Ian J. Deary (Deary, 2020). As I have suggested above, notwithstanding the hypotheses and conclusions that are often stated within these books, I have no qualm or argument with the many evidentiary details that these books provide. These evidentiary details form the backbone of any reasonable study of human intelligence.

To come straight to the heart of the matter, the missing component in the standard description—the aspect that I believe is essential for arriving at a more complete understanding of the workings of human intelligence—is the artificial, structural, and assembled nature of the modern human environment. One of the ironies of this component is that it has become so ubiquitous within the modern world it has become nearly impossible to characterize adequately. We humans now live in an ocean of artificiality, with the structural aspects of that artificial environment influencing almost every one of our perceptions and behaviors. This is in the sharpest contrast to how humans must have lived only a couple hundred thousand years ago, when the species was still purely animal, with not a single artificial feature to be found anywhere within the human surroundings (Klein, 2009). No clothes, no cars, no skyscrapers, no abstract language—just nature and the bestial quest for survival and procreation. I will make the claim that the humans of that earlier era demonstrated no measurable intelligence whatsoever, just as the wild animals of today also demonstrate no measurable intelligence whatsoever, with the absence of this measurable intelligence directly attributable to the dearth of artificial construction to be found within the lived environment. And I will further make the claim that the direct and palpable source of the measurable intelligence that humans now demonstrate in great abundance is the artificial and structural aspects of the surrounding modern world, and not the neurons in the human brain. Human intelligence is originated from the ocean of artificiality in which humans now live, it is not originated from inside the human skull. And until that ocean of artificiality is fully incorporated into our description of human intelligence, the pattern of increasing IQ exam performance—as observed significantly, consistently and population wide throughout the twentieth century (Trahan et al., 2014)—that pattern will continue to make no sense.

To be clear, the brain still plays an important role in this alternative view of human intelligence, a role outlined mostly by the features of the standard description. The brain (and its underlying genetics) is the primary determining factor in general intelligence ability and in individual intelligence differences. But this neural role is not primary. Instead of hosting, producing and originating intelligence, the brain should be seen as responding to the intelligence contained within the surrounding environment, responding to that multitude of artificial construction that now exists all around us. This is entirely consistent with the well understood notion that neural systems have evolved as stimulus-response mechanisms (Meyer et al., 2017). For the brain to produce intelligence would be biologically and evolutionarily extraordinary, but for the brain to respond to the stimulus of intelligence would not. Without the stimulus of artificial construction, the human brain would have only the natural features within the surrounding environment with which to respond, the same condition that humans found themselves in a couple hundred thousand years ago, a time when humans were biologically similar to what they are today, but also a time in which humans were demonstrating no measurable intelligence. For human intelligence to exist—and for human intelligence to be fully understood—both halves of the process must be taken fully into account, both the totality of the artificial structural features contained within the lived environment, as well as the corresponding responsiveness of the human brain.

This being only the introduction, I do not expect the reader to accept such ideas without a reasonable amount of skepticism. At the same time, I do hope the reader will provide me the opportunity to expand upon these concepts in greater detail. The remainder of this essay will outline both the reasoning and the evidence behind this alternative approach to describing human intelligence, an approach that gives full and primary weight to the totality of artificial construction contained within the human environment. To accomplish this task, several topics will be explored upon the way: the scientific definition of intelligence; general intelligence ability versus measurable intelligence; the Flynn effect; the nature and impact of artificial construction; the content of intelligence tests; what the standard description gets right, and what the standard description gets wrong; and finally, a discussion of the true nature of genius. My hope is that by the end of these discussions, it will have become clear to the reader why the standard description of intelligence is fundamentally inadequate, and why a more complete description must be offered in its place.

I cannot overemphasize the importance of this topic. No concept is more critical to a thorough understanding of modern humanity than is the concept of intelligence. It is essential that we get this right.

The Scientific Definition of Intelligence

In common use, intelligence is one of those words that has a broad range of both denotations and connotations. Ask ten different people to define what intelligence is and you will get ten different answers, and this despite the fact that given the choice, most people will concur upon which of their colleagues are more intelligent and which of them are less intelligent. We seem to instinctively know what intelligence is, but we do not seem to explicitly know what intelligence is.

Adding to the difficulty, intelligence has also become a hot-potato topic (Carl & Woodley, 2019). Discussions surrounding human intelligence, especially in any kind of ranked form, lead to unease and argumentation within some portions of the population, causing many to suggest that the subject would be better off left untouched.

Let me be clear about where I stand and what meaning of intelligence will form the basis of the discussion within this essay. I believe intelligence is the most important topic there is for an understanding of the circumstances of modern humanity—we would ignore the subject at our peril. And I am perfectly prepared to provide an explicit definition.

The explicit definition that I will use for intelligence is the scientific definition of intelligence. This definition has been around for more than a century now and it has produced a cornucopia of useful information. No other definition of intelligence has come anywhere near it in providing both productive data and effective insight, and within the social sciences, it is one of the rare instances of a definition actually proving to be scientific. I should probably warn you first, however, that this definition does contain an ironic twist, so be prepared:

Intelligence is what gets measured by intelligence tests.

I did not create that aphorism. It comes from the psychologist Edwin Boring writing more than a century ago, responding to the first iterations of intelligence tests and to the novel statistical analyses that were arising from them (Boring, 1923). I am not exactly sure how much irony and how much literalness Boring intended, but as time has gone on, the aphorism has revealed itself as containing a good deal of both. It would seem that Boring's main objective, perfectly valid, was to remind and to warn early twentieth-century researchers of the self-referencing nature of what they were attempting to do. Imagine, if you will, that I were to tell you that the human population possesses a characteristic called gogobag, and furthermore I was setting out upon a quest to quantify gogobag within each individual and to use that information for comparative purposes. To achieve this quest, I design a series of questions and challenges designed specifically to probe for gogobag ability—gogobag tests, I call them—and I administer these tests to a representative sample of the human population. Afterwards, I perform a statistical analysis upon these results, and lo and behold, what emerges from this analysis is a common factor of correlated performance across all the gogobag tests, something I attribute to a human characteristic I call general gogobag ability. The existence of this common factor and its attributed characteristic proves that gogobag must be real, quantifiable, and comparable across the human population. I next apply for my Nobel Prize.

If you were to witness me making such an effort, you would insist at some point that I must be crazy, or at the very least, must be a very poor researcher. And yet substitute the word intelligence for gogobag and you would have a decent description of what intelligence researchers were doing in the early twentieth century (and are still doing today). The thing is, although my gogobag efforts would almost certainly turn out to be both ineffective and unproductive, the early twentieth-century intelligence efforts proved to be quite the opposite. IQ exams and their resulting analyses seemed to confirm what researchers both intuitively expected and needed to show, that levels of performance on intelligence tests tended to translate to similar performance across all intelligence tasks (Carroll, 1993), and that levels of performance on intelligence tests were very good predictors of academic performance, job success, and an entire host of other life circumstances (Kuncel et al., 2004). How had the researchers pulled off this mostly self-referencing feat? Was it sheer luck, a coincidental accident? Or a case of cheating perhaps? Or maybe a stroke of pure genius? Intelligence researchers never stopped to care. Armed with an effective and powerful tool, researchers then generated more than a century's worth of meaningful and valid study on the subject of human intelligence, including, quite significantly, a preponderance of evidence that individual intelligence abilities are driven primarily by biological, neural, and genetic factors (Deary et al., 2022). The study of human intelligence has been one of the great success stories of the last one hundred years of science, and this constitutes the literal side of Boring's aphorism. For whatever reason, intelligence tests work. Scientifically and productively speaking, intelligence is what gets measured by intelligence tests.

There are two audiences that will routinely dismiss this scientific definition of intelligence. The first audience consists of those whom I would call the lay detractors, those who insist intelligence is too complicated, too abstract, or too ethereal to be captured by something as mundane as an administered test. Consider all that intelligence purports to encompass, the lay detractor will say: problem solving, reasoning, learning, creativity, adaptation, communication—how is a couple-hour exam supposed to assess all of that? And what would an IQ exam performance even mean, given that intelligence is both malleable and impacted by so many different things—education, parenting, social circumstances, and whether or not one is having a good day? And what about those other important human qualities—resilience, maturity, confidence, altruism—where do these show up on an intelligence exam? And can humans even be thought of as intelligent at all, given the fact of their wars, their inequities, their harmfulness to the planet—how have these stupidities been included in the final tally? No, whatever intelligence is, the lay detractor will say—if indeed it is anything at all—it has to be something much different than what can be gauged by a silly test.

I could perhaps muster some empathy for these objections if I were able to make any use of them, but alas they are entirely impotent. The best I can offer to the lay detractor is a splitting compromise, one that would allow us to go our separate ways in peace. What I would suggest is that I relinquish use of the word intelligence, and employ another word instead—flugleheim, for instance. The lay detractor can then define intelligence in any way he or she prefers, while I content myself with claiming that flugleheim is what gets measured by flugleheim tests. Neither party should have an objection to these respective words, because neither party will give a whit about the other's.

But before the lay detractor accepts this compromise, I feel I must issue a warning. If that lay detractor should ever want to understand why some individuals tend to learn quickly and broadly, are more successful in their jobs, have better economic circumstances, engage in more satisfying relationships and stable marriages, experience greater health and lasting longevity, all this and much much more; should that lay detractor ever inquire about why some societies, cultures and organizations advance and thrive, while other societies, cultures and organizations stagnate and flounder; should that lay detractor ever ponder upon the puzzle of how did humanity arise from its animal past and go on to build its modern civilizations; should that lay detractor ever engage in such questions, as well as an entire host of others, he or she will discover that their word intelligence serves them no use at all. No data, no perspective, no insights—just a handful of scattered and amorphous concerns. And should that lay detractor also happen to notice that there are others who have made progress in the answering of such questions, he or she might also take notice of the accompanying use of flugleheim.

I will of course not be using the word flugleheim within this essay (it has been annoying enough already). But do keep in mind that my use of the word intelligence derives strictly from those supposedly silly intelligence tests.

The second audience that routinely dismisses the scientific definition of intelligence is, surprisingly enough, intelligence researchers. It is not that these researchers do not recognize that their entire profession has been built upon the foundation of these tests—as indeed it has (deLeyer-Tiarks et al., 2024)—but giddy with all their resulting success, these researchers have had a tendency to downplay the significance of their origins. This manifests in a handful of different ways. For instance, some intelligence researchers are wont to say that the content of an intelligence exam is essentially irrelevant, that as long as its questions remain psychometrically valid, they could be about any topic at all. Therefore, push-ups, fertility levels, the amount of food one can consume in a minute, these are perfectly viable candidates for potential inclusion on an intelligence exam—except of course they are not. Whether willing to admit it or not, intelligence researchers have always used an intuition for what type of challenge is appropriate for an intelligence test (and what type of challenge is not), the same intuition that guided them in the construction of their very first intelligence tests. And if those researchers have never stopped to figure out what characteristic it is that has allowed exam questions to be so successful in the measuring of intelligence, other than the tautological characteristic that they are being used to measure intelligence, that does not mean that such a non-tautological characteristic does not exist. This essay, respectful of the scientific definition of intelligence, will examine the content of intelligence exams with some seriousness, and will identify the non-psychometric characteristic that a question must possess in order for it to be a viable candidate for inclusion on an intelligence exam, providing a meaningful answer to Boring's warning about the study of intelligence being too self-referencing.

Another way that the current generation of intelligence researchers dismisses the scientific definition of intelligence is with its response to the Flynn effect. The Flynn effect is another topic that will be explored in greater detail later on, but by way of preview, the Flynn effect is the phenomenon observed, ever since intelligence tests were invented, that the average raw score on these tests has been significantly increasing with time (Pietschnig & Voracek, 2015). Nothing comes nearer to the scientific definition of intelligence than the raw scores on intelligence tests, and so the Flynn effect has emerged as the most significant intelligence finding that researchers have uncovered to date. But frustrated in their efforts to explain the Flynn effect, intelligence researchers have been treating the phenomenon dismissively: through an endless stream of non-compelling suggested causes, through a promulgation of obscure and complex models, and through a biased hope that the Flynn effect soon will end (Griswold, 2024a). The Flynn effect is an inconvenient truth for intelligence researchers, which certainly explains their desire to downplay it, but to downplay the significance of the Flynn effect is tantamount to downplaying the scientific definition of intelligence.

Finally, there is the easy acceptance of the notion of animal intelligence, a sentiment many researchers share with the lay detractors. I suspect what has happened is that because researchers have been able to establish such a strong connection between human intelligence and human neural activity, researchers have now taken to using neural activity as the primary indicator of the presence of intelligence. And since all sorts of animals have brains, and since all sorts of animals engage in brain-based behaviors, and since some of these behaviors are downright humanlike, it must seem perfectly reasonable to contemplate and to study intelligence within these other species (Bräuer et al., 2020). Leopards, honeybees, crows, octopuses, etc. Sometimes you even get a whiff of suggestion that in some sense, these other animal species are often more intelligent than humans.

But by the scientific definition of intelligence, all such notions of animal intelligence are pure rubbish. What gives evidence of intelligence is not neural activity, not brain-based behaviors, not humanlike behaviors. What gives evidence of intelligence is an intelligence test. And intelligence tests are not conceivable for wild animals—neither in administration nor in construction—and without an intelligence test there can be no measurable intelligence.

Think about it, what would an intelligence exam consist of for a wild animal? Since wild animals engage solely in survival and procreative activities, one can measure only for such qualities as speed, strength, attractiveness, sense acuity, fertility, social dominance, and so on—natural characteristics, nothing artificial or fabricated. But such natural characteristics do not show up on human intelligence exams, and I suspect every intelligence researcher would quickly disallow any such type of question (and for very good reason). But then what kind of challenge can be set for a wild animal? Vocabulary? Matrix patterns? Digit recall? Even if an intelligence exam could somehow be constructed that was appropriate for a wild animal, how then would one administer such a test? I challenge any intelligence researcher to devise such an exam and then travel to a remote section of Africa and deliver that exam to a pack of wild leopards—I am fairly certain I can predict the results.

The one apparent exception to this discussion about animal intelligence is the case of tamed, caged and domesticated animals, but that exception comes with an extremely large asterisk. Tamed, caged and domesticated animals are distinguished by their established connection to the human species, a connection that cannot be separated from their resulting perceptions and behaviors. Thus, when one supposedly is testing for species intelligence within these creatures, what one is more likely doing is testing the strength of their association to human intelligence. If this were not the case, then the wild version of these creatures would be just as capable of navigating the tests as are the tamed, caged and domesticated version, but this never turns out to be the result. Therefore, the intelligence being measured in these animals is not a species-specific intelligence but is instead an associated human intelligence, and if you were to remove entirely the humans and the human influences from the test scenario, you would not only no longer have tamed, caged and domesticated animals, you would also lose all the measurable intelligence.

There exists an extremely important corollary to these statements about wild animals demonstrating no measurable intelligence. Humans, as recently as a couple hundred thousand years ago, also were wild animals, living in an entirely natural setting and engaging in nothing but survival and procreative behaviors. These early humans, before their turn towards behavioral modernity, they were just as incapable of taking an intelligence test as are the wild animals of today. (Nor could such a test have been devised for these early humans—certainly nothing like Stanford-Binet or Wechsler.) It would be a mistake to assume that just because these early humans possessed a level of neural activity—indeed a level of neural activity comparable to what humans experience today—that this also means that these early humans could demonstrate a significant degree of intelligence. If intelligence is what gets measured by intelligence tests, then the human species—and not that long ago—was once demonstrating no measurable intelligence.

I suspect many of these statements are unlikely to be well received, but they are scientifically valid and correct. Scientists cannot have it both ways. Either researchers admit that by the scientific definition of intelligence modern humans have been the only creatures upon this planet to have ever demonstrated any species-specific level of measurable intelligence, or else researchers must give up the entire foundation of their profession. If intelligence cannot be measured, then there is no evidence that the intelligence exists. And to date, the only effective tool to have been devised for measuring intelligence is an intelligence test. To repeat, intelligence is what gets measured by intelligence tests.

In summary, the purpose of this section is to demonstrate my loyalty to the scientific definition of intelligence, and to insist upon the need for intelligence to be measurable by intelligence exams. In the sections which follow, I will lay out an alternative viewpoint regarding the nature of human intelligence, a viewpoint that strives both to augment and to overturn the standard description of intelligence. But it needs to be recognized that this alternative viewpoint is not being built upon any unusual, popular or novel foundations—for instance, upon the idea of emotional intelligence (Bru-Luna et al., 2021), or upon social intelligence (Kihlstrom & Cantor, 2000), or upon umpteen different kinds of multiple intelligences (Gardner, 1987), or upon any American Psychological Association definition of intelligence, or upon all the other psychobabble descriptions of intelligence that can be found within the literature. No, this alternative viewpoint is being built upon the very same foundation that has buttressed the last one hundred years of scientific pursuit in the study of human intelligence, is being built upon the foundation of intelligence tests. The content of IQ exams, the Flynn effect, the history of human intelligence—all these topics have a direct connection to what we currently know about human performance on intelligence exams. There is still a great deal of understanding that can be gleaned from the scientific definition of intelligence.

General Intelligence Ability Versus Measurable Intelligence

There are two phrases that will be frequently employed within this essay (indeed, they have already been used within the previous sections). The first phrase is general intelligence ability, and the second phrase is measurable intelligence. These two phrases refer to different things, and yet in the standard description of intelligence, there is a strong suggestion that these two phrases are tightly connected, even to the point of being mostly interchangeable. But is this in fact the case?

General intelligence ability (often referred to as g in the literature) refers to the concept that each individual possesses a certain level of aptitude that can be applied to the solving of every type of intelligence task (Gottfredson, 1998). Not visuospatial ability. Not verbal ability. Not chess ability. General ability (hence, the name). This concept arose out of the statistical analyses of the very first intelligence tests, where it was noted that there was a strong correlation of performance across all kinds of intelligence tasks (Spearman, 1904). This is sometimes referred to as the positive manifold, and it can be colloquially described by saying that an individual who does well on one type of intelligence test will also tend to do well on the other types of intelligence tests. Researchers do recognize that each individual will have some specific intelligence strengths and weaknesses, but statistically speaking, these specific strengths and weaknesses have much less influence on intelligence performance than does one's general intelligence ability.

Furthermore, it is general intelligence ability that correlates so well to outcomes in various life circumstances. There is now a large and established body of evidence indicating that those individuals with a higher level of general intelligence ability tend to have greater academic success, more favorable job performance, better economic circumstances, superior health, and so on (Gottfredson, 2002). If you want to make a prediction about how well an individual is going to do in their everyday life, no information is more valuable than to have that individual's general intelligence ability.

Also, through an assortment of family-based studies, scientists have been able to demonstrate that general intelligence ability is determined primarily by one's genetic background (Deary et al., 2022; Plomin & von Stumm, 2018). There are a variety of ways to perform such studies, but the classic and easiest-to-understand version is to find identical twins (same genetics) who have been separated at birth and then raised in different environmental circumstances. The strong tendency is that by the time these identical twins reach adulthood, they will find themselves experiencing similar circumstances and similar characteristics: similar IQs, similar academic achievement, similar job success, and so on. This means that the foundation of general intelligence ability is primarily biological, much more so than, for instance, environmental influences. This mostly biological underpinning to general intelligence ability is now being pursued with great vigor through an assortment of neuroimaging and biogenetic studies (Anderson & Holmes, 2021). When it comes to general intelligence ability, there is really no escaping one's genetic and neural background.

These findings regarding general intelligence ability have formed the backbone of intelligence science for more than one hundred years. These findings are not controversial. They are not dubious. More than a century's worth of scientific study has repeatedly indicated that individuals are born with a certain level of general intelligence ability, due mostly to their genetic background and expressed in their neural makeup, and these individuals will experience life circumstances that are mostly in line with their given level of general intelligence ability. There are exceptions, of course, but the overall correlations remain significant and strong. General intelligence ability is a foundational and enduring component of human intelligence.

Whereas general intelligence ability is a characteristic of the individual taking an intelligence test, measurable intelligence is a characteristic of the test itself. In simplest form, measurable intelligence is the amount of intelligence being demonstrated via the raw score on an intelligence exam. For instance, assuming the questions are of similar difficulty, a raw score of 60% would demonstrate a greater amount of measurable intelligence than would a raw score of 40%. This description might be adequate if there were only one type of intelligence exam in existence, but in reality there are of course many different types of exams, of varying construction and varying difficulty. This then raises the question of calibration.

The analogy to think of here is that of thermometers (representing intelligence tests, the tools doing the measuring) and heat (representing measurable intelligence, the characteristic being measured). Thermometers can be constructed so that a reading of 60 indicates the presence of more heat than does a reading of 40. But if you have two different kinds of thermometer, their readings will not be comparable unless the thermometers are somehow calibrated (as is well known from using thermometers set out against the Fahrenheit and Celsius scales). A raw score of 70% on a particular intelligence exam is not comparable against a raw score of 50% on a second intelligence exam unless the two exams have been calibrated. In rough terms, if the first exam is twice as easy as the second exam, then a score of 70% on the first exam would demonstrate less measurable intelligence than a score of 50% on the second exam.

Although intelligence tests are capable of being calibrated, especially when they are taken by the same group of people, in practice this is not always done. Nonetheless, assuming calibration is achievable, the definition of measurable intelligence can then be made complete. Measurable intelligence is the amount of intelligence being demonstrated via the calibrated raw scores on intelligence exams.

Both general intelligence ability and measurable intelligence use only relative values—there are no known absolute scales by which to assess their readings. An individual's general intelligence ability is typically estimated by that individual's IQ score, and IQ scores are themselves relative measures, determined by one's statistical ranking against an appropriate cohort of test-taking peers. If Person A has an IQ of 120, and Person B has an IQ of 105, one can say only that Person A has a greater (estimated) general intelligence ability than Person B, but one cannot say in any absolute sense by how much these abilities differ. The same is true of measurable intelligence. A raw score of 65% on an intelligence exam indicates a greater amount of measurable intelligence than does a raw score of 45%, but it does not indicate in any absolute sense by how much these amounts differ.

It is important to emphasize once again that general intelligence ability and measurable intelligence are two different concepts, each determined in its own way and each applicable to its own entity. General intelligence ability, typically estimated via an IQ score, is a characteristic of a human individual, and represents that person's overall capacity for demonstrating intelligence, including when taking an intelligence test. Measurable intelligence, determined via raw scores and calibration, is a characteristic of the test itself, and represents the amount of intelligence being directly demonstrated via a performance on that intelligence exam. Nonetheless, despite the fact that these are two differing concepts, the assumptions underlying the standard description of intelligence drive a strong motivation to connect them, to assert, broadly speaking, the equivalency—or at least the near equivalency—of general intelligence ability and measurable intelligence.

To see this, note that in a typical intelligence testing scenario, a low raw score on an exam (low measurable intelligence) would translate to a low IQ score, which would then translate to a low estimated general intelligence ability. A high raw score would produce a high IQ score and lead to a high estimated general intelligence ability. Within the standard description of intelligence, this all makes perfect sense. Within the standard description, where intelligence is regarded exclusively as a brain-based activity, a person's general intelligence ability is depicted as representing that person's neural capacity for producing intelligence, and that person's raw score on an intelligence test is depicted as corresponding to the amount of intelligence the brain has produced in this particular situation. Ability and performance appear to meet directly via the test-taker's brain: capacity delivers performance, performance indicates capacity, and thus it seems that both general intelligence ability and measurable intelligence could serve equivalently well as an indicator of the brain's effectiveness.

But alas, there are known scenarios where the equivalency between general intelligence ability and measurable intelligence does not appear to hold. Some of these scenarios are well understood by intelligence researchers, have reasonable explanations, and can be adjusted for. For instance, if you give a standard intelligence test to a pool of ten-year-olds and then give that very same test to a pool of adults, the ten-year-olds as a group will demonstrate lower raw scores and thus less measurable intelligence than do the adults. This means that if one were to insist on the equivalency of general intelligence ability and measurable intelligence, then one would also have to say that the ten-years-olds as a group possess less general intelligence ability than do the adults. But this is inconvenient for a number of reasons, not the least of which is that it violates the intuition that general intelligence ability should be a stable and consistent characteristic across an individual's lifetime. This intuition can be restored by limiting comparison of ten-year-olds only to other ten-year-olds, and not mixing them in with the adults. Indeed, intelligence researchers have become quite aware that their findings regarding general intelligence ability are prone to being influenced by both the nature of the tests being employed as well as by the characteristics of the population being tested (Detterman & Daniel, 1989). Therefore, when researchers want to make broad statements about general intelligence ability within the human population, they rely upon test scenarios in which a broad battery of tests is used (such as Stanford-Binet or Wechsler) and in which a broadly representative sample of the population is tested. It is only under such circumstances that one can safely inquire about the equivalency of general intelligence ability and measurable intelligence.

Nonetheless, even under these broadly representative test conditions, there are still scenarios in which the equivalency between general intelligence ability and measurable intelligence does not appear to hold, scenarios that do not have a ready explanation nor an easy adjustment. The most prominent of these scenarios arises from the Flynn effect.

To cleanly demonstrate the Flynn effect and to indicate how it plays havoc with the equivalency between general intelligence ability and measurable intelligence, I will employ a hypothetical scenario, and then afterwards discuss how realistic this scenario is. The scenario begins in the year 1925, at which time a broad battery of intelligence tests (similar to the current form of Stanford-Binet or Wechsler) is administered to ten thousand people drawn as a random sample from the adult human population. This population's average raw score on the battery of tests is 30%, with a standard deviation of 10%. These raw results are then normed and set out against the standard IQ scale, from which estimates are made of each individual's general intelligence ability. Studies are then conducted which result in the showing of a strong correlation between these estimated general intelligence abilities and academic performance, job success, and other life circumstances, and also, family-based studies are conducted which indicate that there is a strong genetic/neural component underlying these estimated general intelligence abilities. These are the same kinds of outcomes that have been repeated many times by real-world intelligence researchers ever since intelligence tests were invented—this is all standard fare.

Next, one hundred years later, in the year 2025, the very same test is administered to ten thousand people drawn as a random sample from the extant adult human population. This time, in exact one-to-one correspondence to the 1925 population, everyone doubles their raw score. This means that the average raw score is now 60%, with a standard deviation of 20%. Nonetheless, when these results are normed and applied against the standard IQ scale, and then used to estimate general intelligence abilities, everything turns out to be the same, because the relative and correlative comparisons are still equivalent. (The same is true for the g-loadings on the tests, and for every other intelligence statistic determined by relative values.) Also, when the life circumstance and family-based studies are performed for the 2025 population, the results turn out to be essentially identical to those for the 1925 population—general intelligence ability still predicts life success, general intelligence ability still shows a strong genetic and neurological underpinning, and so on. When it comes to general intelligence ability, these researchers are not going to be able to detect any appreciable difference between the 1925 and 2025 populations, an outcome very much in line with what they might have predicted. Evolutionarily and biologically speaking, one hundred years is a short period of time, too short to foster widespread genetic and neural changes. The genetic/neural characteristics of the 2025 population should be nearly identical to those of the 1925 population, and since general intelligence ability is determined primarily by these genetic/neural characteristics, scientists would have every reason to expect that general intelligence ability across these two populations has not changed appreciably over time, exactly as the results appear to confirm.

But the measurable intelligence for these two populations has very much changed over time. An average raw score increase from 30% to 60% demonstrates a significant increase in measurable intelligence (and this is not a problem with calibration, since the test is exactly the same). Intelligence researchers now face a dilemma. If they wish to maintain the connection of general intelligence ability to measurable intelligence, then this increase in measurable intelligence should translate to a corresponding increase in general intelligence ability, something that neither their studies nor their expectations would indicate. But if they disconnect measurable intelligence from general intelligence ability, this then creates a contradiction for any brain-based explanation of human intelligence. The fact that general intelligence ability has remained stable from 1925 to 2025 would indicate that brain effectiveness has also remained stable within the human population over that time. And the fact that measurable intelligence has significantly increased from 1925 to 2025 would indicate that brain effectiveness has correspondingly increased within the human population over that time. But it is not possible for both of these statements to be true. You could ask the question directly—has brain effectiveness increased for humans over these one hundred years, or has it not? Either answer is going to cause a problem.

Just how realistic is this hypothetical scenario? As it turns out, it is extremely realistic. The first modern intelligence exams were developed in the early twentieth century, and as that century progressed, these exams were regularly updated and expanded. For calibration and comparison purposes, when a new version of an exam was developed, researchers usually administered both the old and the new version to a representative sample from the testing population. This process created a connected time series of comparable results, and it began to be noticed that later generations of test-takers were always demonstrating significantly better raw performance on these exams than had the earlier generations. Several researchers noted this phenomenon, but it was James Flynn in the 1980s who published two influential papers (Flynn, 1984, 1987) which demonstrated that the phenomenon was extremely widespread, thereby drawing greater attention to it, and in the decades which followed, many studies were published verifying the nearly ubiquitous pattern of increasing raw intelligence scores (Wongupparaj et al., 2023). It was during this time that the phenomenon began to be called the Flynn effect.

The real-world Flynn effect data is of course not quite as neat and clean as the perfect doubling of raw scores in the hypothetical scenario. Nonetheless, the overall difference in raw intelligence scores from the early twentieth century to the early twenty-first century is significantly large, far too large to ignore and far too large to support a simple equivalency between general intelligence ability and measurable intelligence. Researchers were puzzled when they first realized the magnitude of the Flynn effect, and they are still puzzled today (Rodgers, 2023).

There is another important scenario that casts significant doubt on any equivalent connection between general intelligence ability and measurable intelligence. This scenario arises from what has been stated previously about wild animals and early humans (pre behavioral modernity) demonstrating no measurable intelligence. The question presents itself, do these creatures also have no general intelligence ability? What we will come to realize is that there is a great deal of reason and evidence to suggest just the opposite, to suggest that these wild animals and early humans should be recognized as possessing a positive level of general intelligence ability. But if this is indeed true, then what must it say about the concept of intelligence overall, knowing that a species can both possess general intelligence ability and at the same time demonstrate no measurable intelligence?

In the previous section it was asserted that early humans once displayed no measurable intelligence. What this means is that there would have been no intelligence exam simple enough to register a score for these early humans (or even simple enough to administer). One might imagine trying to create such a test, for instance by starting with Stanford-Binet or Wechsler and then simplifying it, even making use of calibration to track how much simplicity was being introduced. But just how easy would one have to make such an exam? It could not be written, it could not contain arithmetic, or abstract vocabulary, or geometric patterns, or digit recall, or just about anything that shows up on a modern exam. By the time one had gone through several rounds of such simplification, would there be anything left at all, and would the calibrating test-takers find themselves challenged to any degree? At some point, it would just be easier to say that this population of early humans must be demonstrating a zero level of measurable intelligence, the same that could be said of any of the wild animals.

But does zero measurable intelligence also imply zero general intelligence ability? True, one cannot calculate general intelligence ability without measurable intelligence, but that is not quite the same thing as saying that general intelligence ability does not exist.

One of the most established findings of intelligence science is that general intelligence ability is determined primarily by one's genetic and neural makeup, and furthermore it is not just a small number of genes and a small number of neurons that make this determination—there appear to be many genes and many different sections of the brain involved in general intelligence ability (Toga & Thompson, 2005). For large and representative samples from the human population, the principles of biology and evolution would preclude massive changes to these genetic and neural structures over anything but an extremely long period of time. We have already noted that in going from the year 2025 to the year 1925, one would not expect significant changes to the population's overall genetic and neural makeup, but is it really any different if one goes back a couple hundred thousand years? The Homo sapiens population from that earlier era should have had a similar genetic/neural makeup to the humans of today, including nearly all the genetic and neural structures now being indicated as influential in general intelligence ability. And if those genetic and neural structures can produce general intelligence ability in modern humans, why would they not have been capable of producing general intelligence ability in the humans of an earlier era? Biologically speaking, scientists have every reason to expect that early humans possessed a significant level of general intelligence ability, even during that time period in which they were displaying no measurable intelligence.

Further bolstering this notion that early humans must have possessed a significant level of general intelligence ability is the consideration of what would have taken place shortly after human behavioral modernity began. The early stages of human behavioral modernity would have been marked by some telling changes in human circumstances—for instance, control of fire, introduction of clothing, structured tools and weapons, ornamental jewelry, cave paintings, abstract vocabulary to refer to these new features, and so on (Klein, 2002). These new circumstances would have had the effect of making an intelligence exam conceivable for this population, because now there would have been features to test against other than just survival-and-procreative demand. True, such an exam would have been extraordinarily simple by modern standards, but nonetheless, it could have served a familiar purpose. If such an exam had been available for that early population, some of its members could have been expected to perform well, some could have been expected to do poorly, with the overall population evincing something like a normal distribution of performance. And if those ranked performances could have been matched against early-era life circumstances, a strong correlation might have been anticipated. That is to say, this early population would have been capable of demonstrating the usual evidence of a positive general intelligence ability. And if the human population possessed general intelligence ability right after the turn towards behavioral modernity, then why would it not have possessed general intelligence ability right before the turn towards behavioral modernity?

Finally, there is the evidence of the wild animals and their tamed, caged and domesticated brethren. Intelligence researchers have been able to show that under the right circumstances, tamed, caged and domesticated animals can demonstrate a degree of measurable intelligence (Shaw & Schmelz, 2017), and although this measurable intelligence is likely tainted with human influence, its existence nonetheless indicates a positive level of general intelligence ability within these animals. But the wild version of these animals are genetically the same, and therefore should have the same level of general intelligence ability as do the tamed, caged and domesticated version, even though the wild version of these animals is not capable of demonstrating any measurable intelligence. Again we have a situation—much like with the Flynn effect and much like with the case of early humans—in which any direct connection between general intelligence ability and measurable intelligence appears to be fundamentally broken.

I hope these scenarios do not come across as technical nitpicking, because this fundamental distinction between general intelligence ability and measurable intelligence is what lies at the heart of a more thorough understanding of human intelligence. Of these two concepts, general intelligence ability is by far the most studied and the best understood, and therefore the most influential—it has been the main pursuit of intelligence science for more than one hundred years, and it has served as the cornerstone of nearly every meaningful intelligence finding to date. When scientists speak of intelligence it is mostly through reliance upon this notion of a general intelligence ability, along with its established connection to genetic signatures and neural structures. Thus, it cannot be all that surprising that scientists think of intelligence as being exclusively a brain-based phenomenon.

In contrast, very little attention has been paid to measurable intelligence. Other than the Flynn effect, the calibrated raw scores on intelligence exams have been mostly ignored. I would claim that this neglect is an unfortunate mistake. In the scientific definition of intelligence (that is, intelligence is what gets measured by intelligence tests), measurable intelligence is the direct reference, not general intelligence ability. General intelligence ability is a derived concept, measurable intelligence is the thing itself. And what this section has both demonstrated and emphasized is that general intelligence ability and measurable intelligence are two very different things—not just conceptually different, but also different quantitatively. We have at least the following three scenarios to describe the uneven relationship between general intelligence ability and measurable intelligence:

• Sometimes, when measurable intelligence increases [decreases], so does general intelligence ability, and vice versa (typical testing scenario).
• Sometimes, when general intelligence ability remains constant, measurable intelligence increases (the Flynn effect).
• Sometimes, when general intelligence ability is positive, measurable intelligence can be zero (early humans and wild animals).

Whatever measurable intelligence is, it is fundamentally not the same thing as general intelligence ability, meaning among other things that measurable intelligence is not necessarily a brain-specific phenomenon. Something else is going on here, something that has been overlooked by intelligence researchers to date, and thus we need to arrive at a more thorough understanding of what exactly is being indicated by the readings of measurable intelligence.

The Flynn Effect

The Flynn effect can be described in very few words: measurable intelligence, indicated by the calibrated raw scores on intelligence tests, increases over time. This description links directly to the scientific definition of intelligence, so one might expect that the Flynn effect must be a very big deal within the realm of intelligence science. And yet this is not exactly the case. For instance, in the book The Science of Human Intelligence (Haier et al., 2023)—possibly the most thorough presentation of the current state of intelligence research—the Flynn effect barely gets a mention. It is almost as if it is being swept under the rug. Intelligence researchers are aware of the Flynn effect, but they do not know what to make of it.

The widespread perplexity surrounding the Flynn effect is evidenced most obviously by the large number of causes that have been suggested for its existence, none of which have proven to be compelling: greater hybrid vigor, better nutrition, expanded education, increased exposure to IQ exams, increased exposure to video games and graphical puzzles, the rise of science, a decrease in infectious diseases, etc.—along with various combinations of all of the above. Everyone has a favorite explanation, and yet nowhere is there anything like a consensus. In addition, several complex models have been developed to plot and to predict the Flynn effect patterns, models geared mostly towards matching various environmental factors against the long-term influences underlying general intelligence ability. The Dickens-Flynn model (Dickens & Flynn, 2001) and Woodley's theory of fast and slow life (Woodley, 2012) are two of the more well known of these complex models—each has its proponents but neither has been convincing.

Researcher Linda Gottfredson, in responding to James Flynn's book What Is Intelligence?, once offered the following statement:

The chief riddle posed by the Flynn Effect is this: How can something so heritable as IQ change so quickly from one generation to the next? To my mind, this paradox signals that we have yet to learn something fundamental about intelligence or current measures of it (Gottfredson, 2007).

That is by far the most insightful statement I have ever seen from an intelligence researcher regarding the Flynn effect—it points the way. If I might be allowed to rephrase it somewhat, what it is suggesting is the following: to successfully explain the Flynn effect, it is going to be necessary to reconceptualize human intelligence. Unfortunately, no researcher—including Gottfredson herself—seems to have taken her up on the offer. Every suggested cause and every complex model employed to explain the Flynn effect to date has its roots in the standard description of human intelligence. And for this reason, every suggested cause and every complex model is wrong.

There are two telltale signs that indicate that all the thinking surrounding the Flynn effect has been heavily influenced by the standard description of human intelligence. The first such sign is the widespread belief and insistence that the Flynn effect must be temporary. Intelligence tests did not appear until the early twentieth century, and so there is no actual hard data to indicate whether or not the Flynn effect would have been operative before the year 1900. And yet despite the fact that it would seem the default choice would be to assume that the twentieth century was a continuation of what had come before, the consensus within the intelligence research community appears to be strongly in favor of the Flynn effect being only a twentieth-century phenomenon—a twentieth-century aberration, if you will. Only a couple researchers, such as James Flynn himself, have suggested that the Flynn effect could trace its origin to somewhat earlier days, for instance to the time of the Industrial and Scientific Revolutions (Flynn, 2007; van der Linden & Borsboom, 2019). But I am unaware of any researcher who has suggested that the Flynn effect might have existed any earlier than that.

On the other side of the calendar, the latest trend in intelligence research has been the diligent hunt for evidence that in the early twenty-first century, the Flynn effect is now ending or reversing. Several studies have been published to this effect (Dutton et al., 2016; Pietschnig & Gittler, 2015), claiming that raw intelligence scores have started a downwards trend, and although there have also appeared other studies indicating that the Flynn effect is still continuing (Colom et al., 2023; Liu & Lynn, 2013; Nijenhuis et al., 2012), most of the fanfare has come down on the side of the Flynn effect's imminent demise. Within the intelligence research community, both the belief and the hope seem to be that the Flynn effect cannot go on forever.

It has already been hinted at why the standard description of human intelligence requires that the Flynn effect be temporary. The standard description, based almost entirely upon the concept of general intelligence ability and depicting human intelligence as a brain-specific and brain-produced phenomenon, is thereby bound to the known principles of biology and evolution. As a biological characteristic, especially one with such a broad neural and genetic basis, human intelligence must be anticipated as remaining fairly stable over time. If, for instance, the average human body were to begin doubling in size and weight every century or two, scientists would find themselves utterly dumbfounded, since such a population-wide growth in physical dimensions would violate every known biological and evolutionary principle. But that is exactly what appears to be happening with human intelligence, creating an ongoing conundrum, one that would be challenging enough to explain across the span of a century, and essentially impossible to explain across any time period much longer than that.

Therefore, the solutions usually offered to explain the Flynn effect are presented as short-term boosts to overall human intelligence—twentieth-century boosts. Greater hybrid vigor, better nutrition, expanded education, increased exposure to IQ exams, increased exposure to video games and graphical puzzles, the rise of science, a decrease in infectious diseases, etc.—all these are generally put forth as twentieth-century surges in human experience, not operative during earlier times and soon to run their effective course, covering the known range of the Flynn effect data. Furthermore, the complex models employed to explain the Flynn effect, such as those of Dickens-Flynn and Woodley, these are constructed in such a way as to allow human intelligence to increase, to decrease, and also to remain stable over time, a remarkably flexible feat accomplished mostly through the use of a large number of parameters within the model, all done with the stated intention of allowing intelligence to exhibit an occasional short-term variation while at the same time maintaining its underlying long-term stability. It would seem that whatever explanation is being offered, the one characteristic that the Flynn effect cannot be allowed to have is to be enduring.

But human history gives every reason to expect that the Flynn effect is indeed enduring, and has been a part of the human species ever since the turn towards behavioral modernity. We have already noted that there would have been no intelligence exam simple enough to administer to humans before their behavioral turn, indicating a zero level of measurable intelligence for that population (the same level of measurable intelligence witnessed in wild animals today). And even after the behavioral turn, although intelligence exams would have then become conceivable, they would have had to remain extraordinarily simple for quite some time, certainly much simpler than any of the modern forms of Stanford-Binet and Wechsler. Such early exams could not have been written, they could not have contained arithmetic, nor complex geometrical patterns, nor any of the broad vocabulary that paints the modern world of today, etc. Almost everything that currently shows up on a modern intelligence exam represents a concept or artifact that was introduced into the human species over the course of the last several thousand years, meaning that human intelligence has been increasing steadily and continuously from its initial level of zero a couple hundred thousand years ago to the far more substantial level that is witnessed today. Judging by what we can see of the progress of human history—from the days of being purely hunter-gatherers, to the days of being budding agrarians, to the days of being civilization builders, to the days of being architects of towering skyscrapers—and judging by what would have been available for inclusion on an intelligence exam across all that period of time, one would have to say that measurable intelligence within the human population has been steadily increasing ever since the turn towards behavioral modernity. Therefore, the Flynn effect has been anything but temporary.

Furthermore, these considerations alone should be reason enough to doubt any suggestion that the Flynn effect is currently ending or reversing. What an incredible coincidence it would be to find a human phenomenon that has been in existence for tens of thousands of years, and then just now, right at the very moment of its conscious discovery, that phenomenon screeches to an abrupt halt. That would make no logical sense at all. But more than just this, if the twentieth century is to have taught us anything, it is that an increase in human intelligence was more than just an observed phenomenon over those one hundred years, it must have also been a necessity. Consider the long list of new features and new experiences that the twentieth century brought to nearly the entire swath of the human population—the speed of automobile travel, the reach of airplane flights, the ubiquity of electronic communication, the complexity and hustle-bustle of modern cities, the algorithmic turns of digital computers, the widespread connectedness of the Internet—all this and much much more. Would it have been conceivable for humans to have absorbed and to have mastered these new features and new experiences, while at the same time becoming less intelligent? That population would surely have been overwhelmed by all these new constructions and all these new life requirements if it had not somehow kept pace cognitively, as indeed it did. But if these statements are true about the twentieth century, why would we expect the twenty-first century to turn out any different? There are a host of challenging new innovations now appearing on the human horizon—virtual reality, machine learning, bioengineering, robotics, not to mention discoveries yet to be made. Is it really feasible to think that humanity will somehow navigate these new artifacts and these new experiences while at the same time moving backwards in intelligence? Does that really make any sense at all?

Listen, I am willing to give science its time to work. If enough solid evidence accumulates that measurable intelligence is indeed plateauing or reversing, I will gladly surrender every claim made within this essay. But you cannot expect me to hold my breath. There is no logical reason and no historical basis to believe that the Flynn effect is temporary. What there actually is, is a desire to maintain the sanctity of the standard description of human intelligence. What I believe is currently taking place within the academic community is something like the following. After forty some years of consistent Flynn effect data, still more reports of an increase in raw performance on intelligence exams is no longer interesting news—both journal editors and the academic press have seen this story many times before. But in their desire for something fresh, and in their desire to defend the standard description of human intelligence, both journal editors and the academic press quickly perk up their ears at any hint that the Flynn effect might be ending or reversing. If twenty studies are offered indicating that the Flynn effect is continuing, and just one study is offered suggesting that the Flynn effect has ended, guess which one is going to get published, and guess which one is going to make the news. I believe we are in the middle of a scholarly fad, the fad of the Flynn effect's imminent demise, but since there is no logical reason and no historical basis to believe that the Flynn effect is temporary, I am convinced this fad will soon run its unfortunate course. I have no difficulty making the following claim: the Flynn effect is a fundamental and enduring component of human intelligence, has been with humanity ever since the turn towards behavioral modernity, and barring unusual circumstances (such as civilization collapse), the Flynn effect will continue to be with humanity into the foreseeable future. The Flynn effect is not temporary.

If I might be allowed to pause the discussion for a moment, I would like to make the suggestion that the confusion surrounding the Flynn effect's permanence or transience is actually a result of the confusion surrounding general intelligence ability and measurable intelligence. As has been suggested already within this essay, and as will be discussed in some detail later on, there is very good reason to make the assumption that general intelligence ability is a stable component of human intelligence, and that it exists at nearly the same levels today as it did during the time of the human behavioral turn. Therefore, when researchers speak of intelligence being stable over time, it is likely of general intelligence ability that they are thinking, along with its established connection to genetics and to the human brain. I have no actual argument with this point of view, as long as it remains specific, as long as it remains restricted to the domain of general intelligence ability. But general intelligence ability is not the same thing as measurable intelligence, and the Flynn effect is not an outcome of general intelligence ability, it is an outcome of measurable intelligence. The mismatch between these two concepts, between general intelligence ability and measurable intelligence, this is what is producing the "paradox" of Gottfredson's statement, this is the "something fundamental" that researchers have yet to learn about intelligence and current measures of it.

The second telltale sign that the thinking surrounding the Flynn effect is being heavily influenced by the standard description of human intelligence is that every suggested explanation is offered as a solution that impacts the human brain. In the standard description, brain effectiveness is everything—any increase in an intelligence measure is assumed ipso facto to translate to a corresponding increase in brain efficiency and brain effectiveness (Langer et al., 2012). For some of the suggested Flynn effect solutions, this connection is obvious—greater hybrid vigor and better nutrition, for instance, these are chosen for their presumed organic consequence upon neurons, synapses and biochemical activity. In other instances, the impact might be described as being a little more indirect—expanded education and increased exposure to video games and graphical puzzles, for instance, here one might point to an augmentation in the brain's stored knowledge or to a vigorous exercising of the brain's working memory or something of the like. But direct or indirect, if you dig deeply enough into any of the offered Flynn effect explanations, what you will ultimately find is an insistence that the explanation results in an increase to the brains's effectiveness, the only way within the standard description that intelligence can be enhanced (Marsman et al., 2017).

But this approach creates a logical conflict—a conflict with general intelligence ability. General intelligence ability, believed to be the product of many different genes and understood as being hosted across a multitude of neural regions, already addresses the notion of general brain effectiveness (Colom et al., 2006; Haier et al., 2004). So when an additional boost to brain effectiveness is proposed, the question naturally arises, does this additional boost to brain effectiveness also impact general intelligence ability? The prevailing evidence would appear to say that the answer to this question is no, that the Flynn effect does not represent an increase to general intelligence ability (te Nijenhuis & van der Flier, 2013). But then biologically speaking, how does the proposed additional boost accomplish such a selective feat, increasing the brain's effectiveness overall but not impacting those many brain structures involved with general intelligence ability? Take better nutrition, as an example. How does better nutrition impact the neurons, synapses and biochemical activity that lead to the increase in intelligence corresponding to the Flynn effect, while at the same time somehow managing to avoid all those brain features implicated in general intelligence ability? It sounds too organically magical. The alternative of course is to assume that the Flynn effect does represent an increase in general intelligence ability, but for this there is no actual evidence and there is no other reason to support the conclusion. No matter which way one turns, any suggested additional increase to brain effectiveness inevitably creates a double-edged sword with general intelligence ability.

To help untangle this confusion, let us recall the three scenarios that describe the uneven relationship between general intelligence ability and measurable intelligence:

• Sometimes, when measurable intelligence increases [decreases], so does general intelligence ability, and vice versa (typical testing scenario).
• Sometimes, when general intelligence ability remains constant, measurable intelligence increases (the Flynn effect).
• Sometimes, when general intelligence ability is positive, measurable intelligence can be zero (early humans and wild animals).

If we wish to describe mathematically the relationship between general intelligence ability and measurable intelligence, the three scenarios above will rule out certain combinations. For instance, the following expression of equivalency cannot possibly be true:

measurable intelligence = general intelligence ability

This equation would work well for the first scenario, since within it general intelligence ability and measurable intelligence move up and down in tandem. But the equation cannot be made to fit either the second or the third scenario, including of course the scenario of the Flynn effect.

But with a slight adjustment, we can arrive at something that is a little more promising. Consider, for example, the following equation:

measurable intelligence = general intelligence ability x _________________

The blank represents a yet-to-be-named factor, one that can mediate the relationship between general intelligence ability and measurable intelligence. And if that unnamed factor is assumed to be independent of (orthogonal to) general intelligence ability, then all three scenarios can be easily accommodated:

• When the unnamed factor is constant, general intelligence ability and measurable intelligence will move up and down in tandem, covering the first scenario.
• When general intelligence ability is constant, measurable intelligence will move in tandem with the unnamed factor, and in particular, if the unnamed factor increases over time, measurable intelligence will also increase over time, the scenario of the Flynn effect.
• When the unnamed factor is zero, measurable intelligence will be zero as well, including those instances when general intelligence ability is positive, the presumed scenario of early humans and wild animals.

The requirement of independence between general intelligence ability and the as-yet-to-be-named factor is important. Without independence, the interaction between general intelligence ability and the unnamed factor would create a strong second-order effect, destroying all the mathematical simplicity that the equation contains and conflicting with both the reasoning and the evidence that suggests general intelligence ability remains stable over time. But this requirement of independence also rules out every suggested solution that has ever been made for explaining the Flynn effect. Greater hybrid vigor, better nutrition, expanded education, etc.—all these choices can be ruled out as candidates for the unnamed factor, because all these choices have been put forth as boosts to brain effectiveness, and you cannot boost brain effectiveness and at the same time remain independent of general intelligence ability. The standard description of human intelligence is actually getting in the way, forcing researchers into solutions that necessarily encompass the human brain, when what is actually needed is a solution that does not touch the brain at all. The unnamed factor, whatever it happens to be, should have no influence on neural operations, allowing general intelligence ability to continue all its functioning undisturbed.

Therefore, what we are seeking, using the Flynn effect as our primary guide, is an influence on human intelligence that increases over time, is not temporary, and does not impact the operations of the human brain. What we are seeking will turn out to be the key ingredient to a reconceptualization of human intelligence.

Artificial Construction and the Content of Intelligence Tests

There are two main indicators of the identity of the factor connecting general intelligence ability to measurable intelligence. The first indicator is the material description of human history, beginning with the circumstances of the crossover to human behavioral modernity, and continuing across the entire range of tangible alteration that has brought humanity to the circumstances of the modern world today. The second indicator is the content of intelligence tests.

Prior to their turn towards behavioral modernity, humans were living the lives of pure animals, which is to say they were living in an entirely natural setting and were experiencing only those perceptions and behaviors critical to survival and procreation (Klein, 2009). There was essentially nothing artificial or constructed within these ancient human surroundings, the same as can be witnessed in the lived environment of wild animals today. And as has been noted, there would have been no intelligence exam simple enough to be constructed for or administered to these early humans, indicating a zero level of measurable intelligence for that entire population. Indeed, if one were trying to predict the success or failure in life circumstances of these early humans, an intelligence exam would have been of no help at all. All that mattered to these early humans was their survival-and-procreative ability, and thus, to predict their life outcomes one would have needed to have measured for their strength, speed, fertility, sense acuity, attractiveness, and so on, qualities that of course do not make an appearance on modern intelligence exams. For these early humans, intelligence meant nothing at all—everything hinged entirely upon the immediate and natural needs of survival and procreation, the same as has been true of wild animals throughout the many years, and the same as is true of wild animals today.

The crossover from those ancient circumstances to the beginnings of human behavioral modernity was a pivotal event (Christian, 2018), one that had not been experienced by any other species upon this planet, and one that would mark the start of human intelligence. Many scientists are wont to explain this event with biological causes, such as the presumed emergence of a language gene, or perhaps through an impactful alteration in neural structure (Bradshaw, 2002; Graham & Fisher, 2013). Unfortunately, such explanations remain almost entirely unspecified—no language gene has ever been successfully identified, and no concrete alteration in neural structure has ever been explicitly depicted. Furthermore, no plausible description has ever been offered as to how these proposed biological changes could have managed to spread so quickly to essentially one hundred percent of the human population. Once again, these biological explanations begin to sound too organically magical. What I prefer is an explanation that is not biological at all, and I prefer it because it possesses a multitude of specific evidence, evidence existing right there before our very eyes.

What has never been in dispute about that initial crossover to human behavioral modernity is that it was accompanied by a dramatic and tangible change to the circumstances of the human surroundings (Henshilwood & Marean, 2003; Klein, 2002; Sterelny, 2011). Within a lived environment that had been previously entirely natural, there now began to appear many features of a much different kind: fire pits, animal skin clothing, structured tools and weapons, ornamental jewelry, cave paintings, etc. None of these new features were natural, and none of these new features had anything in common with the circumstances of the other creatures of the animal kingdom. These were the first examples of something artificial appearing within the human surroundings, and these were the first instances of the human environment becoming constructed. It is not possible to overstate the importance of this occasion, this event was truly momentous. Artificial construction—the palpable evidence of the turn towards behavioral modernity, and also the key ingredient to a reconceptualization of human intelligence—it had now arrived upon the human scene. And this was only its beginning.

The tale of modern human history is first and foremost a tale of accumulating and increasingly sophisticated artificial construction (Christian, 2014). From the days of the Upper Paleolithic, with its introduction of bone tools, crafted weapons, and carved intricacies, opening pathways to humanity's ever more widening range; to the days of an agrarian awakening, and its explosion of new artifacts into the human environment—huts, pottery wheels, irrigation trenches, chisels, burial tombs; to the days of the great civilizations, and their remarkable recasting of entire landscapes—aqueducts, monuments, roads, carts, ships and more; to the days of the Scientific and Industrial Revolutions, and their dense cluttering of the now crowded human scene—cathedrals, factories, laboratories, tractors, houses arranged upon row after row after row; to the days of the early twenty-first century, in which nature has been almost entirely eclipsed from human view, eclipsed everywhere by automobiles, highways, towering skyscrapers, books, computers, televisions—the list goes on and on. From surroundings that were once entirely natural, humans now live in an ocean of constructed artificiality, its evidence exists literally every direction one looks. And if one were seeking for a human characteristic that could be relied upon to increase significantly, indeed exponentially, over a long period of time, one could do no better than to turn to the unrelenting and ubiquitous growth of artificial construction.

Plus it has been more than just the features themselves that have marked the human transformation, because artificial construction has also thoroughly altered almost every human perception and behavior (Rączaszek-Leonardi et al., 2019). Each new artifact changes something about what humans apprehend and do: clothing alters where humans live, controlled fire alters what humans eat, structured weapons alter what humans hunt, structured tools alter what humans build, and so on. The entirety of the human transformation rests upon this foundation of artificial construction, and if you do not believe me, then ask yourself what would happen if we were to remove every artificial feature that now exists within the human environment, and if we were to suppress every human behavior that owes its origin to those removed artifacts. Ask yourself, what would then remain? All that would then remain would be the purely natural environment in which humans once used to live, and all that would then remain would be the purely survival-and-procreative behaviors that humans once used to display. All that would then remain would be the pure animals that humans once used to be. Modern humanity is, above all else, a product of its artificial construction.

In the present world, the impact of artificial construction upon human lives has become so pervasive as to be almost overlooked. There are now very few instances of humans engaging in purely survival-and-procreative behaviors, but there are now countless instances of humans engaging in the unnatural behaviors that have been induced by a multitude of unnatural features: we drive because there are cars on the street, we read because there are books on the shelf, we shave because there are razors in the cabinet, etc. For humans, successfully navigating the modern world means successfully navigating an entire surrounding ocean of artificial construction, and those who master that ubiquitous world of artificial construction, they are the ones who do well, and those who do not master that ubiquitous world of artificial construction, they are the ones who do poorly. And what characteristic is it that determines how successful an individual can be in navigating and mastering all this artificial construction? We may not be accustomed to employing the word in just quite this fashion, but the successful navigation and mastery of the surrounding multitude of artificial construction is simply an alternative definition of the word intelligence.

In the previous section, it was suggested that the uneven relationship between general intelligence ability and measurable intelligence could be captured by the following equation:

measurable intelligence = general intelligence ability x _________________

To make this equation conform to the circumstances of human history, and to allow this equation to work with the known data regarding human performance on intelligence exams, including especially the Flynn effect, the blank needs to be filled in by the totality of artificial construction to be found within the human environment. Artificial construction is the feature that connects an individual's general intelligence ability to that individual's specific raw performance on an intelligence exam.

measurable intelligence = general intelligence ability x artificial construction

At any given moment in time and for a particular testing population, the amount of artificial construction to which that population has been exposed is going to be roughly the same (this is particularly true in today's world, where widespread travel and electronic communication exposes nearly every aspect of the artificial environment to nearly every member of the human population). If the amount of artificial construction is roughly the same for all the test-takers, then it can be seen by the equation above that their relative raw performance on intelligence exams is going to be determined solely by their relative levels of general intelligence ability—those with more general intelligence ability will absorb a greater amount of the artificial construction around them and will display more measurable intelligence, and those with less general intelligence ability will absorb a lesser amount of the artificial construction around them and will display less measurable intelligence, the typical testing scenario. But this only works at a given moment in time, because as time progresses the amount of artificial construction continues to grow, increasing its factored contribution to measurable intelligence. This is why earlier test-takers, such as those from the early twentieth century, end up displaying considerably less measurable intelligence than do later test-takers, such as those from the late twentieth century, even though each population possesses roughly the same average level of general intelligence ability. It is the changing amount of artificial construction to which these populations are exposed that accounts for their overall difference in measurable intelligence.

Before the turn towards behavioral modernity, there would have been no artificial construction to be found within the human environment, and although it could be surmised that humans possessed a positive level of general intelligence ability during that time, since there was no artificial construction against which to apply (to multiply, via the equation) this ability, the result would have been a zero level of measurable intelligence. After the turn towards behavioral modernity, with its budding appearance of artificial construction, humans would have then had features to apply (to multiply) their general intelligence ability against, resulting in a beginning level of measurable intelligence. And as human history has progressed, from the days of the hunters-gatherers through the days of the early twenty-first century, the amount of artificial construction contained within the human environment has continued to increase in both size and complexity over all that period of time. And assuming that general intelligence ability has remained fairly stable within the population, this continuous increase in artificial construction will have produced a corresponding increase in measurable intelligence, the precise description of the Flynn effect. Note that this process would have been entirely observable throughout the twentieth century, a century in which artificial construction was increasing by leaps and bounds—automobiles, airplanes, electronic communication, sprawling cities, computers, etc.—and a century in which the raw scores on intelligence tests were also increasing by leaps and bounds. Even though general intelligence ability remains stable within the human population, the increase in artificial construction drives a corresponding increase in measurable intelligence, meaning that the changing amount of artificial construction contained within the human environment is the sole driver and the sole explanation of the Flynn effect.

This analysis also buttresses the claim that there is no reason to expect that the Flynn effect is ending or reversing. In the twenty-first century, the amount and complexity of artificial construction contained within the human environment continues to grow at a significant pace, just as it has during all the previous centuries (Nijs, 2015). Future generations, if they are going to be successful in their everyday lives, will be obliged to navigate and to master all this newly created artificial construction (as well as continue to navigate and to master all the previously created artificial construction). Through this accumulative process, future generations will display knowledge and skills that translate into higher levels of measurable intelligence, significantly surpassing the raw intelligence scores of all the previous generations. As long as the amount and complexity of artificial construction continues to grow within the human environment, the Flynn effect will not end.

The second indicator that artificial construction serves as the factor connecting general intelligence ability to measurable intelligence is the distinctive content of intelligence exams. There is a certain irony to the fact that a wide range of questions and challenges can serve as appropriate content for an IQ test—vocabulary queries, game puzzles, arithmetic, traffic sign recognition, sports trivia, etc.—all these items, and many more, can adequately serve the purpose of differentiating individual intelligence abilities. All these items conform to the positive manifold of human intelligence (Kovacs & Conway, 2016). This great variety of productive content for intelligence tests might lead one to think that just about any type of challenge could serve as a possible candidate for inclusion on an IQ exam, and indeed some intelligence researchers have made the leap to claiming that the content of intelligence tests does not matter much at all, that just about any collection of questions will do. But this is absolutely not the case.

There are entire categories of queries and challenges that have never made an appearance on an intelligence exam, and never will. One's capacity to beget children, for instance. One's digestive rate. One's forestalling of bacterial disease. One's running speed, lifting strength, throwing accuracy, etc. These are all important human attributes, they could make a difference in the quality and success of a person's life, and yet they remain entirely out-of-bounds for an intelligence exam. But why is this the case? One could point to the likelihood that performance on such challenges will not correlate to performance on items that traditionally appear on intelligence tests, but this is simply confirming a conclusion we already knew. The developers of the very first intelligence tests understood which types of challenges needed to be included, and which types of challenges needed to be excluded, and this was before they had done any psychometric analysis at all. We intuitively recognize that there is something about these alternative queries and challenges that is fundamentally incompatible with the types of queries and challenges that do show up on intelligence tests. We have just never bothered to make the distinction clear.

What should come to mind of course is that these alternative queries and challenges—such as digestive rate, running speed, etc.—these are derived exclusively from the natural characteristics of human beings, their origins are owed to the evolutionary and organic underpinnings of the species. As such, these qualities could have been tested and measured even before the turn towards human behavioral modernity; these characteristics are intimately connected to the species' long biological quest for survival and procreation. Anything which is natural, physical, biological, organic, it does not belong on an intelligence exam. And indeed if you inspect every single question on every existing intelligence test, you will not find the slightest hint of anything natural, physical, biological, organic.

So what then remains? What type of query or challenge is acceptable for inclusion on an intelligence exam? If everything natural, physical, biological, organic needs to be excluded from an intelligence test, then all that then remains is that which is artificial and constructed. The content of every single intelligence test is content that is exclusively derived from artificial construction. Words, digits, matrices, arithmetic, grammar, logic, geometric patterns, analogies, codes, connections, pictures, blocks, symbols, etc. Inspect every single question on every existing intelligence test, and all you will ever find are examples of artificial construction. All you will ever see are features that were synthetically introduced into the human species sometime after the turn towards behavioral modernity.

What this implies is that an intelligence test serves as a proxy. It serves as a proxy for the artificial construction that humans must navigate and master in their everyday lives. The same structure that underlies the multitude of artificial features that can be found in the modern world—structure identified by such concepts as grammar, logic, mathematics, etc.—that very same structure is recast into the form of artificial challenges and questions, providing the means for making a measurable and comparable assessment. Those individuals who can successfully navigate and master the artificial construction contained on an intelligence test are the same individuals who can successfully navigate and master the artificial construction ubiquitously contained in the human environment, making the correlation of IQ exam performance to general life circumstances little more than a tautology. The key is artificial construction—it serves as the linchpin between intelligence tests and the experienced world.

A major consequence of this realization that intelligence tests serve as a proxy for the artificial construction contained within the surrounding human environment is that intelligence tests cannot remain static over time. The amount and complexity of artificial construction contained within the surrounding environment is continuously increasing, and so to remain effective and valid as proxies, intelligence tests must be constantly reconstructed to contain greater variety and greater intricacy as time goes on. This is not news to the IQ test developers of the past century (Berger, 1986), and it will not be news to the IQ test developers of the coming century—intelligence tests have been, and will be, frequently modified, with the trend always being towards greater variety of content and greater difficulty. Think of all the additional artificial complexity that humans must now confront in their present world—greater demands to multitask, constant requirements to program "smart" machines, novel needs to decipher emojis and other addendums that have come with the widespread use of electronic communication, etc. If corresponding challenges are not added to the IQ tests, then the tests will eventually stop behaving as adequate proxies for the surrounding environment, and they will begin to fail in their purpose of differentiating individual intelligence abilities. This necessary fluidity to intelligence exam content is also made clear from the course of human history: a test such as Stanford-Binet or Wechsler would have completely overwhelmed, for instance, an ancient agrarian population, but a test appropriate for that ancient agrarian population would turn out to be too simplistic for the average twenty-first century human. The content of intelligence tests must correspond to the type of artificial construction contained within a testing population's lived environment, and because the amount and complexity of this artificial construction continues to increase over time, the sophistication of intelligence tests must be continuously boosted as well, just one more piece of evidence cementing artificial construction's role in the driving of the Flynn effect.

Finally, it should be noted that this emphasis on artificial construction provides an effective answer to Edwin Boring's warning about intelligence being too self-referencing. The developers of the very first intelligence tests were relying upon their intuition (Terman, 1916), looking to such terms as mental, reasoning, cognition, etc. to help guide them in their choice of questions and challenges, not quite recognizing that all these terms (just like the term intelligence) are intricately linked to the concept of artificial construction. But once those first tests had been developed, and once the positive manifold of human intelligence had been recognized, the future course was then set for all further test questions, via the correlation of their underlying connection to artificial construction. Without recognizing this role that artificial construction is playing in the process, intelligence exam development and psychometric analysis will indeed seem to be too self-referencing, but in truth the process is actually well grounded. It is grounded in the everyday lives and in the everyday endeavors of modern humans, grounded in their continuous interaction with an entire surrounding ocean of artificiality. When intelligence gets recognized and defined as the ability to navigate and to master artificial construction, intelligence stops being self-referencing.

Augmenting and Overturning the Standard Description of Human Intelligence

The standard description of human intelligence is not entirely wrong. Its popularity derives from its extremely successful depiction of general intelligence ability, a depiction that has been built upon more than one hundred years of detailed study regarding relative performance on human intelligence exams. In the equation measurable intelligence = general intelligence ability x artificial construction, general intelligence ability is one of the two critical components in the determination of measurable intelligence, and of these two components, it is the one that has been the most analyzed and the best understood. General intelligence ability explains individual intelligence differences; general intelligence ability predicts academic success, job performance, and an assortment of important life circumstances; general intelligence ability is expressed via neural characteristics; and general intelligence ability is determined primarily by genetic background. There is no doubting the validity and the importance of these many informative findings, and therefore, there is also no doubting the essential role that general intelligence ability plays in a thorough understanding of human intelligence behavior (Gottfredson, 1998).

The problem with the standard description—its fatal flaw—is that it places the entire burden of explaining intelligence on general intelligence ability alone, alongside its corresponding neural and genetic foundations. As alluring as this has been, it cannot be made to work. As we have already observed, the standard description's most glaring deficiency is its complete inability to account for the Flynn effect, an inability that becomes painfully obvious once it is recognized that the Flynn effect is not a product of general intelligence ability at all. This then has led to the bizarre hope within the intelligence research community, still desirous of maintaining the sanctity of its standard description, that the Flynn effect will prove to be little more than a temporary and unimportant phenomenon, a hope that has no logical, scientific, or historical basis.

But more than just its inability to explain the Flynn effect, the standard description suffers from an obvious difficulty that arises from its emphasis on intelligence being strictly a biological and brain-produced phenomenon. What this emphasis does is it forces an attribution of functionality onto the human brain that does not align with the primary task of neural systems. Neural systems within animal species have been well studied and are generally well understood, being most commonly and most accurately portrayed as stimulus-response mechanisms (Simmons & Young, 2010). These mechanisms have been in place for hundreds of millions of years and they have been evolutionarily honed to support basic survival-and-procreative demands. The human neural system is of course no different. Although very few humans today are raised and trained to live a predominantly survival-and-procreative existence, that biological capacity still remains. And therefore, if one wishes to discuss human neurons supporting stimulus-response behavior, then one would find oneself on very solid biological and evolutionary ground. But if on the other hand, one wishes to discuss human neurons producing and hosting intelligence, then one has now stepped into the realm of the biologically and evolutionarily extraordinary.

As the standard description of intelligence would appear to have it, around a couple hundred thousand years ago, the human brain, until then simply a component of a typical stimulus-response mechanism, began to take on the additional role of producing and hosting intelligence. Abstract language, calculation, pattern recognition, logic, etc.—all these characteristics that are commonly associated to intelligence, characteristics that had not been seen before within the entire animal kingdom, they somehow began to emerge as tangible and special features somewhere inside the hominin brain. It is these very same cerebral features that are the target of today's sophisticated neuroimaging studies, and it is these very same cerebral features that are depicted as being the palpable source of human intelligence (Jung & Haier, 2007).

But there are a myriad of troubling questions that arise from this widely held view. To begin with, how is it that these extraordinary cerebral features became so quickly and so thoroughly suffused throughout the entire human population? This question is made all the more challenging by the fact that intelligence is not associated to just a single neural characteristic, nor is it the consequence of a specific genetic cause—many neural regions and many genetic markers have been implicated in the expression of human intelligence ability (Basten et al., 2015; Plomin & von Stumm, 2018). So did all these regions and all these markers emerge within the population suddenly and together? Or perhaps instead, were most of them already fortuitously in place, just waiting for some final piece of the evolutionary puzzle to happen along? Either way, it begins to sound a bit convenient. Another troubling question is how does one account for the progression of intelligence? If during the days of the hunter-gatherers and the early agrarians, the human brain already contained the neural features necessary for the production of intelligence, then why was there no appearance during that time of algebra, calculus, formal logic, scientific method, polyphonic music, literature, and so on? If the brain was physically capable of such feats, and if the brain is supposedly the source of such feats, then why would there need to be any delay? For that matter, if there are going to be future intelligence advances—say, an important scientific understanding or an innovative computational technique—then why does the current brain not pop out that information right now, what could it possibly be waiting for? Perhaps you will say that the brain's journey towards intelligence is not altogether complete, that the brain is always in the process of becoming more effective generation after generation, that its capacity has been, and still is, organically advancing over time. But this then would require a biological explanation, would it not, a plausible mechanism to describe how such a generational and population-wide improvement is supposed to work, or are we again going to resort to some organic magic?

All of these concerns can be quickly dispelled by removing the insistence that human intelligence be regarded as strictly a biological, brain-produced phenomenon, and by giving instead equal weight to the influence that the totality of artificial construction has had upon the development and support of human intelligence. That is what is being conveyed by the equation measurable intelligence = general intelligence ability x artificial construction. Intelligence is not strictly a biological concept, nor is it strictly a non-biological concept. Measurable human intelligence is most accurately and most thoroughly portrayed as the orthogonal interaction of both biology and artifact, as the orthogonal interplay of general intelligence ability and the totality of artificial construction.

The key to understanding the workings behind this alternative description of human intelligence is to recognize that intelligence is not being housed somewhere inside the human skull. Instead, logic, reasoning, number, composition, calculation, rationality, grammar, etc., all these concepts commonly associated to human intelligence, they have a far more plausible and far more observable home right there in the structural features of the surrounding artificial environment, and there is no need to duplicate these features somewhere inside the human head. The tangible form of human intelligence exists literally all around us, and no neural depiction of intelligence can possibly be as elucidative as what we witness with our very own eyes. What neural instance of abstract language is going to be more informative than an actual conversation or the pages of a written book? What cerebral example of calculation is going to be more instructive than a performed integration? What cognitive bit of geometry is going to be more enlightening than a towering skyscraper? And what synaptic path of logic is going to be more compelling than a highway system, a legal brief, or a proof sketched out upon the chalkboard? Human intelligence first appeared in the form of the species' very first artifact, and human intelligence has continued to progress with the accruing size and connected complexity of the world's synthetic structure. Everything that needs to be understood about the material nature of human intelligence can be observed directly within the artificial environment, and to go seeking for that information somewhere inside neurons, synapses, and biochemical activity would be to engage in nothing more than a redundant folly. Corporeally speaking, artificial construction is human intelligence.

The role of the human brain in this alternative description of human intelligence is to respond to the stimulus of artificial construction, and in that sense, the human brain continues in its age-old purpose, as a stimulus-response mechanism. The stimulus has of course changed greatly from former hominin times, and therefore so has the response, but the neural mechanism remains essentially the same. Instead of perceiving predators, prey and sexual targets, today's human brain spends far more time discerning traffic, restaurants and the latest fashion. Instead of responding vigorously to rivals, the terrain and impending droughts, today's human brain reacts more robustly to computers, a building's layout and the evening news. Natural stimuli and organic responses produce what we call survival-and-procreative or biological behaviors, and artificial stimuli and constructed responses produce what we commonly call intelligence behaviors, but despite the differing labels, the supporting apparatus remains much the same. It could be surmised that if today's neuroscientists could be sent back in time a couple hundred thousand years, and could hook up their neuroimaging equipment to the human brains then engaged in purely survival-and-procreative activities, the neural patterns observed would be quite similar to what is witnessed in test subjects today, test subjects engaged in intelligence activities.

Therefore, it is dubious to speak of such things as cognitive modules within the human brain (Coltheart, 1999; Hilger et al., 2017). Instead of a language module, what there actually is is a language artifact within the artificial environment—a gesture, an uttered sound, a mark upon the page—and when the human brain perceives that artifact and responds meaningfully and productively to it, it produces what we call intelligent language behavior. Instead of a calculation module, what there actually is is a numerical or geometric artifact within the artificial environment—a grouping, a triangle, an algorithmic recipe—and when the human brain encounters that artifact and responds meaningfully and productively to it, it produces what we call intelligent mathematical behavior. Instead of a logic module, what there actually is is a formal artifact within the artificial environment—a key, a rule, an outlined argument—and when the human brain engages with that artifact and responds meaningfully and productively to it, it produces what we call reasoned intelligence behavior. Some human brains are more effective at perceiving the stimulus of artificial construction, some human brains are more effective at responding to that artificial construction, and some human brains are more effective at both, giving rise to individual intelligence differences and the concept of general intelligence ability. But no human brain has undergone a biologically radical or evolutionarily extraordinary alteration from its structure and operations of former hominin times. The dynamics of human intelligence are being driven almost entirely by the changing features of the surrounding artificial environment, and it is only because of this accumulating artificial stimulus that the human brain can now so frequently produce an entirely different kind of response—produce what we call an intelligence response.

The most prominent consequence that arises from this alternative description of human intelligence is that it provides for an elegant and extremely straightforward explanation of the Flynn effect. The overall increase in raw performance on human intelligence exams—observed directly ever since intelligence tests were invented, and demonstrated indirectly from the progressive course of human history—this increase can be characterized as being driven by the accruing amount and growing complexity of artificial construction contained within the human environment, with no need to posit any biological alteration at all. This allows general intelligence ability to continue all its functioning undisturbed and to continue to be regarded as a stable component of human intelligence, likely to be found at similar levels today within the human population as it would have been during the time of the turn towards behavioral modernity, exactly as might be anticipated for a characteristic determined primarily by neural and genetic factors. And yet despite this biological stability, measurable human intelligence can still grow indefinitely and without restriction, because measurable human intelligence is determined by more than just its neural influence. Due to the tandem impact of artificial construction, measurable human intelligence can grow as swiftly as the structural features of the surrounding environment allow, structural features not hindered by any biological limitations.

Genius

One manifestation of the misunderstandings that can arise from the standard description of human intelligence is the popular misconception surrounding the topic known as genius. In common parlance, genius is typically equated to the highest levels of IQ. For instance, famous historical figures such as Mozart, Newton and Einstein are frequently proclaimed to have possessed IQs surpassing the 150 level, even reaching well into the 200s. These assertions are of course specious—intelligence tests had not yet been invented during the days of Newton and Mozart, and there is no evidence Einstein ever took an intelligence test. The actual IQ of almost every known historical genius remains entirely unknown. And yet the reason these claims can be so widely promulgated and at the same time so broadly accepted is that there would appear to be no grounds for assuming otherwise. After all, genius means the highest levels of intelligence, does it not, and so once genius has been recognized, it would seem the prodigious IQ must of necessity follow. In the standard description of human intelligence, intelligence is IQ, intelligence is general intelligence ability, intelligence is brain effectiveness, and so genius has no other role than to occupy the far end of this one-factor scale.

And it is not just in common parlance that these notions regarding genius have taken such a firm hold, they can be frequently seen also within the intelligence research literature. Here is one telling quote from The Science of Human Intelligence:

In the longer term, if we learn how to tinker with brain mechanisms to increase reasoning ability, we might enter a new phase of personal achievement and societal well-being. Such knowledge might even create more geniuses on the level of Einstein, Newton, Cervantes, or Da Vinci (Haier et al., 2023).

Yet one more example of how higher levels of intelligence are equated to the concept of genius comes from the domain of artificial intelligence research, where one prominent figure from the field has recently boasted that advances in artificial intelligence will soon create a "country of geniuses in a datacenter" (Amodei, 2024). It seems that every direction one turns, the concept of genius has been decisively determined: genius is achieved solely and necessarily by possessing the highest levels of intelligence.

But if you find yourself actually believing such notions regarding the concept of genius, I would ask you to take a careful look at the twentieth century. The twentieth century was notable for having produced a substantial list of recognized geniuses: Einstein, de Broglie, Heisenberg, Stravinsky, Picasso, Joyce, Wittgenstein, Turing—a list likely not altogether complete. The twentieth century was also notable for having produced a substantial and observed Flynn effect, with the overall level of human intelligence increasing sizably from the beginning of that century through its end. Therefore, we have a century in which genius was initially flourishing, and we also have a century in which the population-wide level of intelligence was increasing significantly, and so if you believe in the equivalency between genius and the highest levels of human intelligence, then the only logical conclusion would be to say that by the year 2000, there must have been a genius standing on just about every street corner, you could not have walked more than fifty feet without bumping into another one. But the reality of course turned out to be just the opposite. Nearly all the genius of the twentieth century was front loaded, it occurred almost exclusively within the century's early decades, at a time when intelligence was at its relative ebb. Since then, there has been almost nothing—I personally cannot think of a single individual from the last fifty to seventy-five years whom I would deem worthy of inclusion alongside that list of names from above. As the twentieth century progressed, and as the overall level of human intelligence advanced to ever more impressive heights, genius went from being relatively abundant to being essentially disappeared. Do we really want to stick with the notion that genius is the same thing as the highest levels of intelligence?

Or try this experiment on for size (this is an experiment that can actually be performed, by the way, so I would encourage all the academicians to apply for a grant): give a standard intelligence test to a very large portion of the human population, then gather up the top one hundred scores. Put those extremely high IQ people into a room for a week and then encourage them, singularly or together, to develop as many ingenious ideas and constructs as they can. At the week's end, just open the door and allow all that ingenuity to flow out into the hallway, being careful not to let any of the more brilliant insights escape.

Does anyone actually think that such an experiment would prove to be productive in any way? I know what I would predict—I would predict that those efforts would amount to absolutely nothing in the way of genius, that the only insight emanating from that room of extremely high IQ people would be the deafening sound of silence. Genius is not the same thing as high IQ, not even close. The entire thought is conceptually wrong. What genius actually is is the solution to how human intelligence grows. And since human intelligence is composed fundamentally of the artificial construction contained within the human environment, genius can also be described as the characteristic that ignites the accumulation and advancement of the world's synthetic structure. Genius is the spark that fires the Flynn effect, and it does not require an extraordinary amount of intelligence to do its work. What it does require is an extraordinary amount of atypicality.

As humans swim in their current ocean of artificial construction, a question that might enter their thoughts is how have all these artifacts come to be? The biological world passed hundreds of millions of years without producing one constructed invention, and hominins themselves went for many millions of years without manufacturing a single thing (Klein, 2009). What prompted the first fire pit? How and why were structured tools and weapons originally crafted? What could have possibly driven an individual to draw a representational picture upon a cave wall? What launches the assembled abundance of a modern city? As it happens, growth in artificial construction can be accomplished in a variety of ways, some of which are quite common and rather mundane, not resembling very much at all the characteristics of genius. For instance, there is copying and adjusting. Whereas the very first automobile was quite the revolutionary achievement—adding a significant amount of novel and influential structure into its surrounding environment—every other automobile that then followed would have been essentially a replication of, or a variation upon, that very first template. The majority of the artificial construction that enters the human world does so through these means of replication and improvement—these are the time-honored and widely practiced techniques that allow the content of human intelligence to spread universally and ubiquitously, they are the techniques that permit the Flynn effect to become impactful population wide.

But activities such as copying and adjusting can only go so far. For instance, these activities cannot produce the original construct, and eventually they will reach a point of saturation, where further copying and adjusting produces only diminishing returns. To initiate artificial construction and to continue to expand its variety and depth, novel instances of structure must be periodically inserted into the surrounding environment—richer patterns, more fruitful forms, additional scintillating symmetries. What is required for this innovative advancement of artificial construction, and for the subsequent expansion of human intelligence, is not a perception and reproduction of the world's synthetic structure as it currently is, but instead a perception and introduction of the world's synthetic structure as it soon will be. Not a typical understanding of the current human surroundings, but instead an atypical vision of the ambient state of affairs—a coming revolution, an impending paradigm shift, an iconoclastic act of genius.

The quintessential example of these operations would be Isaac Newton's introduction of the laws of motion and gravity, followed by the resulting explosion of artificial construction produced during the era of the Industrial Revolution (Jacob & Stewart, 2004). Newton's distinctive act was to insert into his surrounding environment (insert quite literally, for instance via the original pages of the Principia) the novel structure contained inside his formulaic laws. But beyond this igniting act, Newton himself did little to add to the artificial construction of the surrounding world: Newton built no steam engines, no airplanes, no rocket ships, he did not expand the mathematics of his calculus nor oversee a multitude of physics labs. These furthering and constructive activities, various forms of copying, adjusting and improving, all traceable back to Newton's original conceptions, were carried out by countless other individuals, indeed by nearly the entire human population. The twofold advantage to this mass reproduction of Newton's original structure was that it tangibly spread that structure all around the planet, and it also encouraged a multitude of individuals to engage with that structure personally and directly. Very few people learn to master Newton's laws by reading the pages of the Principia. Most people learn to master Newton's laws by interacting with the artifacts of the modern world. Acceleration, momentum, force, energy, differentiation, etc., these concepts are embodied in the many constructions of the modern human environment, and by engaging with and navigating these constructions, the human population achieves a degree of mastery comparable to that of Newton at the moment of his conception.

Copying, adjusting, improving, navigating, mastering—these are the typical activities of engagement with the structural features of the surrounding world, they are the artifact-inspired behaviors that remain open to nearly all, and we measure the strength of an individual's ability to engage in these typical activities by formulating intelligence tests out of those very same structural features. But these typical activities, important as they are, would be entirely inconceivable without a preceding atypical event. Without the initial introduction of an unforeseen instance of novel environmental structure, every resulting structural consequence would have no opportunity to be—no copying, no adjusting, no improving, no navigating, no mastering, no anything. That initiating event, that seemingly out-of-nothing insertion of a new and momentous artificial feature into the human environment, this is an activity that is not open to all, it is in fact a behavior that is exceedingly rare. So extraordinarily exceptional, and at the same time so consequentially productive, have been these atypical events throughout the course of human history that humans have felt compelled to celebrate that original moment (celebrate it often very much in retrospect), celebrate it for what it has proven to be—revolutionary, atypical, ingenious.

Ironically, the requirement of atypicality means that an extremely intelligent individual would actually be a very poor candidate for producing genius. An extremely intelligent individual, one who possesses the highest of IQs, is demonstrating an extraordinary ability to navigate and to master the artificial structure contained within the surrounding environment as it currently exists. This would be reflected through the contents of the intelligence exam itself, which can only be composed out of questions containing structural features that are already widely recognized and well known, because no test developer would ever be able to formulate a question out of structural features that have yet to be invented. But this extraordinary ability to navigate and to master the world's artificial structure as it currently exists would serve as a hindrance to seeing that world in a much different light. A high IQ individual has already mastered the world's surrounding structure, and would have little motivation to consider radically changing it.

Therefore, I suspect there are very few, and perhaps no, historical geniuses who have possessed an extremely high IQ. Yes, these individuals would have possessed a decent amount of general intelligence ability—enough to understand the details of their relevant problem domain—but beyond that, more intelligence would have only had the tendency to get in the way. If Beethoven had been more intelligent, then perhaps he could have done a more reliable job of mimicking Haydn and Mozart. If Einstein had been more intelligent, then perhaps he could have been more effective at confirming Newton and Maxwell. Geniuses are not known for their ability to navigate and to master conventional wisdom and established paradigms—Newton and Einstein, for instance, were rather mediocre students. What geniuses are known for are their atypical characteristics—taciturn, irascible, isolated, iconoclastic—characteristics that give evidence of the genius's unusual mode of perception (Snyder, 2004). Geniuses have not mastered their surrounding environment so much as they have chafed against it, and it is out of their resulting irritation that the most consequential instances of innovative structure have been introduced into the human world.

Although not entirely essential to the discussion here, I have suggested elsewhere that autism is the the key to a deeper understanding of the atypical characteristics associated with genius (Griswold, 2024b). Autism has been traditionally researched and described as a medical condition, but fifty plus years of failed efforts along those lines have put that conjecture into doubt (Myers et al., 2020; Parellada et al., 2023; Whitehouse et al., 2021). A more effective approach to describing autism would be to focus on the concept of conspecific perception, the inherent biological tendency for an organism to perceive first and foremost the other members of its own species, a characteristic essential for holding a species together and for providing its newborn members with a sensory grounding. Autistic individuals can be characterized as experiencing a significant lack of conspecific perception, causing them to be less attached to the other members of their own species, and also causing them to be less able to achieve a sensory grounding by the usual species-focused means. Needing a sensory grounding to overcome their potential sensory chaos, most autistic individuals begin to perceptually focus on those environmental features that contain inherent form—patterns, repetitions, symmetries, etc. These structure-focused perceptions are evidenced in autistic individuals' unusual interests and behaviors—fascination with geometric objects, insistence on repetitive routines, migration towards technical disciplines, etc.—interests and behaviors targeted almost always towards ambient structure. This inherent fascination with environmental form is what ties autism to the concept of artificial construction, and it is my contention that the natural autistic proclivity towards surrounding structure, and also the natural autistic proclivity towards atypicality, is what drives the idiosyncratic perceptions that underlie the innovative and synthetic constructions of genius.

But whether or not one chooses to entertain this connection between autism and genius, the one lesson from this section that cannot be held in dispute is that the brain-specific standard description of human intelligence is inadequate for describing and explaining genius. All that the standard description can offer is the default conviction that genius must be the consequence of the highest levels of intelligence, a conjecture that is demonstrably and observably false. Just as was the case with the Flynn effect, a deeper understanding of human intelligence is necessary for describing and explaining genius—brain effectiveness by itself is not sufficient. By incorporating the impact of artificial construction into a more complete understanding of human intelligence, and by recognizing that novel growth in artificial construction is the ultimate source of the Flynn effect, one sheds meaningful light on the role that genius plays. Not a high IQ, but instead an atypical perception for the surrounding environment, that is what prompts genius to add innovative structural features into the human environment, thereby opening a pathway for those features to be copied and spread, raising the amount and complexity of artificial construction everywhere, and raising the level of measurable intelligence for the entire population.

Conclusion

What makes the study of human intelligence a scientific pursuit is that human intelligence can be measured. The instrument for making these measurements is an intelligence test, of which many examples exist, Stanford-Binet and Wechsler being two of the more commonly known and widely used. But there is an irony and a bit of a mystery currently accompanying these tests. Although researchers have been successfully employing intelligence tests for more than a century now, they still do not understand the nature of the entity that is being measured. That is, they do not yet know what is being indicated by the calibrated raw scores on intelligence exams. This unusual situation would be the equivalent of physicists using thermometers in their everyday work while having almost no conception of the nature of heat.

The primary reason for this state of affairs is that intelligence researchers have been able to make a surprisingly productive use of a derived measure of human intelligence—the IQ score. Analysis of IQ scores from contemporaneous cohorts has led to the discovery of general intelligence ability, an extremely important concept for explaining individual intelligence differences, for predicting general life circumstances, and for uncovering the cognitive influence of neural and genetic characteristics. Intelligence researchers are certainly to be applauded for these many insightful efforts, efforts that have become crystallized in the standard description of human intelligence, but researchers also need to be reminded that these derived results have been distracting them from what could have been their original task, namely figuring out what it was that was being measured via the raw scores on intelligence exams. The continuing impact of this distraction is made evident these days in the research community's response to the Flynn effect, the observed phenomenon that raw scores on intelligence tests have been increasing ever since intelligence tests were invented. The best that the research community has been able to offer in response to the Flynn effect has been a multitude of unconvincing attempts to reconcile the Flynn effect to general intelligence ability, an approach that appears to have no scientific merit, along with the somewhat desperate hope that the Flynn effect will just quietly go away, a hope that has no professional merit.

This essay has made the argument that it is the totality of artificial construction accruing within the human environment (and also appearing universally on intelligence exams) that is the key to a more comprehensive understanding of human intelligence than can be offered by the standard description alone. Measurable intelligence—the entity directly being measured via the calibrated raw scores on intelligence tests—it is the result of the orthogonal combination of general intelligence ability and the totality of artificial construction. That is to say, an individual's raw performance on an intelligence test is determined by that individual's particular neural capacity to navigate and to master artificial construction, in combination with the total amount of artificial construction to which that individual has been exposed. Or stated as a relational equation, measurable intelligence = general intelligence ability x artificial construction. Therefore, what is being directly measured via an intelligence test is not just neural effectiveness—as implied by the standard description. What is being directly measured via an intelligence test is neural effectiveness as applied to the growing ocean of artificiality in which humans now live. Both aspects of this orthogonal combination must be taken fully into account before human intelligence can be thoroughly and successfully described.

I would like to conclude by claiming that this essay's alternative description of human intelligence—in some sense an augmentation to the standard description—possesses three distinct advantages that speak in its favor. The first advantage is that it offers a depiction of the material nature of human intelligence that is specific and observable, highlighting the formal characteristics of the world's many constructed artifacts, characteristics that exist right before our very eyes and that have been deeply fathomed via the tools of mathematics, science, logic and so on. Compare these precise descriptions of the structural aspects of the surrounding artificial environment to the vague offerings emerging from the domain of neuroscience (Haier, 2021)—the difference in specificity could not be greater.

The second advantage of this essay's alternative description of human intelligence is that it does not require any biologically or evolutionarily extraordinary alterations to the operations of the human brain, claiming instead the retention of the human neural system's long-standing role as a stimulus-response mechanism. By attributing the material source of intelligence to the artificial environment—and not to the neurons, synapses and biochemical activity inside the human head—the brain can remain tasked with simply responding to the stimulus of artificial construction, an operation for which it is already well suited. This allows general intelligence ability to be depicted as a stable component of human intelligence, likely existing at similar levels today within the human population as it did several hundred thousand years ago, exactly as to be expected for a characteristic determined primarily by neural and genetic factors.

The third advantage of this essay's alternative description of human intelligence is that it outlines a straightforward and observable explanation of the Flynn effect. With general intelligence ability remaining stable within the human population, and with measurable intelligence being the orthogonal combination of general intelligence ability and artificial construction, the Flynn effect can be described as being driven entirely by the increasing amount and complexity of artificial construction contained within the human environment, an increase unhindered by any biological limitations. And as an important corollary to this understanding regarding the Flynn effect, it can also be stated with conviction that the Flynn effect is not temporary, that it has been a fundamental part of humanity ever since the species' turn towards behavioral modernity, and it will continue to be a fundamental part of humanity into the foreseeable future.

References

Amodei, D. (2024). Machines of Loving Grace. https://www.darioamodei.com/essay/machines-of-loving-grace

Anderson, K. M., & Holmes, A. J. (2021). Predicting individual differences in cognitive ability from brain imaging and genetics. In A. K. Barbey, S. Karama, & R. J. Haier (Eds.), The Cambridge handbook of intelligence and cognitive neuroscience (pp. 327–348). Cambridge University Press. https://doi.org/10.1017/9781108635462.021

Basten, U., Hilger, K., & Fiebach, C. J. (2015). Where smart brains are different: A quantitative meta-analysis of functional and structural brain imaging studies on intelligence. Intelligence, 51, 10–27. https://doi.org/10.1016/j.intell.2015.04.009

Berger, M. (1986). Toward an Educated Use of IQ Tests. In: Lahey, B.B., Kazdin, A.E. (eds) Advances in Clinical Child Psychology. Advances in Clinical Child Psychology, vol 9. Springer, Boston, MA. https://doi.org/10.1007/978-1-4613-9823-3_1

Boring, E. G. (1923). Intelligence as the tests test it. New Republic, 36, 35–37.

Bradshaw, J. L. (2002). The evolution of intellect: Cognitive, neurological, and primatological aspects and hominid culture. In R. J. Sternberg & J. C. Kaufman (Eds.), The evolution of intelligence (pp. 55–78). Lawrence Erlbaum Associates Publishers.

Bräuer, J., Hanus, D., Pika, S., Gray, R., & Uomini, N. (2020). Old and New Approaches to Animal Cognition: There Is Not "One Cognition". Journal of Intelligence, 8(3), 28. https://doi.org/10.3390/jintelligence8030028

Bru-Luna, L. M., Martí-Vilar, M., Merino-Soto, C., & Cervera-Santiago, J. L. (2021). Emotional Intelligence Measures: A Systematic Review. Healthcare (Basel, Switzerland), 9(12), 1696. https://doi.org/10.3390/healthcare9121696

Carl, N., & Woodley of Menie, M. A. (2019). A scientometric analysis of controversies in the field of intelligence research. Intelligence, 77, Article 101397. https://doi.org/10.1016/j.intell.2019.101397

Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. Cambridge University Press. https://doi.org/10.1017/CBO9780511571312

Christian, D. (2014). This fleeting world: a short history of humanity. Berkshire Publishing Group.

Christian, D. (2018). Origin story: a big history of everything. First edition. New York, Little, Brown and Company.

Colom, R., Jung, R. E., & Haier, R. J. (2006). Distributed brain sites for the g-factor of intelligence. NeuroImage, 31(3), 1359–1365. https://doi.org/10.1016/j.neuroimage.2006.01.006

Colom, R., Karama, S., Jung, R. E., & Haier, R. J. (2010). Human intelligence and brain networks. Dialogues in clinical neuroscience, 12(4), 489–501. https://doi.org/10.31887/DCNS.2010.12.4/rcolom

Colom, R., García, L. F., Shih, P. C., & Abad, F. J. (2023). Generational intelligence tests score changes in Spain: Are we asking the right question? Intelligence, 99, 101772. https://doi.org/10.1016/j.intell.2023.101772

Coltheart M. (1999). Modularity and cognition. Trends in cognitive sciences, 3(3), 115–120. https://doi.org/10.1016/s1364-6613(99)01289-9

Deary, I. J. (2020). Intelligence: a very short introduction. Second edition. Oxford University Press.

Deary, I. J., Cox, S. R., & Hill, W. D. (2022). Genetic variation, brain, and intelligence differences. Molecular psychiatry, 27(1), 335–353. https://doi.org/10.1038/s41380-021-01027-y

deLeyer-Tiarks, J. M., Caemmerer, J. M., Bray, M. A., & Kaufman, A. S. (2024). Assessment of Human Intelligence-The State of the Art in the 2020s. Journal of Intelligence, 12(8), 72. https://doi.org/10.3390/jintelligence12080072

Detterman, D. K., & Daniel, M. H. (1989). Correlations of mental tests with each other and with cognitive variables are highest for low IQ groups. Intelligence, 13(4), 349–359. https://doi.org/10.1016/S0160-2896(89)80007-8

Dickens, W. T., & Flynn, J. R. (2001). Heritability estimates versus large environmental effects: the IQ paradox resolved. Psychological review, 108(2), 346–369. https://doi.org/10.1037/0033-295x.108.2.346

Dutton, E., van der Linden, D., & Lynn, R. (2016). The negative Flynn Effect: A systematic literature review. Intelligence, 59, 163–169. https://doi.org/10.1016/j.intell.2016.10.002

Flynn, J. R. (1984). The mean IQ of Americans: Massive gains 1932 to 1978. Psychological Bulletin, 95(1), 29–51. https://doi.org/10.1037/0033-2909.95.1.29

Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101(2), 171–191. https://doi.org/10.1037/0033-2909.101.2.171

Flynn, J. R. (2007). What is intelligence?: Beyond the Flynn effect. Cambridge University Press.

Gardner, H. (1987). The Theory of Multiple Intelligences. Annals of Dyslexia, 37, 19–35. http://www.jstor.org/stable/23769277

Gottfredson, L. S. (1998). The general intelligence factor. Scientific American Presents, 9, 24-29.

Gottfredson, L. S. (2002). g: Highly general and highly practical. In R. J. Sternberg & E. L. Grigorenko (Eds.), The general factor of intelligence: How general is it? (pp. 331–380). Lawrence Erlbaum Associates Publishers.

Gottfredson, L. S. (2007). Shattering logic to explain the Flynn Effect. Cato Unbound, November 8.

Graham, S. A., & Fisher, S. E. (2013). Decoding the genetics of speech and language. Current opinion in neurobiology, 23(1), 43–51. https://doi.org/10.1016/j.conb.2012.11.006

Griswold, A. (2024a). Rethinking the Flynn effect. In Autistic Études. iUniverse.

Griswold, A. (2024b). Prototypical autism is transformatively atypical. In Autistic Études. iUniverse.

Haier, R. J., Jung, R. E., Yeo, R. A., Head, K., & Alkire, M. T. (2004). Structural brain variation and general intelligence. NeuroImage, 23(1), 425–433. https://doi.org/10.1016/j.neuroimage.2004.04.025

Haier, R. J. (2016). The Neuroscience of Intelligence. Cambridge: Cambridge University Press.

Haier, R. J. (2021). Are we thinking big enough about the road ahead? Overview of the special issue on the future of intelligence research. Intelligence, 89(C). https://doi.org/10.1016/j.intell.2021.101603

Haier, R. J., Colom, R., & Hunt, E. (2023). The Science of Human Intelligence (2nd ed.). Cambridge: Cambridge University Press.

Henshilwood, C. S., & Marean, C. W. (2003). The Origin of Modern Human Behavior: Critique of the Models and Their Test Implications. Current Anthropology, 44(5), 627–651. https://doi.org/10.1086/377665

Hilger, K., Ekman, M., Fiebach, C. J., & Basten, U. (2017). Intelligence is associated with the modular structure of intrinsic brain networks. Scientific reports, 7(1), 16088. https://doi.org/10.1038/s41598-017-15795-7

Jacob, M. C., & Stewart, L. (2004). Practical Matter: Newton's Science in the Service of Industry and Empire, 1687–1851. Harvard University Press.

Jung, R. E., & Haier, R. J. (2007). The Parieto-Frontal Integration Theory (P-FIT) of intelligence: converging neuroimaging evidence. The Behavioral and brain sciences, 30(2), 135–187. https://doi.org/10.1017/S0140525X07001185

Kihlstrom, J. F., & Cantor, N. (2000). Social intelligence. In R. J. Sternberg (Ed.), Handbook of intelligence (pp. 359–379). Cambridge University Press. https://doi.org/10.1017/CBO9780511807947.017

Klein, R. (2002). The Dawn of Human Culture. New York: Wiley.

Klein, R. G. (2009). The human career: Human biological and cultural origins. University of Chicago Press.

Kovacs, K., & Conway, A. R. A. (2016). Process Overlap Theory: A Unified Account of the General Factor of Intelligence. Psychological Inquiry, 27(3), 151–177. https://doi.org/10.1080/1047840X.2016.1153946

Kuncel, N. R., Hezlett, S. A., & Ones, D. S. (2004). Academic performance, career potential, creativity, and job performance: can one construct predict them all?. Journal of personality and social psychology, 86(1), 148–161. https://doi.org/10.1037/0022-3514.86.1.148

Langer, N., Pedroni, A., Gianotti, L. R., Hänggi, J., Knoch, D., & Jäncke, L. (2012). Functional brain network efficiency predicts intelligence. Human brain mapping, 33(6), 1393–1406. https://doi.org/10.1002/hbm.21297

Liu, J., & Lynn, R. (2013). An Increase of Intelligence in China 1986-2012. Intelligence, 41(5), 10.1016/j.intell.2013.06.017. https://doi.org/10.1016/j.intell.2013.06.017

Marsman, A., Mandl, R. C. W., Klomp, D. W. J., Cahn, W., Kahn, R. S., Luijten, P. R., & Hulshoff Pol, H. E. (2017). Intelligence and Brain Efficiency: Investigating the Association between Working Memory Performance, Glutamate, and GABA. Frontiers in psychiatry, 8, 154. https://doi.org/10.3389/fpsyt.2017.00154

Meyer, A. F., Williamson, R. S., Linden, J. F., & Sahani, M. (2017). Models of Neuronal Stimulus-Response Functions: Elaboration, Estimation, and Evaluation. Frontiers in systems neuroscience, 10, 109. https://doi.org/10.3389/fnsys.2016.00109

Myers, S. M., Challman, T. D., Bernier, R., Bourgeron, T., Chung, W. K., Constantino, J. N., Eichler, E. E., Jacquemont, S., Miller, D. T., Mitchell, K. J., Zoghbi, H. Y., Martin, C. L., & Ledbetter, D. H. (2020). Insufficient Evidence for "Autism-Specific" Genes. American journal of human genetics, 106(5), 587–595. https://doi.org/10.1016/j.ajhg.2020.04.004

Nijenhuis, J.T., Cho, S.H., Murphy, R., & Lee, K.H. (2012). The Flynn effect in Korea: Large gains. Personality and Individual Differences, 53, 147-151. https://doi.org/10.1016/j.paid.2011.03.022

Nijs, D. E. L. W. (2015). Introduction: Coping with Growing Complexity in Society. World Futures, 71(1–2), 1–7. https://doi.org/10.1080/02604027.2015.1087223

Parellada, M., Andreu-Bernabeu, Á., Burdeus, M., San José Cáceres, A., Urbiola, E., Carpenter, L. L., Kraguljac, N. V., McDonald, W. M., Nemeroff, C. B., Rodriguez, C. I., Widge, A. S., State, M. W., & Sanders, S. J. (2023). In Search of Biomarkers to Guide Interventions in Autism Spectrum Disorder: A Systematic Review. The American journal of psychiatry, 180(1), 23–40. https://doi.org/10.1176/appi.ajp.21100992

Pietschnig, J., & Gittler, G. (2015). A reversal of the Flynn effect for spatial perception in German-speaking countries: Evidence from a cross-temporal IRT-based meta-analysis (1977–2014). Intelligence, 53, 145–153. https://doi.org/10.1016/j.intell.2015.10.004

Pietschnig, J., & Voracek, M. (2015). One Century of Global IQ Gains: A Formal Meta-Analysis of the Flynn Effect (1909–2013). Perspectives on Psychological Science, 10(3), 282-306. https://doi.org/10.1177/1745691615577701

Plomin, R., & von Stumm, S. (2018). The new genetics of intelligence. Nature reviews. Genetics, 19(3), 148–159. https://doi.org/10.1038/nrg.2017.104

Rączaszek-Leonardi, J., Krzesicka, J., Klamann, N., Ziembowicz, K., Denkiewicz, M., Kukiełka, M., & Zubek, J. (2019). Cultural Artifacts Transform Embodied Practice: How a Sommelier Card Shapes the Behavior of Dyads Engaged in Wine Tasting. Frontiers in psychology, 10, 2671. https://doi.org/10.3389/fpsyg.2019.02671

Rodgers, J. (2023). Eleven articles and 27 authors pay tribute to James Flynn: A summary and critique of special issue articles on the Flynn effect. Intelligence, 101. https://doi.org/10.1016/j.intell.2023.101794

Shaw, R. C., & Schmelz, M. (2017). Cognitive test batteries in animal cognition research: evaluating the past, present and future of comparative psychometrics. Animal cognition, 20(6), 1003–1018. https://doi.org/10.1007/s10071-017-1135-1

Simmons, P., & Young, D. (2010). Nerve cells and animal behaviour (3rd ed.). Cambridge University Press. https://doi.org/10.1017/CBO9780511782138

Snyder, A. (2004). Autistic genius?. Nature 428, 470–471. https://doi.org/10.1038/428470a

Spearman C. (1904). "General intelligence," objectively determined and measured. The American Journal of Psychology, 15, 201–292. http://www.jstor.org/stable/1412107

Sterelny K. (2011). From hominins to humans: how sapiens became behaviourally modern. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 366(1566), 809–822. https://doi.org/10.1098/rstb.2010.0301

te Nijenhuis, J., & van der Flier, H. (2013). Is the Flynn effect on g?: A meta-analysis. Intelligence, 41(6), 802–807. https://doi.org/10.1016/j.intell.2013.03.001

Terman, L. M. (1916). The measurement of intelligence. Houghton, Mifflin and Company. https://doi.org/10.1037/10014-000

Toga, A. W., & Thompson, P. M. (2005). Genetics of brain structure and intelligence. Annual review of neuroscience, 28, 1–23. https://doi.org/10.1146/annurev.neuro.28.061604.135655

Tooley, U. A., Bassett, D. S., & Mackey, A. P. (2021). Environmental influences on the pace of brain development. Nature reviews. Neuroscience, 22(6), 372–384. https://doi.org/10.1038/s41583-021-00457-5

Trahan, L. H., Stuebing, K. K., Fletcher, J. M., & Hiscock, M. (2014). The Flynn effect: a meta-analysis. Psychological bulletin, 140(5), 1332–1360. https://doi.org/10.1037/a0037173

van der Linden, M., & Borsboom, D. (2019). Affluence boosted intelligence? How the interaction between cognition and environment may have produced an eighteenth-century Flynn effect during the Industrial Revolution. The Behavioral and brain sciences, 42, e211. https://doi.org/10.1017/S0140525X19000190

Whitehouse, A., Varcin, K., Waddington, H., Sulek, R., Bent, C., Ashburner, J., Eapen, V., Goodall, E., Hudry, K., Roberts, J., Silove, N., Trembath, D. (2021). Interventions for children on the autism spectrum: A synthesis of research evidence. Autism CRC, Brisbane.

Wongupparaj, P., Wongupparaj, R., Morris, R. G., & Kumari, V. (2023). Seventy years, 1000 samples, and 300,000 SPM scores: A new meta-analysis of Flynn effect patterns. Intelligence, 98, 1–9. https://doi.org/10.1016/j.intell.2023.101750

Woodley, M. A. (2012). A life history model of the Lynn-Flynn effect. Personality and Individual Differences, 53, 152-156. https://doi.org/10.1016/j.paid.2011.03.028