5000 words

Introduction

“General Intelligence” or g is championed as the hallmark “discovery” of psychology. First “discovered” by Charles Spearman in 1904, noting that schoolchildren who scored highly on one test scored highly on others and vice versa for lower-scoring children, he assumed that due to the correlation between tests, that there must be an underlying physiological basis to the correlation, which he posited to be some kind of “mental energy”, stating that the central nervous system (CNS) explained the correlation. He proclaimed that g really existed and that he had verified Galton’s claim of a unitary general ability (Richardson, 2017: 82-83). Psychometricians then claim, from these intercorrelations of scores, that what is hidden from us is then revealed, and that the correlations show that something exists and is driving the correlation in question. That’s the goal of psychometrics/psychology—to quantify and then measure psychological traits/mental abilities. However, I have argued at length that it is a conceptual impossibility—the goal of psychometrics is an impossibility since psychometrics isn’t measurement. Therefore, claims that IQ tests measure g is false.

First, I will discuss the reification of g and it’s relation to brain properties. I will argue that if g is a thing then it must have a biological basis, that is it must be a brain property. Reductionists like Jensen have said as much. But it’s due to the reification of g as a concrete, physical thing that has people hold such beliefs. Second, I will discuss Geary’s theory that g is identical with mitochondrial functioning. I will describe what mitochondria does, and what powers it, and then discuss the theory. I will have a negative view of it, due to the fact that he is attempting to co-opt real, actual functions of a bodily process and attempt to weave g theory into it. Third, I will discuss whether or not psychological traits are indeed quantifiable and measurable, and whether or not there is a definition psychometricians can use to ground their empirical investigations. I will argue negatively for all three. Fourth, I will discuss Herrnstein and Murray’s 6 claims in The Bell Curve about IQ and provide a response to each in turn. Fifth, I will discuss the real cause of score variation, which isn’t reduction to a so-called assumed existence of a biological process/mechanism, but which is due to affective factors and exposure to the specific type of knowledge items on the test. Lastly, I will conclude and give an argument for why g isn’t a thing and is therefore immeasurable.

On reifications and brain properties

Contrary to protestations from psychometricians, they in fact do reifiy correlations and then claim that there exists some unitary, general factor that pervades all mental tests. If reification is treating the abstract as something physical, and if psychometrics treat g as something physical, then they are reifying g based on mere intercorrelations between tests. I am aware that, try as they might, they do attempt to show that there is an underlying biology to g, but these claims are defeated by the myriad arguments I’ve raised against the reducibility of the mental to the physical. Another thing that Gould gets at is that psychometricians claim that they can rank people—this is where the psychometric assumption that because we can put a number to their reified thing, that there is something being measured.

Reification is “the propensity to convert an abstract concept (like intelligence) into a hard entity (like an amount of quantifiable brain stuff)” (Gould, 1996: 27). So g theorists treat g as a concrete, physical, thing, which then guides their empirical investigations. They basically posit that the mental has a material basis, and they claim that they can, by using correlations between different test batteries, we can elucidate the causal biological mechanisms/brain properties responsible for the correlation.

Spearman’s theory—and IQ—is a faculty theory (Nash, 1990). It is a theory in which it is claimed that the mind is separated into different faculties, where mental entities cause the intellectual performance. Such a theory needs to keep up the claim that a cognitive faculty is causally efficacious for information processing. But the claim that the mind is “separated” into different faculties fails, and it fails since the mind is a single sphere of consciousness, it is not a complicated arrangement of mental parts. Physicalists like Jensen and Spearman don’t even have a sound philosophical basis on which to ground their theories. Their psychology is inherently materialist/physicalist, but materialism/physicalism is false and so it follows that their claims do not hold any water. The fact of the matter is, Spearman saw what he wanted to see in his data (Schlinger, 2003).

I have already proven that since dualism is true, then the mental is irreducible to the physical and since psychometrics isn’t measurement, then what psychometricians claim to do just isn’t possible. I have further argued that science can’t study first-personal subjective states since science is third-personal and objective. The fact is the matter is, hereditarian psychologists are physicalist, but it is impossible for a purely physical thing to be able to think. Claims from psychometricians about their “mental tests” basically reduce to one singular claim: that g is a brain property. I have been saying this for years—if g exists, it has to be a brain property. But for it to be a brain property, one needs to provide defeaters for my arguments against the irreducibility of the mental and they also need to argue against the arguments that psychometrics isn’t measurement and that psychology isn’t quantifiable. They can assume all they want that it is quantifiable and that since they are giving tests, questionnaires, likert scales, and other kinds of “assessments” to people that they are really measuring something; but, ultimately, if they are actually measuring something, then that thing has to be physical.

Jensen (1999) made a suite of claims trying to argue for a physical basis for g,—to reduce g to biology—though, upon conceptual examination (which I have provided above) these claims outright fail:

g…[is] a biological [property], a property of the brain

The ultimate arbiter among various “theories of intelligence” must be the physical properties of the brain itself. The current frontier of g research is the investigation of the anatomical and physiological features of the brain that cause g.

…psychometric g has many physical correlates…[and it] is a biological phenomenon.

As can be seen, Jensen is quite obviously claiming that g is a biological brain property—and this is what I’ve been saying to IQ-ists for years: If g exists, then it MUST be a property of the brain. That is, it MUST have a physical basis. But for g proponents to show this is in fact reality, they need to attempt to discredit the arguments for dualism, that is, they need to show that the mental is reducible to the physical. Jensen is quite obviously saying that a form of mind-brain identity is true, and so my claim that it was inevitable for hereditarianism to become a form of mind-brain identity theory is quite obviously true. The fact of the matter is, Jensen’s beliefs are reliant upon an outmoded concept of the gene, and indeed even a biologically implausible heritability (Richardson, 1999; Burt and Simons, 2014, 2015).

But Jensen (1969) contradicted himself when it comes to g. On page 9, he writes that “We should not reify g as an entity, of course, since it is only a hypothetical construct intended to explain covariation am ong tests. It is a hypothetical source of variance (individual differences) in test scores.” But then 10 pages later on pages 19-20 he completely contradicts himself, writing that g is “a biological reality and not just a figment of social conventions.” That’s quite the contradiction: “Don’t reifiy X, but X is real.” Jensen then spent the rest of his career trying to reduce g to biology/the brain (brain properties), as we see above.

But we are now in the year 2023, and so of course there are new theoretical developments which attempt to show that Spearman’s hypothesized mental energy really does exist, and that it is the cause of variations in scores and of the positive manifold. This is now where we will turn.

g and mitochondrial functioning

In a series of papers, David Geary (2018, 2019, 2020, 2021) tries to argue that mitochondriaal functioning is the core component in g. At last, Spearman’s hypothetical construct has been found in the biology of our cells—or has it?

One of the main functions of mitochondria is to oxidative phosphorylation to produce adenosine triphosphate (ATP). All living cells use ATP as fuel, it acts as a signaling molecule, it is also involved in cellular differentiation and cell death (Khakh and Burnstock, 2009). The role of mitochondrial functioning in spurring disease states has been known for a while, such as with cardiovascular diseases such as cardiomyopathy (Murphy et al, 2016, Ramaccini at al, 2021).

So due to the positive manifold, where performance in one thing is correlated with a performance in another, Geary assumes—as Spearman and Jensen did before him—that there must be some underlying biological mechanism which then explains the correlation. Geary then uses established outcomes of irregular mitochondrial functioning to then argue that the mental energy that Spearman was looking for could be found in mitochondrial functioning. Basically, this mental energy is ATP. I don’t deny that mitochondriaal functioning plays a role in the acquisition of disease states, indeed this has been well known (eg, Gonzales et al, 2022). What I deny is Gary’s claim that mitochondrial functioning has identity with Spearman’s g.

His theory is, like all other hereditarian-type theories, merely correlative—just like g theory. He hasn’t shown any direct, causal, evidence of mitochondrial functioning in “intelligence” differences (nor for a given “chronological age). That as people age their bodies change which then has an effect on their functioning doesn’t mean that the powerhouse of the cell—ATP—is causing said individual differences and the intercorrelations between tests (Sternberg, 2020). Indeed, environmental pollutants affect mitochondrial functioning (Byun and Baccarelli, 2014; Lambertini and Byun, 2016). Indeed, most—if not all—of Geary’s hypotheses do not pass empirical investigation (Schubert and Hagemann, 2020). So while Geary’s theory is interesting and certainly novel, it fails in explaining what he set out to.

Quantifiable, measurable, definable, g?

The way that g is conceptualized is that there is a quantity of it—where one has “more of it” than other people, and this, then, explains how “intelligent” they are in comparison to others—so implicit in so-called psychometric theory is that whatever it is their tests are tests of, something is being quantified. But what does it mean to quantify something? Basically, what is quantification? Simply, it’s the act of giving a numerical value to a thing that is measured. Now we have come to an impasse—if it isn’t possible to measure what is immaterial, how can we quantify it? That’s the thing, we can’t. The g approach is inherently a biologically reductionist one. Biological reductionism is false. So the g approach is false.

Both Gottfredson (1998) and Plomin (1999) make similar claims to Jensen, where they talks about the “biology of g” and the “genetics of g“. Plomin (1999) claims that studies of twins show that g has a substantial heritability, while Gottfredson (1998) claims that heritability of IQ increases to up until adulthood where it “rises to 60 percent in adolescence and to 80 percent by late adulthood“, citing Bouchard’s MISTRA (Minnesota Study of Twins Reared Apart). (See Joseph 2022 for critique and for the claim that the heritability of IQ in that study is 0 percent.) They, being IQ-ists, of course assume a genetic component to this mystical g. However, there arguments are based on numerous false assumptions and studies with bad designs (and hidden results), and so they must be rejected.

If X is quantitative, then X is measurable. If X is measurable, then X has a physical basis. Psychological traits don’t have a physical basis. So psychological traits aren’t quantitative and therefore not measurable. Geary’s attempt at arguing for identity between g and mitochondrial functioning is an attempt at a specified measured object for g, though his theory just doesn’t hold. Stating truisms about a biological process and then attempting to liken the process with the construct g just doesn’t work; it’s just a post-hoc rationalization to attempt to liken g with an actual biological process.

Furthermore, if X is quantitative, then there is a specified measured object, object of measurement and measurement unit for X. But this is where things get rocky for g theorists and psychometricians. Psychometry is merely pseudo-measurement. Psychometricians cannot give a specified measured object, and if they can’t give a specified measured object they cannot give an object of measurement. They thusly also cannot construct a measurement unit. Therfore, “the necessary conditions for metrication do not exist” (Nash, 1990: 141). Even Haier (2014, 2018) admits that IQ test scores don’t have a unit that is like inches, liters, or grams. This is because those are ratio scales and IQ is ordinal. That is, there is no “0-point” for IQ, like there is for other actual, real measures like temperature. That’s the thing—if you have a thing to be measured, then you have a physical object and consequently a measument unit. But this is just not possible for psychometry. I then wonder why Haier doesn’t follow what he wrote to its logical conclusion—that the project of psychometrics is just not possible. Of course the concept of intelligence doesn’t have a referent, that is, it doesn’t name a property like height, weight, or temperature (Midgley, 2018:100-101). Even the most-cited definition of intelligence—Gottfredson’s—still fails, since she contradicts herself in her very definition.

Of course IQ “ranks” people by their performance—some people perform better on the test than others (which is an outcome of prior experience). So g theorists and IQ-ists assume that the IQ test is measuring some property that varies between groups which then leads to score differences on their psychometric tests. But as Roy Nash (1990: 134) wrote:

It is impossible to provide a satisfactory, that is non-circular, definition of the supposed ‘general cognitive ability’ IQ tests attempt to measure and without that definition IQ theory fails to meet the minimal conditions of measurement.

But Boeck and Kovas (2020) try to sidestep this issue with an extraordinary claim, “Perhaps we do not need a definition of intelligence to investigate intelligence.” How can we investigate something sans a definition of the object of investigation? How can we claim that a thing is measured if we have no definition, and no specified measured object, object of measurement and measurement unit, as IQ-ists seem to agree with? Again, IQ-ists don’t take these conclusions to their further logical conclusion—that we simply just cannot measure and quantify psychological traits.

Haier claims that PGS and “DNA profiles” may lead to “new definitions of intelligence” (however ridiculous a claim). He also, in 2009, had a negative outlook on identifying a “neuro g” since “g-scores derived from different test batteries do not necessarily have equivalent neuro-anatomical substrates, suggesting that identifying a “neuro-g” will be difficult” (Haier, 2009). But one more important reason exists, and it won’t just make it “difficult” to identify a neuro g, it makes it conceptually impossible. That is the fact that cognitive localizations are not possible, and that we reify a kind of average in brain activations when we look at brain scans using fMRI. The fact of the matter is, neuroreduction just isn’t possible, empirically (Uttal, 2001, 2014, 2012), nor is it possible conceptually.

Herrnstein and Murray’s 6 claims

Herrnstein and Murray (1994) make six claims about IQ (and also g):

(1) There is such a thing as a general factor of cognitive on which human beings differ.

Of course implicit in this claim is that it’s a brain property, and that people have this in different quantities. However, the discussion above puts this claim to bed since psychological traits aren’t quantitative. This, of course comes from the intercorrelations of test scores. But we will see that most of the source of variation isn’t even entirely cognitive and is largely affective and due to one’s life experiences (due to the nature of the item content).

(2) All standardized tests of academic aptitude or achievement measure this general factor to some degree, but IQ tests expressly designed for that purpose measure it most accurately.

Of course Herrnstein and Murray are married to the idea that these tests are measures of something, that since they give different numbers due to one’s performance, there must be an underlying biology behind the differences. But of course, psychometry isn’t true measurement.

(3) IQ scores match, to a first degree, whatever it is that people mean when they use the word intelligent or smart in ordinary language.

That’s because the tests are constructed to agree with prior assumptions on who is or is not “intelligent.” As Terman constructed his Stanford-Binet to agree with his own preconceived notions of who is or is not “intelligent”: “By developing  an exclusion-inclusion criteria that favored the  aforementioned groups, test developers created a norm “intelligent” (Gersh, 1987, p.166) population “to differentiate subjects of known superiority from subjects of known inferiority” (Terman, 1922, p. 656)” (Bazemoore-James, Shinaprayoon, and Martin 2017). Of course, since newer tests are “validated”(that is, correlated with) older, tests (Richardson, 1991, 2000, 2002, 2017; Howe, 1997), this assumption is still alive today.

(4) IQ scores are stable, although not perfectly so, over much of a person’s life.

IQ test scores are malleable, and this of course would be due to the experience one has in their lives which would then have them ready to take a test. Even so, if this claim were true, it wouldn’t speak to the “biology” of g.

(5) Properly administered IQ tests are not demonstrably biased against social, economic, ethnic, or racial groups.

This claim is outright false and can be known quite simply: the items on IQ tests derive from specific classes, mainly the white middle-class. Since this is true, it would then follow that people who are not exposed to the item content and test structures wouldn’t be as prepared as those who are. Thus, IQ tests are biased against different groups, and if they are biased against different groups it also follows that they are biased for certain groups, mainly white Americans. (See here for considerations on Asians.)

(6) Cognitive ability is substantially heritable, apparently no less than 40 percent and no more than 80 percent.

It’s nonsense to claim that one can apportion heritability into genetic and environmental causes, due to the interaction between the two. IQ-ists may claim that twin, family, and adoption studies show that IQ is X amount heritable so there must thusly be a genetic component to differences in test scores. But the issue with heritability has been noted for decades (see Charney, 2012, 2016, 2022; Joseph, 2014, Moore and Shenk, 2016, Richardson, 2017) so this claim also fails. There is also the fact that behavioral genetics doesn’t have any “laws.” It’s simply fallacious to believe that nature and nurture, genes and environment, contribute additively to the phenotype, and that their relative contributions to the phenotype can be apportioned. But hereditarians need to keep that facade up, since it’s the only way their ideas can have a chance at working.

What explains the intercorrelations?

We still need an explanation of the intercorrelations between test scores. I have exhaustively argued that the usual explanations from hereditarianism outright fail—g isn’t a biological reality and IQ tests aren’t a measure at all because psychometrics isn’t measurement. So what explains the intercorrelations? We know that IQ tests are comprised of different items, whether knowledge items or more “abstract” items like the Raven. Therefore, we need to look to the fact that people aren’t exposed to certain things, and so if one comes across something novel that they’ve never been exposed to, they thusly won’t know how to answer it and their score will then be affected due to their ignorance of the relationship between the question and answer on the test. But there are other things irrespective of the relationship between one’s social class and the knowledge they’re exposed to, but social class would still then have an effect on the outcome.

IQ is, merely, numerical surrogates for class affiliation (Richardson, 1999; 2002; 2022). The fact of the matter is, all human cognizing takes place in specific cultural contexts in which cultural and psychological tools are used. This means, quite simply, that culture-fair tests are impossible and, therefore, that such tests are necessarily biased against certain groups, and so they are biased for certain groups. Lev Vygotsky’s sociocultural theory of cognitive development and his concepts of psychological and cultural tools is apt here. This is wonderfully noted by Richardson (2002: 288):

IQ tests, the items of which are designed by members of a rather narrow social class, will tend to test for the acquisition of a rather particular set of cultural tools: in effect, to test, or screen, for individuals’ psychological proximity to that set per se, regardless of intellectual complexity or superiority as such.

Thinking is culturally embedded and contextually-specific (although irreducible to physical things), mediated by specific cultural tools (Richardson, 2002). This is because one is immersed in culture immediately from birth. But what is a cultural tool? Cultural tools include language (Weitzman, 2013) (it’s also a psychological tool), along with “different kinds of numbering and counting, writing schemes, mnemonic technical aids, algebraic symbol systems, art works, diagrams, maps, drawings, and all sorts of signs (John-Steiner & Mahn, 1996; Stetsenko, 1999)” (Robbins, 2005). Children are born into cultural environments, and also linguistically-mediated environments (Vasileva and Balyasnikova, 2019). But what are psychological tools? One psychological tool (which would also of course be cultural tools) would be words and symbols (Vallotton and Ayoub, 2012).

Vygotsky wrote: “In human behavior, we can observe a number of artificial means aimed at mastering one’s own psychological processes. These means can be conditionally called psychological tools or instruments… Psychological tools are artificial and intrinsically social, rather than natural and individual. They are aimed at controlling human behavior, no matter someone else’s or one’s own, just as technologies are aimed at controlling nature” (Vygotsky, 1982, vol. 1, p. 103, my translation). (Falikman, 2021).

The source of variation in IQ tests, after having argued that social class is a compound of the cultural tools one is exposed to. Furthermore, it has been shown that the language and numerical skills used on IQ tests are class-dependent (Brito, 2017). Thus, the compounded cultural tools of different classes and racial groups then coalesce to explain how and why they score the way they do. Richardson (2002: 287-288) writes

that the basic source of variation in IQ test scores is not entirely (or even mainly) cognitive, and what is cognitive is not general or unitary. It arises from a nexus of sociocognitive-affective factors determining individuals’ relative preparedness for the demands of the IQ test. These factors include (a) the extent to which people of different social classes and cultures have acquired a specific form of intelligence (or forms of knowledge and reasoning); (b) related variation in ‘academic orientation’ and ‘self-efficacy beliefs’; and (c) related variation in test anxiety, self-confidence, and so on, which affect performance in testing situations irrespective of actual ability.

Basically, what explains the intercorrelations of test scores—so-called g—are affective, non-cognitive factors (Richardson and Norgate, 2015). Being prepared for the tests, being exposed to the items on the tests (from which are drawn from the white middle-class) explains IQ score differences, not a mystical g that some have more of than others. That is, what explains IQ score variation is one’s “distance” from the middle-class—this follows due to the item content on the test. At the end of the day, IQ tests don’t measure the ability for complex cognition. (Richardson and Norgate, 2014). So one can see that differing acquisition of cultural tools by different cultures and classes would then explain how and why individuals of those groups then attain different knowledge. This, then, would license the claim that one’s IQ score is a mere outcome of their proximity to the certain cultural tools in use in the tests in question (Richardson, 2012).

The fact of the matter is, children do not enter school with the same degree of readiness (Richardson, 2022), and this is due to their social class and the types of things they are exposed to in virtue of their class membership (Richardson and Jones, 2019). Therefore, the explanation for these differences in scores need not be some kind of energy that people have in different quantities, it’s only the fact that from birth we are exposed to different cultures and therefore different cultural and psychological tools which then causes differences in the readiness of children for school. We don’t need to posit any supposed biological mechanism for that, when the answer is clear as day.

Conclusion

As can be seen from this discussion, it is clear that IQ-ist claims of g as a biological brain property fail. They fail because psychometrics isn’t measurement. They fail because psychometricians assume that what they are “measuring” (supposedly psychological traits) have a physical basis and have the necessary components for metrication. They fail because the proposed biology to back up g theory don’t work, and claiming identity between g and a biological process doesn’t mean that g has identity between that biological process. Merely describing facts about physiology and then attempting to liken it to g doesn’t work.

Psychologists try so very hard for psychology to be a respected science, even when what they are studying bares absolutely no relationship to the objects of scientific study. Their constructs are claimed to be natural kinds, but they are merely historically contingent. Due to the way these tests are constructed, is it any wonder why such score differences arise?

The so-called g factor is also an outcome of the way tests are constructed:

Subtests within a battery of intelligence tests are included n the basis of them showing a substantial correlation with the test as a whole, and tests which do not show such correlations are excluded. (Tyson, Jones, and Elcock, 2011: 67)

This is why there is a correlation between all subtests that comprise a test. Because it is an artificial creation of the test constructors, just like their normal curve. Of course if you pick and choose what you want in your battery or test, you can then coax it to get the results you want and then proclaim that what explains the correlations are some sort of unobserved, hidden variable that individuals have different quantities of. But the assumption that there is a quantity of course assumes that there is a physical basis to that thing. Physicalists like Jensen, Spearman, and then Haier of course presume that intelligence has a physical basis and is either driven by genes or can be reduced to neurophysiology. These claims don’t pass empirical and conceptual analysis. For these reasons and more, we should reject claims from hereditarian psychologists when they claim that they have discovered a genetic or neurophysiological underpinning to “intelligence.”

At the end of the day, the goal of psychometrics is clearly impossible. Try as they might, psychometricians will always fail. Their “science” will never be on the level of physics or chemistry, and that’s because they have no definition of intelligence, nor a specified measured object, object of measurement and measurement unit. They know this, and they attempt to construct arguments to argue their way out of the logical conclusions of those facts, but it just doesn’t work. “General intelligence” doesn’t exist. It’s a mere creation of psychologists and how they make their tests, so it’s basically just like the bell curve. Intelligence as an essence or quality is a myth; just because we have a noun “intelligence” doesn’t mean that there really exists a thing called “intelligence” (Schlinger, 2003). The fact is the matter is, intelligence is simply not an explanatory concept (Howe, 1997).

IQ-ist ideas have been subject to an all-out conceptual and empirical assault for decades. The model of the gene they use is false, (DNA sequences have no privileged causal role in development), heritability estimates can’t do what they need them to do, how the estimates are derived rest on highly environmentally-confounded studies, the so-called “laws” of behavioral genetics are anything but, they lack definitions and specified measured objects, objects of measurement and measurement units. It is quite simply clear that hereditarian ideas are not only empirically false, but they are conceptually false too. They don’t even have their concepts in order nor have they articulated exactly WHAT it is they are doing, and it clearly shows. The reification of what they claim to be measuring is paramount to that claim.

This is yet another arrow in the quiver of the anti-hereditarian—their supposed mental energy, their brain property, simply does not, nor can it, exist. And if it doesn’t exist, then they aren’t measuring what they think they’re measuring. If they’re not measuring what they think they’re measuring, then they’re showing relationships between score outcomes and something else, which would be social class membership along with everything else that is related with social class, like exposure to the test items, along with other affective variables.

Now here is the argument (hypothetical syllogism):

P1: If g doesn’t exist, then psychometricians are showing other sources of variation for differences in test scores.

P2: If psychometricians are showing other sources of variation for differences in test scores and we know that the items on the tests are class-dependent, then IQ score differences are mere surrogates for social class.

C: Therefore, if g doesn’t exist, then IQ score differences are mere surrogates for social class.