H. macrocephalus

There has been a long standing assumption that the rate of cultural and technological evolution is so much faster than biological evolution that we can safely ignore the latter.  Also, speciation is taken to be the result of reproductive isolation and humans, instead of becoming more isolated, are becoming substantially more reproductively connected.  However, there is something called sympatric speciation where assortative mating causes effective reproductive isolation through behaviors rather than through natural barriers.  It is controversial, but, I would argue, is demonstrably taking place in humans.

There are all sorts of television shows and movies about the next human evolutionary step and they typically involve the emergence of paranormal abilities.  However, the reality is that nearly all evolution involves movement along dimensions that already exist within the species.  For example, when we look at our immediate evolutionary precursor, H. erectus, we see that they had an average cranial capacity of about 1,000 cm³.  This is significantly more than the 600 cm³ capacity of its predecessor H. habilis but much less than the modern human capacity of 1,350 cm³. 

In fact, if we were to project a successor to H. sapiens, we would first imagine that it would have an average cranial capacity in the range of 1,700 cm³.  Such people should exist within the human population at the rate of one in 4,300.  In fact, they are much more common than that.  A preliminary analysis suggests that the 'fat tail' of cranial capacity matches the fat tail of ratio IQ.  In other words, cranial voumes of 1,700 cm³ are found in one in 260 individuals.

Clearly, the supposition is, as is the case in Physical Anthropology, that larger brains will lead to greater intelligence and essentially new intellectual abilities.  We accept that H. erectus was smarter than H. habilis and that H. sapiens is smarter than H. erectus and that cranial capacity is evidence of that.  However, we resist the notion that larger brains in modern humans result in higher intelligence despite the anthropological tradition that militates for the relationship. 

Also, and this is critical, the notion that one form of homo directly supplants the other is incorrect.  H. habilis lived from 2.3 to 1.4 million years ago.  Its successor, H. erectus, lived from 1.8 million to 300 thousand years ago.  For 400 thousand years both were extant.  Its successor H. sapiens (including archaic forms) lived from about 600 thousand years ago to the present.  So, H. Erectus and H. sapiens were simultaneously extant for about 300 thousand years.  For at least 100ky, H. sapiens, H.neanderthalensis coexisted in the Levant.  A successor species to H. sapiens, what I call H. macrocephalus or big headed man, could co-exist with it for quite some time, at least in theory.

We actually have good evidence that sympatric speciation is already creating H. macrocephalus.  It shows up first in our efforts to measure the intelligence of children.  IQ was originally a ratio, Mental Age/Chronological Age X 100.  In other words a 150 IQ eight year old will perform on an IQ test in a similar way to a normal twelve year old.  When these tests were first developed it was found that the standard deviation was about 16 points.  In other words, there should be no children with IQs over 200 and only about one in three million should have IQs over 180.

However, extraordinary IQ turned out to be much more common than predicted.  180 IQ children turned up at the rate of about one in thirty thousand or one hundred times more often than they should.  200 IQ children, instead of being non-existent, were found at the rate of about one in five hundred thousand.  In reality, there were quite a few children who were so smart that, statistically speaking, they should not have existed.

Spearheaded by David Wechsler, this anomaly was simply eliminated by edict.  He argued that mental age lost its meaning in adults.  People typically reach intellectual maturity at age 16.  There was no way, he argued, to prove that the excess high IQ people found would retain the same ratio superiority in adulthood.  With the Wechsler IQ test and now with essentially all IQ tests, scores are rendered as a 15 point standard deviation that is forced to conform to the normal distribution.  In other words, now there are no 200 IQ people and 180 IQ people are found one in three million.

Of course, these very high, anomalous IQs would need to correspond to some actual intellectual performance difference and preferably different in kind rather than degree.  Oddly, David Wechsler said, 'The genius (as regards intellectual ability) not only has an IQ of say 50 points more than the average person, but in virtue of this difference acquires seemingly new aspects (potentialities) or characteristics. These seemingly new aspects or characteristics, in their totality, are what go to make up the 'qualitative' difference between them."

When one questions the members of ultra high IQ societies, they often talk about a qualitative difference in intellectual processes and often mention 150 or so IQ as a line of demarcation between 'regular' people and themselves.  D.K. Simonton stated, "Even if the exceptionally bright individuals are able to target their use of language to the needs of their audience, the complexity of their ideas may be less accessible to listeners with IQs more than one standard deviation lower than their own."

In other words, there is a preponderance of evidence that the very high IQ people are not simply doing the same problem solving, just faster and more accurately, but rather are actually doing something different.  This qualitatively different intellectual activity is precisely what we would expect if we are talking about a different kind of species.  I, myself, have often been mystified that lines of reasoning that are absolutely compelling to me seem to elude those around me.

Now, we know that greater cranial capacity doesn't necessarily mean more intelligence.  There are factors of brain morphology and differences in neurotransmitters, etc.  That is why, when looking in the normal +/- 2 σ range, the correlation between intelligence and cranial capacity is weak.  However, what happens when we look in the range of 160 D15IQ?

Before we start, I will give a very short lesson in statistics.  Suppose that there is no correlation between IQ and cranial capacity.   Then, if a person has an D15IQ of 160, what is their likely cranial capacity?  It is 1,350cm³ just like everyone else.  Suppose that we find that they have a cranial capacity of 1,750 cm³ or about +4σ.  What is the probability that it is a random variation?  It is about 0.003%.  In other words we can be quite confident from just a single data point that at the extremes IQ and cranial capacity are highly correlated.

Suppose we find two people, selected at random, from the population of 160+ D15IQ and both have cranial capacity of over 1,750 cm³.  The probability that there is no statistical significance to this is 0.003%² or 0.00000009% or effectively zero.  Because of this and because I happen to know quite a few people with IQs over 160, I decided to find out if high IQ and large cranial capacity are correlated.
It was easy enough to find a formula that would estimate cranial capacity, they say within 4%.  It is ((L-1.26)*(W-1.26)*(H-0.63))*.365+406.  L=length of skull at greatest, W= width of skull at its broadest, H=height of skull vault from top of ear hole to top of head.  All measurements are in cm. 

My brain capacity is 2007 cm³ or well above the 1750 cm³ criterion.  I have a friend who is known for having one of the highest IQs on record.  I asked him to take a measurement, too.  His result was 1986 cm³.  Because of a height adjustment, (+/- 2 cm³/cm) his adjusted cranial capacity is actually very slightly more than mine.  I received two other submissions from people with 160+ D15IQs and cranial volumes of 1,936 cm³ and  1,700 cm³.  Because all were male, I used the mean of 1,406 cm³ and a standard deviation of 109 cm².

Even though the sample is quite small, the fit is so good that it has a very high statistical significance.  Furthermore, I received measurements from people with IQs below 160 but still very high.  From a metastudy of IQ/cranial volume databases I was able to perform a linear regression and determine that 'Cranial Capacity' IQ fits the line IQ=.2×V- 181.  With adjustment, for example, my IQ calculates to about 215.  What we find is that the linear regression fits ratio IQs!  In other words, it appears that David Wechsler was wrong.  The fat tail is valid and ratio IQs probably are a better measure of intellectual distance.

From this, we find that the IQ of a person with a 1,750 cm³ cranial volume is 169 R16IQ.  This is an equivalent of 153 D15IQ.  From this, we can estimate that H. macrocephalus comprises about 0.04% of the population.  While this is a very small percentage, it still represents about 2,800,000 people.  So, the population exists and it appears that they differ qualitatively, not just quantitatively, in their cognitive abilities.  However, are they engaging in assortative mating behavior that would support the theory of sympatric speciation?

Anyone who is familiar with high IQ societies knows that they represent a hotbed of assortative mating.  The reason for this is obvious upon reflection.  The intimate relationship of pair bonding requires a high degree of mutual understanding.  Research has, not surprisingly, found that the correlation of spousal IQ is substantially higher than for other personality traits.  Additionally, as IQ increases, its significance in the search for a mate increases.

Suppose we set a reasonable limit that IQ should be within 20 points in order to have an acceptable degree of mutual understanding in mates.  For a person with an IQ of 100, that means that the population of acceptable mates will be from 80 to 120 IQ.  In other words, 80% of the population will have acceptable IQs.  However, for the 130 IQ person, their range of acceptable IQ is 110 to 150.  This reduces the mating pool to just 25% of the population.  However, for our typical H. macrocephalus, the acceptable range is 133 to 173.  Now the mating pool has been reduced to less than 2% of the population.

Clearly, for this population, finding a mate with an acceptable IQ becomes the single biggest challenge in their mate selection.  Especially, over the last 50 years with the emergence of high IQ societies and the Internet, the tools for the exceptionally high IQ people to find one another for the purposes of mating have improved.  While no strong research has verified it, the evidence would suggest that the 153 D15IQ (s.d. 3.5) population is now a semi-isolated breeding population.

It is less than clear when, precisely, a population becomes a new species.  Clearly, if mating between two populations results in no offspring or no fertile offspring, the populations are clearly separate species.  However, in cases where fertile offspring result but the reproductive rate is below replacement, we have two species de facto if not de jure.  It is only a matter of time before they will, through genetic drift, become reproductively separate.  In the case of sympatric speciation there is an extended period of time when the two populations in theory could re-integrate.  However, by behavior, they will not and, as such, they comprise a nascent species.

While cranial capacity has been a hallmark of species classification in hominims, post cranial characteristics are also important.  I suspect that we will find traits such as a continued trend toward neoteny, gracility and longer lifespans to be likely differences.  We are seeing some evidence from the initial group suggesting these traits.    

The four cases also all exhibited very significant brachycephaly which is likely an adaptation to the difficulty of birthing a head of large cross-sectional area.  The more circular the head the less circumference for a given cross-sectional area.

Two of the four have noted very low basal body temperature.  This makes sense in that a normal sized brain consumes about 20% of the calories.  A brain 30% larger would consume 26% of the body's calories.  Lowering the prevailing body temperature, by cooling the blood, will assist in heat transfer.

Because we are looking at the extreme, four examples is actually sufficient to establish the statistical validity of the IQ/brain size correlation.  While not part of our circle, the very high IQ Chris Langan has commented on his enormous head, stating that it is over 3 standard deviations above the norm. The purpose of further research would be to determine any other traits common to the group and to measure the degree of assortative mating within the population.

The research, I believe, would be worthwhile and, with research assistants, I would undertake funding and executing the research.   The newly formed experiment.com may be able to fund this research.

An approximate estimate of ratio IQ from 15 point deviation IQ can be calculated by subtracting the deviation IQ from 128, multiplying the remainder by 1.8 and then adding back 128.  So, the TNS requirement of 148 is (148-128)x1.8+128=164.  This is approximately equivalent to the IQ that Cox, et alia estimated for Darwin, Locke, Bayle, Hegel, etc.  According to that set of estimates, it probably represents the floor for historical science and philosophical genius.


  1. Mr. Ferguson, it seems like you use a formula cranial capacity x 0 .2 -186.

  2. Your CC is 2007 cc. Using your formula (CC\5-181) we can come to your ratio IQ 220, but you report is 215, e.g. 5 points less. Any explanation would be appreciated

  3. In my case, the 2,007 is adjusted down to 1,980 cm³ because of my height. You add or subtract 2 cm³ for every cm ± 175 cm.

    1. Why adjust down? What has your height got to do with your cranial capacity?

    2. We are not entirely sure, but research shows that the cranial volume/IQ fit is best when we make the adjustment.

    3. Dear Michael
      Do you add 2 cm3 to the cranial capacity for every cm above 175 cm and substract 2 cm3 for every cm below 175 cm or vice versa?
      Best regards, Stefan

  4. Do have any more statistics on cranial capacity population frequency? What specific method did you utilize to convert ratio IQ to deviation IQ?

    1. http://hiqnews.megafoundation.org/John_Scoville_Paper.htm

  5. DIQ sd 15 vs "cranial IQ" sd 15, 4 cases, adult males, russians,
    cc\5- 181 formula is used
    142 - 156

  6. wais viq was taken in all cases

    but i got an additional info
    cases (in the same order)
    1/ jcti rix= 137
    2 jcti rix=142, wpt =40\50, the Simplex HRIQ test IQ=141
    3. WPT=21/50
    4/ none but wais viq
    next time i'll try to recalculate them having their heights

    thank you

  7. How did H. Macrocephalus assortively mate before Internet and high IQ societies?

    1. Not as well. It probably started with the emergence of elite universities that brought together people of extraordinary intellect. The Harvard/Vassar interactions, for example, was probably the start.

    2. Excuse me, Radcliffe. Vassar is Yale.

  8. If this turns out to be true, what about H.sapiens? Reading this is somewhat disheartening. It also reminds me of Clifford Simak's The city. Could it be possible that eventually every culture of affluance can evolve/bioengineer themselves into what they choose?

  9. Ivan, when I first analyzed this, my first reaction was to not write on it. The potential dynamic is not good. Sympatric speciation, especially with advancing technology, could result in dozens of nascent species.

  10. Thanks for the very interesting post. I was very surprised I did not make the connection between cranial capacity and results of intelligence measures, even after many years of thought about anthropology.

    I have a set of criticisms I would like to add, each of which perhaps could be overcome.

    1. Neanderthal's had a cranial capacity of approximately 1600 cm^3. Similarly, Inuit have a higher cranial capacity, and I've heard reports in the past that they also exceed 1600 cm on average. The size of the brain in both cases is not necessarily connected with greater measured IQ, but other brain functions (metabolic regulation and other). The natives of the north are not known for having extremely high intelligence, and I would not expect all Neanderthals to be 2-3G on average.

    2. The case of Neanderthal exhibits another point: advantage is tied to other aspects of development, like communication (I'm sure you are aware they were lacking anatomy for sophisticated oral communication). So it is possible that a large brain will not relate to the required advantageous brain organization.

    3. Studies concerning Einsteins brain seem to focus on organization rather than size. Differing organization would result in qualitative difference. One television show I saw recently reported that he had additional surface area due to an additional fold, in a portion of the brain focused on executive function. In that case, Einstein may have been adept at brining his ideas to fruition. My point with this, is that this might be a more significant divergence than increase in brain size, and that other factors like enhanced myelination or connectivity (or other) could be a bigger factor. Perhaps weight is also to be preferred over volume.

    4. I'm not sure there is a significant qualitative jump at 3G/4G or other. I think it is more likely that there is a continuous qualitative increase with increase in intelligence along a gradient. The wider the difference between two individual's IQ, the greater the apparent difference in quality. Put another way, what is the qualitative difference, and why believe it is at a particular threshold?

    5. I think you need to substantiate the normalization and SD of cranial capacity based on direct measure. In other words, skulls have to be emptied for a huge number of people. (or else some study based on MRI has to be done). I found my cranial capacity to be about 1740 (which is eerily close to what I would expect based on your notes), but it was a difficult task without a caliper. Small mis-measures result in huge differences when multiplied out. I had to work with an average of measures done by my wife. It doesn't feel like an accurate procedure (Mismeasure plus 4% could be *highly* significant).

    6. I worry if there is some egocentricity built into this. Obviously, I would like to think myself some improvement on humanity, as one of the people with a head at >1700 cubic centimeters. I also like to think there is a clear cut qualitative difference between myself and others, because they don't understand much of what I say to them. But it seems to me, that this just one area of differentiation and doesn't seem to be incredibly useful for procreation. It also doesn't seem to create a big enough barrier to procreation with everyone else, so what is to prevent my traits from being blended in with everyone else's? It is hard to see how the divergence would take place, especially since it depends on how social groups organize/collide in the future.

    1. I will respond to each of you six points separately. On the first.

      Both Neanderthals and Inuits are different breeding populations and, as such, cannot be considered in this analysis. In the case of Neanderthals the current belief is that the additional brain volume resided in the occipital lobe and was most likely dedicated to superior visual and kinesthetic abilities, not cognitive functions.

      As to the Inuit I could find no data, but as hunters, one might speculate the same explanation.

      I see no criticism here.

    2. As to #5, first, after you measure w and l, calculate wπ+2(l-w) and this should equal the circumference of your head. Then, skull vault should be measure by marking your height on a wall, marking the top of your ear hole on the wall, then, using a level, determine the vertical distance, using a level.

      I sometimes comment on this, and it is a part of the difference between IQ and intellectual sophistication. Some people can apply their intellect on an IQ test, but, in the real world, function at a much lower intellectual level.

    3. On #2, the location and conformation of the hyoid in H. neanderthalis was once thought to limit the range of sounds it was able to produce which, in turn, would limit the complexity of its language. However, since 1983, more complete specimens have shown this to be incorrect.

      Here I am speculating that IQ and brain size are highly correlated at the high end extreme and this suggests a relatively simple genetic cause upon which sympatric speciation could act. H. Neanderthalis is not relevant to that consideration.

    4. Again, on #5, a large number of data points are not necessary and, parenthetically, moderate mismeasures, while significant to the individual, are not important to the general conclusion. Let me go over this again. We know in the +/- 2 sigma range that IQ and brain size are mildly, but significantly, correlated. You are correct that other factors are more significant.

      Let's test the hypothesis that brain size is not a significant contributor to IQ at the very highest range. Then, if I found one four sigma IQ person, the probability that (s)he would also be a four sigma deviant in brain size would be 1/31,574 plus a small amount for the mild correlation which we can disregard for our purposes. If I find a second 160+ D15IQ person and they, too, have a four sigma brain size, the probability that this is coincidence is 1/31,574^2 or 1/996,917,476. In other words, with just two data points, we can conclude that in the +4 sigma range IQ and brain size are highly correlated. I don't need large populations to demonstrate it.

  11. Michael, thank you for taking the time to responding to my comments. After more thought, I think the entire article hinges on #5. I wish now to add some additional comments regarding your response above. I cannot shake the feeling that the methodology above is overly simplistic and that the multiplication of rarity at 4 sigma is too convenient.

    Point one: direct measurement can rule out head size as a merely secondary or loosely correlated consideration, or even rule it out from being worthy of attention.

    Supposing that intelligence actually correlates to differential arrangement of the brain, and overall grey matter in critial areas, then there is no need to measure cranial capacity at all. It is possible that one would not expect two people with the same cranial capacity to have the same quantity of grey matter in any particlar region. Variation in the intracranial space could be of greater interest (which is what I point to when I said people are interested in Einsteins brain arrangemetn more than its size). In a sense, to think one could make this determination (quantity of critical grey matter) by head size alone is to assume a phrenological perspective. It is not so implausible as that, but clearly visibility to the actual strucures is of vital importance. Cranial capacity does seem to correlate to overall intelligence, but maybe not to the extent that one could know the intelligence of a person at a particular capacity. If we find a stronger correlate we might lose interest in cranial capacity, and as a result, even more interest in overall head size.

    Point two: The human head size might be the result of a large number of possible variations that have nothing to do with intelligence

    Humans have a huge set of traits in which they can vary. One could use the same reasoning in your response to say that the odds that a person with *any* 4 sigma variation combined with 4 sigma intelligence is roughly 1/10^9. The odds that my big toe has a particular shape may even be greater than 4 sigma--but I cannot simply state that the odds that these are not correlated is 1/10^9. On the face of it that is absurd because toes have nothing to do with intelligence. Likewise although we might expect head size to be an important factor, it might be almost as irrelevant, even if that seems counterintuitive. It is an assumption we must justify using another approach. If we find that specific grey matter quanitities/densities/arrangements offers a higher correlation, and nearly causal relationship to intelligence, even at the *expense* of capacity in other areas, then we must forget measing the head in favor of measuring quantities of grey matter.

    I think the conclusion of this observation is that we need to forget measuring the head, and rely on MRI imaging and direct brain measurement alone (with a *huge* sample). We follow that path alone until we find the most significant relationships, and only then, see if those relationships correlate to overall head size. I'm willing to bet overall head size is only loosely correlated.

    A highly intelligent person may be born with variation in tissues or simply have mild hydrocephalus, or one of any innumerable traits which result in a large head that have little to do with intelligence.

    1. Here, let me explain it this way. Suppose that, as I calculated, IQ = .2V-181 but plus or minus 10%. In other words, a person with a V of 1,406 would have a calculated IQ of 100 but that is +/- 10 or 90-110. In other words, cranial volume would not be very predictive. However, take a person with a V of 1,750. Their calculated IQ is 169 but may vary 159 to 179. Now it is very significant, even though the percentage error is identical.

      As to the second point, I found a formula on the internet that claimed a 4% margin of error based upon measurements and MRI. So, the accuracy of the measurement is by citation, not by my work, either good or bad. I confess that I was surprised by the degree of accuracy claimed. Still on my measurement it is 1,927 to 2,087 which is still a heafty absolute error.

  12. I have been wondering what are the odds of two H.S. to give birth to a H.M. Have you thought and read about this one? Once Polymathica emerges what kind of further populational genetics' dynamics(?) do you expect?

    1. We can't know at this point, but I suspect that the macrocephalic allele is male sex linked. I suspect that because high IQ women don't seem to have the large cranial capacity. This actually portends greater concerned over speciation. Here is why. The H. macrocephalus gene passes down 100% father to son. However, in order to get within range of mutual understanding, dad is going to need to marry a woman with a very high IQ per CC of brain capacity. That means that over time, the volume difference will be married to more efficient brains and the phenotypic divergence will become greater over time.

  13. I am a male. By the measurement formula proposed here, my cranial capacity is quite average. My Terman Concept Mastery Test score correlates to a SB L-M ratio I.Q. of around 180. I hit the ceiling on the verbal part of the the WAIS III.

    So, I'll leave it to others to draw the appropriate inferences....

    1. I would think that a person with an R16IQ of 180 would immediately see the flaw in your implication. That you do not suggests that you are an extreme example of H. sapiens. With regard to IQ, H. sapiens and H. macrocephalus do have interpenetrating distributions.

  14. This comment has been removed by the author.

  15. Measured with calipers: 1868cc

    The head size to IQ formula gave a number eerily close to my average on high range tests.


  16. Hello Michael! Very interesting article.
    But it seems to me the formula you’re using overestimates the cranial capacity significantly, because of the way the height of the skull vault is measured. Here you can see the average dimensions for the head: https://en.wikipedia.org/wiki/Human_head#Anthropometry
    As you can see the height is measured there from the nasal root to the top of the head. If you enter the average male numbers in cm into your formula it gives 1414 as a result, which makes sense. From my experience though the top of the ear hole is way lower than the nasal root. Can you see if the dimensions of your skull and the others are still that extreme compared to the Wikipedia chart?

    All the best,

    1. Thanks, your calculation based upon my data renders 1,358 cm³ for males. The average of 1,406 suggests a differential bridge of nose and top of ear hole is ,52 cm. Seems reasonable.

  17. Interesting. My not very scientific measurements (60cm circumference, 16cm vault height) are putting me in the 2K cc plus range. That's consistent with my 5+ SD qualitative estimates of IQ.