I have received reputable reports from the Department of Redundancy Department that this long post is long—over 4000 words, in point of fact.

Human evolution is a topic that interests not just researchers specialized in paleoanthropology, but also other scientists and the general public. A number of conflicting hypotheses have been put forward to explain why humans have become strikingly different from other primates. Most scientists in relevant fields (such as paleoanthropology, paleontology, ecology, evolution and human biology) have never published their views on the drivers of human evolution in general, nor on which of the proposed hypotheses on the origin of specific human traits they find most substantiated. No recent summary of the mainstream view among paleoanthropologists has been published either, so there is uncertainty as to whether scientists agree on the driving forces behind human evolution or not.

[…]

A survey was performed using an online form in early 2013. […] The focus was on journals of paleontology, zoology, ecology, evolutionary biology, and human biology. [...] The survey was performed anonymously, and all who responded did so voluntarily. After a few reminders had been sent, a total of 1,266 persons had submitted their responses to the survey.

[…]

The main results of our survey can be summarized as follows:

(1) There was no general agreement among the respondents on why any of the uniquely human traits have evolved: None of the proposed hypotheses was universally either accepted or rejected.

(2) For any individual trait, the percentage of respondents who found none of the hypotheses “very likely” was between >30% (bipedalism) and >65% (nakedness).

(3) In general, opinions on the credibility of the hypotheses were independent of a person’s background (gender, age, field of expertise, degree of scientific experience), but (paleo)anthropologists were clearly more critical than representatives of other fields.

(4) The hypotheses that mention adaptation to swimming or diving as an explanatory factor were found much less credible by (paleo)anthropologists and slightly more credible by human biologists than by biologists and representatives of other fields.

(5) Most respondents were critical about the aquatic ape hypothesis (AAH), but only a small minority considered it to be unscientific.

— Tuomisto, Tuomisto, and Tuomisto, “How scientists perceive the evolutionary origin of human traits: Results of a survey study” (2018)

The practitioners of Eld Science have failed once more to sufficiently concentrate their probability mass, and their temples are in disarray!

Beizutsukai, danketsu seyo!

It is a lesser power to explain why selection would favor any particular set of adaptations.

A greater power lies in explaining how to exapt our way, step by incremental step, from a common chimpanzee-like phenotypic and behavioral baseline, all the way to a human.

Obviously, almost everyone is thinking about sex already, but it seems to me that there are serious limitations on the abductive power of explanatory efforts invoking sexual selection on brains, this early in the process at least.

Once you have any model at all of how our ancestors broke out of the chimp equilibrium, surely you would see all sorts of places where you could explain parts of this history by means of sexual selection.

And yet, when your probability mass is as diffuse as that of Earth’s current scientific consensus (which we should not necessarily attribute to any specific Earth Scientist), this strikes me as a sign that no particular history has risen to the level of your attention, such that you could hope to explain any part of it using sexual selection.

That is, at least, if we are meeting the standard that we have set for ourselves, to explain not why you would evolve a human, but how.

By this standard, and with respect to Eld Science’s diffuse probability mass on most fine details, the phrase ‘sexual selection’ is a fake explanation, admitting us to the literary genre of evolutionary biology, while failing to constrain our beliefs about the evolutionary history of humans in a sufficiently powerful way.

But of course, sexual selection is a fine explanatory format! That is, provided that you have updated hard enough, to be able to execute the Form.

Such rapid phenotypic change as that observed in the hominin lineage could suggest many things, but in particular, it is strongly suggestive of positive feedback.

Thus, the phrase ‘sexual selection’ is also tempting to repeat in your head, and out loud, because a Fisherian runaway is some kind, any kind, of positive feedback, you think wordlessly, despairing of any origin story at all.

I am not throwing out the observation that some kind of positive feedback is probable.

In fact, I think this is simply true, and I will describe a model of this, but mostly via non-sexual selection with some sexual selection sprinkled in, and no sexual selection on brains in the highly specific, mate preference-based, peacock feather-derived sense about which everyone always seems to be thinking.

Eld Science shines brightest in the light of overwhelming empirical evidence, which allows us to inherit various beloved results of Settled Science, such as African origins and a chimpanzee-human last common ancestor baseline (these not always having been the scientific consensus), but in particular, bipedalism and endurance running adaptations, masticatory adaptations, life history adaptations, and endocranial volumes, which all fossilize in some sense.

Some of these fossils have also been the target of isotope and wear analysis, constraining our beliefs about the habitats and diets of fossil hominins, and paleoclimatological evidence informs us about the climatic conditions endured by these species. The stone tool record requires no introduction.

We should also consider the rich body of comparative evidence. In particular, the spotted hyena is an interesting comparison, persistence hunting and scavenging in open habitats, consuming marrow and brains, thermoregulating via nocturnality, converging on the same social system as the Hamadryas baboon, and coercing males into investing heavily in offspring, which could support the extended development of their bone-crushing jaws. They aren't lacking in brains either.

In general, this tells us we should not think overmuch of phylogenetic constraint until we have concentrated our probability mass a little more, for distantly related species can converge on very similar equilibria under similar selection pressures.

These cases constitute equilibria our model should not retrodict: we do not want to build this qualitative, quantifiable model and find that a quantified version of it actually retrodicts that female humans have pseudopenises and maintain rigid matriarchal hierarchies whence they coerce the male investment necessary to support the extended development of human brains.

After a particular history has risen to the level of our attention, phylogenetic constraint will become a wellspring of hypotheses about how humans are not approximately spotted hyenas, but not before.

Observations of ethnographic foragers are less important this early in the process, because to explain the end, we must explain the beginning.

I make a parody of many things, but not of these extraordinary empirical efforts.

We’re talking about the kind of people willing to trek through the most remote parts of Africa, to do crazy shit like tranquilize spotted hyenas with carbon dioxide-powered rifles and tag them with GPS and accelerometer collars.

And if you think that sounds hard, consider the fact that they had to endure the meek, ineffectual interrogation of the Ethics Committee first.

Without these clearly insane people, I would have no empirical evidence to do powerful abductive inference with. I can believe that there is room for improvement in the social epistemology of Eld Science the institution, without underestimating the power of individual Scientists, or their willingness to tranquilize me.

From the middle of the Miocene epoch, a relatively smooth, drying and cooling baseline trend began.

At the beginning of the Pleistocene, a cyclical humidity and temperature trend was imposed upon this baseline, consisting of glacial periods, usually lasting between 80,000 and 90,000 years, which were punctuated by warm, wet interglacial periods, usually lasting anywhere from 10,000 to 15,000 years. These climatic trends caused cyclical recessions of African forest cover over evolutionary time.

Homeostatic costs are higher in the canopy, so you should think not of bipedalism, but terrestriality itself, as a thermoregulatory behavior.

Arboreality also significantly reduces the sleep quality of apes, who must maintain enough awareness to prevent themselves from falling. This is why the human sleep duration mean is lower than the chimp mean, and why you sometimes awaken in shock, in a futile, involuntary effort to restore balance.

Early hominins reallocated sleeping time to moving time, and the thermoregulatory benefits of pelage were reduced, even if they were not immediately eliminated.

Baboon resource patches are less patchy and dense than those of chimps, because baboons are adapted to be far less selective about what they eat.

As Robin Dunbar has demonstrated via his behavioral-ecological models, for these reasons, feeding time is the primary constraint on whether or not a population of baboons can survive in a particular habitat. Chimps, on the other hand, are limited primarily by their moving time between resource patches.

Bipedalism reduces moving time, and requires only a quarter of the energy consumed by a comparable knuckle-walker, per unit of distance traveled.

This would have increased their caloric budget on a fixed diet, and they would have had the behavioral option to reallocate some of their feeding time to moving time. Resources are even patchier when the climate is dry, because aridity decreases the number of patches and their individual size, thereby increasing range size. Reallocating time in this way would have helped the earliest bipedal apes adapt to habitats with less forest cover than the rainforest habitats chimps usually occupy, particularly in the midst of dry glacial periods.

Bipedalism also reduces the solar cross section of the animal in smooth fashion as it becomes more erect, which allowed early hominins to reallocate time during the hottest part of the day, that otherwise would have been allocated to thermoregulatory rest. This source of selection would have been strongest during interglacial periods.

Even with the thermoregulatory benefits of terrestriality, it still may have been too cold at night to be as naked as later hominins.

However, once they started consuming significantly more fat, subcutaneous fat deposits would have provided the same thermoregulatory benefits as pelage, with more option value via a vascular mechanism, again providing a smooth gradient we can climb. If we explain the increased dietary fat intake, we will have explained subcutaneous fat and nakedness in one fell swoop.

But humans are not completely naked, which we must also explain. The head and shoulders both still lie in the solar cross section, so the hair is conserved to insulate the animal from solar radiation. The axillary and pubic hair is conserved to retain apocrine, as opposed to eccrine, sweat. The fine hairs on the rest of the body increase our tactile sensitivity to ectoparasites crawling on the skin.

Ectoparasitism is a constraint on nakedness, hardly the ultimate cause of it, perhaps speeding things up a bit by providing additional relative fitness benefits once significant amounts of pelage have already been lost.

The gait and respiration of quadrupedal apes are strongly coupled, but bipedalism removed this constraint, permitting the evolution of human-like laughter.

Laughter produces enough internal pressure to cause the same endorphin response as the primitive grooming behavior of chimps, compensating for any reduction in social bonding efficiency caused by improved ectoparasite detection.

Early hominins couldn't have told jokes without language, but they could have played together and surprised multiple conspecifics at once in humorous ways.

Laughter increased the number of concurrently reinforceable bonds, while the ancestral grooming mechanism is limited to the reinforcement of mere dyads, which reduced either grooming time or constraints on community size.

In all probability, community sizes increased, as this is one of the primary strategies by which apes minimize predation risk, which would have increased after adopting a terrestrial lifestyle.

The Dutch primatologist Adriaan Kortlandt reported an experiment in which he placed meat behind an electrically operated rotor, upon which some thorned Vachellia nilotica branches were attached. Every time a lion closely approached the bait, the rotor was momentarily operated, causing the lions to startle and jump away. This may be an exaptation of a canalized fear of porcupines.

Early hominins could have used Vachellia branches to ward off predators, just as modern bush pilots use them to protect their airplanes.

If the benefit of seeing over tall grass, a hypothesis paleoanthropologists find credible, has a qualitative threshold effect, then it cannot provide an incremental account of the evolution of bipedalism. However, even if such an effect exists, this would provide additional relative fitness benefits after selection for bipedalism was caused in some other way.

If bipedalism eliminated a couple of constraints, surely it could have created some.

It is, so far as I know, a mystery unto the paleontologists why exactly the australopithecines exhibited constant or increasing body mass dimorphism, and reduced canine dimorphism, at once.

This is exactly what we should observe if the evolution of early hominin sociosexual systems still obeyed the empirical laws discovered in extant nonhuman primates, while the evolution of canine dimorphism operated under novel constraints induced by adaptation to bipedalism.

Despite an apparent lack of overwhelming empirical evidence, theoretical analyses, or even a canonical citation, paleoanthropologists seem to implicitly accept the claim that bipedalism would have disrupted the operation of the ancestral canine honing complex, out of all the claims they could implicitly accept for once, so I won't look a gift horse in the mouth.

This significantly increased the cost-benefit ratio of male-male sexual and foraging competition. When canine displays ceased to be a credible threat, competitive situations escalated to violence more often. This destabilizing influence became very important later in the history.

Now let us consider the first radiation.

Paranthropus spp., Homo habilis, and Homo erectus coexisted for a while, so habilis is not a lucky snapshot of the ascent to erectus, if anyone is thinking that.

If we want this to be formalizable at the end, among other things we need to qualitatively obey the math of Lande and Arnold, and think of Paranthropus (the robust australopithecines), habilis and erectus (descendants of the gracile australopithecines) as three adaptive peaks on a vast fitness landscape, where a causal shooting gallery can change the number, position, elevation, and slope of these peaks.

If you're just walking in, the Paranthropus peak is to the left, and habilis to the right. The erectus peak is behind the habilis peak, and there are eldritch horrors on either side, so if you're headed to the erectus peak, you have to climb the habilis peak first.

But I can't imagine why you'd want to walk toward the erectus peak. Everybody knows it moves away when you walk toward it, probably because of the eldritch horrors, but we can't be certain.

I'm going to start with Paranthropus, because they are part of the mystery, and a good foil, and it will build suspense, and also maybe it will finally be over and I will never have to spill this much ink on them again.

Our model shouldn't just retrodict that gracile hominins developed sophisticated stone tool industries, it should also retrodict that no other animals did.

Most animals lack the manual and cognitive baseline to use stone tools at all, to say nothing of transmitting this innovation socially, and most of those with a sufficient baseline won't need time savings or access to hard targets badly enough for tool use to become an environmental invariant.

Baboons, for example, aren't picky enough, and frankly already have pretty robust jaws, and all this while developing rich, socially learned herbals (bodies of lay botanical knowledge), implying that they are perfectly capable of preserving, transmitting, and accumulating innovations under certain conditions. It is notable that this constitutes what one might tentatively describe as a body of declarative knowledge, as opposed to a procedural one.

Sea otters, crab-eating raccoons, and Paranthropus have robust masticatory adaptations that reduce the need for tool use on hard targets, although sea otters use tools on the very hardest, largest targets, and Paranthropus may have used them as well, despite, in all probability, never having developed stone tool industries as sophisticated as the Oldowan.

No raccoon has ever been observed using stone tools, despite having quite a suggestive manual and cognitive baseline, so raccoons may not actually have a sufficient baseline, and having robust masticatory adaptations, as in the case of Procyon cancrivorus, would only make the situation worse. Their hands may be too small, or their cognitive adaptations inadequate.

The wild chimp populations that have stone tool traditions live in habitats significantly drier than average. This at least demonstrates that you can start a stone tool tradition from a chimp manual and cognitive baseline in a dry environment.

Enculturated captive chimps have been taught to knap, as opposed to merely using hammerstones like their wild brethren, which strongly suggests that the primary constraint on accumulating stone tool traditions is cultural, at the chimp baseline.

So Paranthropus is an example of a species that had the baseline capacities for stone tool use, but already had path-dependent adaptations that minimized the relative utility of it, and the relative fitness benefits of adaptations that might have supported its sophistication. These factors stalled the accumulation of culture.

And fucking with obsidian isn't risk-free, you know. Blindness, hemorrhage, and septicemia are all on the table. You need a good reason to fuck with it, especially during the Lower Paleolithic.

Robust masticatory adaptations are developmentally and energetically costly, and they're costs the gracile australopithecines weren’t paying.

Robust australopithecines were already at least as competitive as chimps, if not as competitive as gorillas, which makes them bad learning targets for juveniles.

Adult males would have been aggressive, incapable of effective canine threat signaling, and wielding obsidian, and juveniles would not have the well-calibrated, canalized fear of this kind of situation required for obsidian to constitute a threat display per se, because the presence of obsidian was not yet an invariant, and it never would be, not for them.

All of this feedback made their central competitive dynamic even worse, as they fought over increasingly scarce food and mates in dwindling groups.

If anyone was reliably transmitting cultural knowledge about stone tools, it most likely would have been the females, but probably only via vertical transmission from mother to offspring, and if Paranthropus still exhibited male philopatry, then their daughters would leave the group, taking their cultural knowledge with them, which is good for spreading innovations but not for accumulating them within a single group. Their sons would remain, and still be bad learning targets.

If robust australopithecines had a sociosexual system closer to the chimp baseline, then there could have been some horizontal transmission among fraternal coalitions as well, but if they were polygynous like gorillas, then it seems like you wouldn't even get that. Horizontal transmission could occur in bachelor groups, but these would disperse as each member founded its own harem, or didn't.

Also keep in mind that the gigantic fossil baboon Dinopithecus was competing for the same fallback foods as Paranthropus at this time, and may have even occasionally preyed upon them.

So you’ve got a species under progressively worse climatic conditions competing with other species in a similar niche, some of whom even eat them at times, and a syndrome of path-dependent adaptations that kill culture in the cradle.

The kind of lesser daedra you'd find in Grimes’ plane of Oblivion, that’s Paranthropus.

The gracile lineage had less sexual dimorphism (and thus less sexual competition), far less robust masticatory adaptations, and their diets were more general. Importantly, effective stone tools would have granted access to more nutritious hard targets, which in turn would have reduced foraging competition.

For habilis, these targets could have included eggs, tortoise shells, seeds, nuts, hard fruits, and mineral licks, to give you an idea.

In particular, if habilines foraged for lacustrine and riverine mollusks, accessing their contents with stone tools, then we will have finally explained the elevated response of the human diving reflex with respect to the chimp baseline.

Indeed, availability of the brain-selective fatty acid DHA (docosahexaenoic acid) could have been a serious constraint on brain size, given how little dietary DHA we should expect from the chimp baseline or the extrapolated diet of early hominins, and how inefficiently both chimps and humans enzymatically convert AA (arachidonic acid) into DHA. 

Our model merely does not forbid this. Both dietary brains and aquatic animal meat lift this constraint, if it exists.

Once you obtain sufficient invariant caloric and micronutrient returns from the improved diet, you have raised the metabolic and developmental constraints on brain size.

The ancestral manual foraging system already smoothly trades off speed for accuracy, so in this very particular case, even simple scaling would suffice to reduce the risk of knapping injury by increasing the accuracy of strikes at a fixed speed. So we have an existence proof of fitness returns on cognition.

This is also when morphological adaptations began to emerge in the hands themselves.

Larger brains take longer to grow, so you have to extend the gestation period too.

Even this early, we already see a simple cycle, between life history, cognitive adaptations, and diet, to say nothing of culture.

To say something of culture, the invariant presence of obsidian obliterated the fitness benefits of Paranthropus-type violent sexual competition, and the improved diet reduced the need to engage in competitive foraging.

These are the primary incentives to disperse from the natal group in the first place. Female chimps do not want to compete with their sisters for mates or food, and beta-male gorillas do not want to be killed by alpha males.

For the first time, sons and daughters could remain with their mothers, and learn how to knap.

The sclerae became bright, to make it easier for juveniles to follow the gaze of adults in the midst of flintknapping. Competition had been reduced enough, and the fitness returns of social learning had increased enough, to render obscurity of the gaze obsolete.

Children could finally afford to take the risk of closely observing the knapping activities of adults, and of playing with their discarded flakes and cores. They could afford to be curious. This evolved into a stronger innate drive.

The inclusive genetic fitness benefits of being an asshole were falling precipitously, and bullies faced a serious risk of injury or death, even from individual targets, let alone coalitions.

The suppression of the ancestral dominance hierarchy had begun. Selection favored the evolution of an egalitarian instinct.

The modest baseline of male investment observed in chimps was increased, as males had more food to give, more time to spend, and each child became more expensive.

The complex of sperm competition adaptations also became less extreme than the chimp baseline as a result of this reduced sexual competition, and the copulatory plug was lost entirely, but the complex still remained more extreme than those of truly monogamous species, for both sexes still stood to benefit from extrapair copulations.

Females exploited and amplified this dynamic by concealing their ovulation, so that males would remain present as often as possible, in order to maximize paternity certainty.

Males in turn evolved a sexual attraction to females that was entirely insensitive to estrous cycles, an ignoble motivation to remain present, but good enough for paleoanthropological work.

Mating pairs could engage in social grooming through intercourse, much like the bonobos. Stable pairs were better positioned to raise increasingly more developmentally and metabolically expensive offspring.

Innovations could finally begin to accumulate, and once they did, children would have benefited from a longer juvenile period to grow larger brains, and to allow for the time necessary to learn.

They could not yet speak, as they formed pair bonds, had families, learned from one another, and made each other laugh.

They ran with the wind at their backs, toward a summit heretofore unexplored, perpetually receding unto the horizon, but it was only the beginning.

That's everything up to the point just before erectus appears, which is a whole ‘nother bag o’ worms.

Question, extend, formalize it. I'm way more interested in being directionally correct than getting every little detail right.

This is all about accumulating the first 27 bits of evidence.

I'll issue a bibliography at some point.