• Your life and health are your own responsibility.
• Your decisions to act (or not act) based on information or advice anyone provides you—including me—are your own responsibility.


Big Brains Require An Explanation, Part VI:
Why Learning Is Fundamental, Even For Australopithecines

Our Story So Far (abridged)

  • By 3.4 MYA, Australopithecus afarensis was most likely eating a paleo diet recognizable, edible, and nutritious to modern humans. (Yes, the “paleo diet” predates the Paleolithic age by at least 800,000 years!)
  • The only new item on the menu was large animal meat (including bone marrow), which was more calorie- and nutrient-dense than any other food available to A. afarensis—especially in the nutrients (e.g. animal fats, cholesterol) which make up the brain.
  • Therefore, the most parsimonious interpretation of the evidence is that the abilities to live outside the forest, and thereby to somehow procure meat from large animals, provided the selection pressure for larger brains during the middle and late Pliocene.
  • A. africanus was slightly larger-brained and more human-faced than A. afarensis, but the differences weren’t dramatic.

(This is Part VI of a multi-part series. Go back to Part I, Part II, Part III, Part IV, or Part V.)

And here’s our timeline again, because it helps to stay oriented:

Timeline of hominin evolution

Click the image for more information about the chart. Yes, 'heidelbergensis' is misspelled, and 'Fire' is early by a few hundred KYA, but it's a solid resource overall.

It Doesn’t Take Much Selection Pressure To Change A Genome (Given Enough Time)

When we’re talking about the selection pressure exerted by the adaptations our ancestors made to different dietary choices, it’s important to remember that it only takes a very small selective advantage to make an adaptation stick.

The path to fixation

Remember, these are based on the most pessimistic assumptions possible.

The math is complicated, and I don’t want to drag my readers through it—but even under the most pessimistic initial assumptions (Haldane 1957), the following rules of thumb hold:

  • A mutation that confers a 10% selective advantage on a single individual takes, on average, a couple hundred generations to become fixed (present in 100% of the population).
  • Even a mutation that confers a tiny 0.1% selective advantage takes only a few thousand generations to become fixed.
  • Therefore, a 10% selective advantage would have become fixed in just a few thousand years—a fraction of an instant in geological time.
  • Even a 0.1% selective advantage would have taken perhaps 50,000 years to reach fixation—still an instant in geological time, and well beyond the precision of our ability to date fossils from millions of years ago.

I’m using approximate figures because they depend very strongly on initial assumptions and the modeling method used…not to mention the idea of a precisely calculated figure for “selective advantage” is silly.

Why is this important? First, because we need to remember that we are thinking about long, long spans of time. All of what we blithely call “human history” (i.e. the history of agriculture, from the Sumerians to the present) spans less than 10,000 years, versus the millions of years we’ve covered so far!

Second, and most critically, it’s important because we don’t need to posit that australopithecines ate lots of meat in order for the ability and inclination to be selected for—and to reach fixation. Even if rich, fatty, calorie-dense meat (including marrow and brains) only provided 5% of the australopith diet—and 4.9% of that advantage was lost due to the extra effort and danger of getting the meat (it doesn’t matter if you’re better-fed if a lion eats you)—the remaining 0.1% advantage still would have reached fixation in perhaps 50,000 years.

In other words: the ability and inclination to eat meat when available might have been a tiny advantage for an individual australopith…but given hundreds of thousands of years, that tiny advantage is more than sufficient to explain the existence and spread of meat-eating.

Most Mutations Are Lost: Why Learning Is Fundamental (Even For Australopithecines)

The flipside of the above calculations is that most mutations occurring from a single individual—even strongly beneficial ones—are lost.

Using the simple mathematical model, the probability that even a beneficial mutation will achieve fixation in the population, when starting from a single individual, is extremely low. J.B.S. Haldane calculated it at approximately 2 times the selective advantage—so even a 10% advantage is only 20% likely to reach fixation if it begins with a single individual! And for a 0.1% selective advantage, well, 0.2% doesn’t sound very encouraging, does it?

For those interested in the dirty mathematical details of simulating gene fixation, see (for instance) Kimura 1974 and Houchmandzadeh & Vallade 2011.

This low probability is because any gene carried by only one individual, or only a few individuals, is usually lost right away due to random chance while we’re on the initial part of the S-curve in the graph above. (As the number carrying the gene increases, the probability that everyone carrying it will die decreases.) So according to this naive model, we would expect individual australopithecines to have discovered meat-eating over and over again, hundreds if not thousands of times, before sheer luck finally allowed the behavior to spread throughout the population! Is that why it took millions of years to make progress?

Perhaps—but it seems doubtful. Meat-eating isn’t a single action: even if we assume that australopithecines were pure scavengers, it’s still a long, complicated sequence of behaviors involving finding suitable scraping/smashing rocks; looking for unattended carcasses; watching for their owners or other predators to return, which is probably a group behavior; grabbing any part that looked tasty; and using the rocks found earlier to help scrape off meat scraps, or to smash them open for marrow or brains. And hunting behavior is even more complex!

Of course, the naive mathematical model assumes that behavioral changes are purely under genetic control, and that individuals are not capable of learning. Since we know that the ability of humans to communicate knowledge by teaching and learning (known generally as “culture”) is greater than that of any other animal, it seems likely that the ability and inclination to learn from other australopiths was the primary mechanism by which our ancestors adapted a new mode of life that involved survival outside the forest—including meat-eating.

Note that chimpanzees can be taught all sorts of complicated skills, including how to make Oldowan stone tools—but they don’t seem to show any particular interest in teaching other chimps what they’ve learned.

Evidence That Increased Learning Ability Was The Key Hominin Adaptation During The Late Pliocene

We’ve just established that it’s very unlikely for a behavior discovered by one individual to spread throughout the population if it’s purely driven by a genetic mutation, even if it confers a substantial survival advantage—because the mathematics show that most individual mutations, even beneficial ones, are lost.

Here’s a summary of the physical evidence that our ancestors’ behavioral change was driven, at least in large part, by the ability to learn:

  • Body mass decreased by almost half between Ardipithecus ramidus (110#, 50kg) and Australopithecus africanus (65#, 30kg). Height also decreased slightly, from 4′ (122cm) to about 3’9″ (114cm). Clearly our ancestors’ adaptation to bipedal, ground-based living outside the forest didn’t depend on being big, strong, or physically imposing!
  • None of the physical changes appear to be a specific adaptation to anything but bipedalism, or to a larger brain case: faces became flatter and less prognathic, canines became shorter and less prominent, etc.
  • Despite a much smaller body, brain size increased from 300-350cc to 420-500cc. As brains are metabolically expensive (ranking behind only the heart and kidney by weight, and roughly equal to the GI tract—see Table 1 of Aiello 1997), this suggests that it was very important to conserve them.

Furthermore, it’s probably not a coincidence that bone marrow and brains are high in the same nutrients of which hominin brains are made—cholesterol and long-chain fats.

World Rev Nutr Diet 2001, 90:144-161.
Fatty acid composition and energy density of foods available to African hominids: evolutionary implications for human brain development.
Cordain L, Watkins BA, Mann NJ.

Scavenged ruminant brain tissue would have provided a moderate energy source and a rich source of DHA and AA. Fish would have provided a rich source of DHA and AA, but not energy, and the fossil evidence provides scant evidence for their consumption. Plant foods generally are of a low energetic density and contain virtually no DHA or AA. Because early hominids were likely not successful in hunting large ruminants, then scavenged skulls (containing brain) likely provided the greatest DHA and AA sources, and long bones (containing marrow) likely provided the concentrated energy source necessary for the evolution of a large, metabolically active brain in ancestral humans.

The learning-driven hypothesis fits with other facts we’ve already established. General-purpose intelligence is an inefficient way to solve problems:

…Intelligence is remarkably inefficient, because it devotes metabolic energy to the ability to solve all sorts of problems, of which the overwhelming majority will never arise. This is the specialist/generalist dichotomy. Specialists do best in times of no change or slow change, where they can be absolutely efficient at exploiting a specific ecological niche, and generalists do best in times of disruption and rapid change.” –Efficiency vs. Intelligence

Yet our hominin ancestors found success via greater intelligence rather than specific adaptations—most likely because of the cooling and rapidly oscillating climate previously discussed in Part I and Part IV. I’ll quote this paper again because it’s important:

PNAS August 17, 2004 vol. 101 no. 33 12125-12129
High-resolution vegetation and climate change associated with Pliocene Australopithecus afarensis
R. Bonnefille, R. Potts, F. Chalié, D. Jolly, and O. Peyron

Through high-resolution pollen data from Hadar, Ethiopia, we show that the hominin Australopithecus afarensis accommodated to substantial environmental variability between 3.4 and 2.9 million years ago. A large biome shift, up to 5°C cooling, and a 200- to 300-mm/yr rainfall increase occurred just before 3.3 million years ago, which is consistent with a global marine δ18O isotopic shift.

We hypothesize that A. afarensis was able to accommodate to periods of directional cooling, climate stability, and high variability.

The temperature graphs show that this situation continued. How did it affect our ancestors’ habitat and mode of life?

J Hum Evol. 2002 Apr;42(4):475-97.
Faunal change, environmental variability and late Pliocene hominin evolution.
Bobe R, Behrensmeyer AK, Chapman RE.

This study provides new evidence for shifts through time in the ecological dominance of suids, cercopithecids, and bovids, and for a trend from more forested to more open woodland habitats. Superimposed on these long-term trends are two episodes of faunal change, one involving a marked shift in the abundances of different taxa at about 2.8+/-0.1 Ma, and the second the transition at 2.5 Ma from a 200-ka interval of faunal stability to marked variability over intervals of about 100 ka. The first appearance of Homo, the earliest artefacts, and the extinction of non-robust Australopithecus in the Omo sequence coincide in time with the beginning of this period of high variability. We conclude that climate change caused significant shifts in vegetation in the Omo paleo-ecosystem and is a plausible explanation for the gradual ecological change from forest to open woodland between 3.4 and 2.0 Ma, the faunal shift at 2.8 +/-0.1 Ma, and the change in the tempo of faunal variability of 2.5 Ma.

In summary, 2.8 MYA is when things started to get exciting, climate-wise…and 2.6 MYA (the beginning of the Pleistocene) is when they started to get really exciting.

None of this is to say that the ability to learn was the only adaptation responsible for meat-eating: learning ability could easily have combined with other adaptations like inquisitiveness, aggressiveness, or a propensity to break things and see what happens.

Conclusion: A Tiny Difference Can Make All The Difference

  • Given the time-scale involved, a small selective advantage conferred by a small amount of meat-eating could easily have produced the selection pressure for meat-eating behavior to reach fixation in australopithecines.
  • Several lines of evidence—the mathematics of population genetics, the trends of australopithecine physical evolution, the ability of the nutrients in meat to build and nourish brains, and the increasingly colder, drier, and more variable climate—all point towards intelligence and the ability to learn (as opposed to physical power, or specific genetically-driven behavioral adaptations) being the primary source of the australopithecines’ ability to procure meat.

Don’t stop here! Continue to Part VII, “The Most Important Event In History”.

Live in freedom, live in beauty.