How did early humans reproduce with a low life expectency?

My girlfriend was studying the other night, and I glanced across a question that asked about the life expectancy of early humans. It basically stated that the average life expectancy 50,000 years ago was 10 years old. How much different were humans back then? Did we reach sexual maturity a lot younger, or did most of us just not reproduce?

Lots of births, lots of deaths. Contraception wasn’t such a thing back then.

eta: reminded me of a tv programme a while ago excavating a building near Hadrians Wall (the Roman wall between England and Scotland).

They found a grave with over a hundred skeletons of newborns. The presenter assumed it was some kind of ancient pedophile ring :smiley:

It was a whore house.

If 50% of the babies don’t make it to age 1, than the 50% that do would average between 19 and 20 for the whole population to average at 10. I suppose the infant death rate was even higher than 50%.

The average was low because large numbers died in infancy or early childhood, before reproducing.

Then, as now, if you survived to adolescence your risk of death dropped dramatically, and continued low for probably ten to fifteen years, before starting to rise again.

Thus, if you made it to sexual maturity, you had a pretty good chance not only of reproducing, but of sticking around long enough to raise your children to their own sexual maturity (though, of course, a fair proportion of your children would themselves die in infancy/childhood). Plus, extended families and tribal structures probably did more than nowadays to assist in raising the surviving offspring of those who died after reproducing, but before completing the raising of their children.

Must admit, I’ve never heard a LE as low as 10.

Well, it is a claim about the state of affairs prevailing 50,000 years ago, so I’m guessing there’s a fair degree of conjecture, estimation and finger-in-the-wind involved.


Ah. This makes sense. But how would they really know what the infant death rate was? Guessing, huh?

And lots of women in childbirth. If a woman dies at 20 and her husband at 70, the average comes out to 45. Add a dead baby, and your average is 30.

No guessing. Archeological remains and written records. Take into account that there have been consistent (if often destroyed by wars) records by religious authorities for thousands of years; centralized civil ones are newer (c. 1800), but the drops in infant and childbirth mortalities are more recent than either.

If I understand correctly, you are saying that if 50% of children die at 1, the other 50% have to die at 20 give an average life expectancy of 20.

It doesn’t work like that. Unless otherwise defined, average = mean. The mean is the sum of all ages divided by the sample size. So if we have 10 children, and 5 die at 1 and the other 5 die at 20, the average is:

(5+5+5+5+5+1+1+1+1+1)/10 = 3. Not 20.

If 50% of children die at 1, then the rest of the population needs to live to ~40 to give an average of 20. eg, for a population of ten children:
(40+40 +40+40+40+1+1+1+1+1)/10 = 20.5

If we throw in an occasional person living to 80, the mean rises quite a lot.

As noted earlier, any sorts of figures on life expectancy before the 20th century are mostly speculation. I’ve seen figures that suggest 3/4 of hunter gatherer children die from homicide or neglect before they turn two. I’ve seen similar figures for infants in the Industrial Revolution. Because each child drowned at birth skews the mean extremely heavily, but is almost impossible to detect in historical records, any figure on average life expectancy is largely guesswork. Most of the figures we have are based on funeral records or burial grounds. But people who kill their children at birth rarely record the birth, let alone the death, and even less often have a burial for the corpse.

The figure of 20 years seems entirely plausible to me, but we simply have no way of knowing for sure.

I think you got the first equation wrong. It’s five 20’s and five one’s, for an average of 10.5

Both of which are essentially useless when dealing with infanticide. Archaeological remains and written records don;t generally exist for children drowned at birth.

based on written records, we know that infanticide was widely practiced and socially acceptable in ancient Rome. But we have very few archaeological remains or written records of such infants. The records we do have consist mostly of offhand remarks that child x was killed at birth or records that the practice was widespread without reference to any specific numbers.

If infanticide was just as widely practiced in Victorian England, what written records or archeologically remains would you expect to find? If it was practiced just as widely in Australia in 1066, what written records or archeologically remains would you expect to find?

Written records or archeological remains only work for societies where infanticide was socially acceptable *and *where the victim was given the normal burial ceremony. I don’t think any such society has ever existed.




There aren’t any written records from 50,000 years ago, and precious few archaeological remains of human settlement. Written records and archaeological remains might give us a reasonable handle on human life expectancy 3,000-5,000 years ago, but ten times further back? We’re talking about a time when Neanderthals were still around, and the Cro-Magnons had yet to reach Europe. We don’t have agriculture, we don’t have domesticated animals, we don’t have permanent settlements - not even hamlets. All these things, and more, are going to happen before we get to a point for which life expectancy can be meaningfully assessed though archaeological remains and written records, and plainly they are all things likely to have an impact on life expectancy.

True, I hadn’t realized the OP was asking about 50K years ago. For that time yes, I wouldn’t so much call it guessing as fear extrapolation done by someone who can’t extrapolate.

Let’s run the numbers.

To simplify things we’ll assume everyone dies either at birth (age 0) or at the age of 20. This gives us an average lifespan of 10. We’ll also assume men and women have equal mortality rates. And finally we’ll assume everyone becomes fertile on average at the age of 12 and all the women have an average of one child every year.

So let’s start with a sample population of 1000 babies. 500 of them immediately die. The remaining 250 men and 250 women grow up. The 250 women will have an average of eight children apiece in their lifetime for a total of 2000 children. 1000 of these children will die at birth. The other 1000 children will survive to adulthood - 500 men and 500 women.

So it’s possible that the population would not only sustain itself but it would experience robust growth.

Nitpick: No, it doesn’t. To get the average of 10 you have to assume that equal numbers die at 0 and at 20. If three-quarters die at 0, then to get an average lifespan of 10 the other quarter must live to 40.

This matters, because in fact the likely mortality pattern is a larger number of individuals dying below the mean age, and a smaller number dying above the mean age but some of those living well beyond twice the mean age, or even three times the mean age. The result of this is that, in any generation, even if it’s true that the average lifespan is 10, there’s a significant cohort that survives to well beyond 20. This is crucial in accounting for the demographic sustainability of the community, and - equally important - the transmission of cultural knowledge.

On your model, yes. With the caveat that you assume children who are orphaned at age 8 or below all survive to adulthood.

But I doubt that your model bears much resemblance to any real-world human community.

Even fairly recently there was a lot of infant mortality (which is generally defined as deaths under one year). In 1950 the average rate of infant mortality around the world was 152 deaths per 1000 births. Now it’s 40 deaths per 1000. It varies a great deal around the world, from 1.80 to 121.63: