This article says life expectancy in ancient societies may not be as low as people tend to think. Many people are probably aware that child mortality brings down the average, but something I never thought about is how hard it is to measure the upper limit for populations that existed before any type of written records. There are people who specialize in digging up graves and doing exactly this, which seems a bit macabre.

We are all able to instinctively label people as ‘young’, ‘middle-aged’ or ‘old’ based on appearance and the situations in which we encounter them. Similarly, biological anthropologists use the skeleton rather than, say, hair and wrinkles. We term this ‘biological age’ as our judgment is based on the physical (and mental) conditions that we see before us, which relate to the biological realities of that person. These will not always correlate with an accurate calendar age, as people are all, well, different. Their appearance and abilities will be related to their genetics, lifestyle, health, attitudes, activity, diet, wealth and a multitude of other factors. These differences will accumulate as the years increase, meaning that once a person reaches the age of about 40 or 50, the differences are too great to allow any one-size-fits-all accuracy in the determination of the calendar age, whether it is done by eye on a living person or by the peer-preferred method of skeletal ageing. The result of this is that those older than middle age are frequently given an open-ended age estimation, like 40+ or 50+ years, meaning that they could be anywhere between forty and a hundred and four, or thereabouts.

The very term ‘average age at death’ also contributes to the myth. High infant mortality brings down the average at one end of the age spectrum, and open-ended categories such as ‘40+’ or ‘50+ years’ keep it low at the other. We know that in 2015 the average life expectancy at birth ranged from 50 years in Sierra Leone to 84 years in Japan, and these differences are related to early deaths rather than differences in total lifespan. A better method of estimating lifespan is to look at life expectancy only at adulthood, which takes infant mortality out of the equation; however, the inability to estimate age beyond about 50 years still keeps the average lower than it should be.

Archaeologists’ age estimates, therefore, have been squeezed at both ends of the age spectrum, with the result that individuals who have lived their full lifespan are rendered ‘invisible’. This means that we have been unable to fully understand societies in the distant past. In the literate past, functioning older individuals were mostly not treated much differently from the general adult population, but without archaeological identification of the invisible elderly, we cannot say whether this was the case in non-literate societies.