Finally an article that blasts the preposterous mythology suggesting that human longevity is a relatively recent phenomenon and primarily due to advances in medical technology.
http://www.livescience.com/health/090821-human-lifespans.html
In reviewing some of the comments made to the article it is clear that there is still a great deal of confusion surrounding the difference between "expectancy" and "lifespan".  The basic point in the article is that human life span is fundamentally unchanged over 2,000 years and quite possibly for a much longer period before that.

Life expectancy is a much abused term because it is really a statement about probabilities and has little to do with life span(1).  In other words, expectancy is determined by taking all the things that can kill you and averaging them out to determine what your "expectation" of survival is.  One obvious element is infant and child mortality, so if you're an adult then that aspect of your "expectation" no longer affects the probability of your surviving.

Similarly, if you live in a war zone your expectancy is obviously going to be different than if you live on a farm in Iowa, since the risks you are subject to will be different.  Expectancy is going to be related to the risks to which one is exposed and become relevant only in those cases where a comparable risk is associated with an individual.

The same can be said for diseases, where your expectancy is obviously going to be different if you've been diagnosed with cancer, for example, versus if you don't have cancer.  

While all these elements play a role, the mythology seems to have coalesced around an implicit assumption that it is modern technology, specifically medical, which is responsible for human longevity.  This creates the illusion that people living during other periods were simply dying at ridiculously young ages. It could certainly be argued that the risks (expectancy) might be greater to an individual during other periods of history, it would be incorrect to say that the human lifespan was affected.

Another oft mentioned explanation is that modern medical technology is responsible for increasing human lifespans, however this isn't accurate.  While technology has produced many new lifesaving procedures that affect expectancy there have also been higher risks created by technological advances.  An individual traveling in a car with hundreds of other drivers has a dramatically different risk than an individual riding a horse or walking.  Therefore one could argue that many of the medical advances are simply keeping pace with new and creative ways in which we can injure ourselves.

There's no question that surgical advances for dealing with organ failures, like heart or kidney transplants will affect outcomes, but these capabilities simply mitigate the probabilty of death without affecting the normal lifespan.  In other words, they may raise the life expectancy to a degree but do little with respect to overall longevity (except for the affected individual). 

Part of the problem in assessing life expectancy is in determining what all the risks are that a particular group is subject to. This becomes especially important when one considers the population size that is being explored and what's included.  Native American life expectancy was certainly impacted by European settlers, but if we included both populations in such an evaluation it would be misleading in interpreting human life expectancy.  Similarly it is misleading when all risks are simply grouped together under the "human" heading and then making statements about life expectancy.

So why does this matter?

The problem occurs when the life expectancy or probability of reaching a certain age becomes confused and interpreted as an absolute indicator of age.  This creates the false impression that human longevity has been increased, when what actually happened is that the probabilities of an early death have been reduced.  What the specific reasons for that reduction are would have to be separately analyzed, but they have nothing to do with lifespan.

Consider this contrived example where we simply state that 50% of the population lives to be 80 years of age.  This doesn't tell us anything about the other 50%, but life expectancy can give us some ideas.  To keep it simple, let's assume that the other 50% die at age 20.  This would give us a life expectancy of 50 years old.  What does this mean?  Absolutely nothing, because as we can see from our example, it really means that you have a 50/50 chance of making it to 50, however if you do, then you have a 100% chance of living to 80 since you're obviously part of the long-lived group.  The only thing that life expectancy showed was the one age that NO ONE would actually die at.  

Similarly if we improve things so our lower 50% now lives to age 40, the life expectancy will have increased to 60 years.  This has done nothing to the original 80-year old group.  It has simply shifted the probabilities of dying to a later period for the younger group.

Therefore it becomes reasonable to argue that advances in technology have probably contributed little to human longevity while it may have contributed (along with numerous other social and lifestyle changes) to improving the probabilities of surviving the earlier years. While this may not be an earth-shattering revelation, hopefully it will stop the type of preposterous misrepresentations indicated below:

"However, two-thousand years ago average life expectancy was less than 20 years or about 7,000 days. It is difficult to imagine, but most of our ancestors kicked the bucket before our modern legal drinking age."

"In 1796, life expectancy hovered around 24 years. A hundred years later it doubled to 48. In our modern world of air conditioners, hand washing and booster shots, you have a good chance of living 63 years, which is the world average. However, for those fortunate enough to live in a first-world country, lifespan jumps considerably."

http://www.longevitymeme.org/articles/viewarticle.cfm?article_id=11
As exemplified in Gender and Poverty in Nineteenth-century Europe By Rachel Ginnis Fuchs

"Infant and child mortality varied by region, social class, and family income, but the underserved and poorer populations of Europe experienced a higher likelihood of infant and child mortality than the more affluent.  The worn-out bodies and poor nutrition of pregnant poor mothers contributed to premature births and infant death."

"Infant mortality in Italy, Russia, and the German states was far greater with about 220 per 1,000 in Italy, 260 in Germany, and 280 in Russia."

This would suggest that it is social and lifestyle changes which have had a more profound affect on the probabilities affecting "expectancy" and that the suggestion that people were dying in their 20's or 30's is largely a myth.


(1) One obvious problem in these definitions is that even "old age" will cause you to die of something, which makes the entire discussion more difficult.  However, I'm going to use the assumption that what we consider "old age" to be the commonly accepted point at which we expect physiological failures to occur and lead to death. Even without precise values there are implicit understandings of what we mean when we say "old" or "young", however generally it refers to the longest lived 10% of a given cohort (maximum life span).