We discussed both demography and paleodemography; the former involves population patterns among current populations, while the latter refers to the study of some demographic parameters for past populations (which may be skeletal or fossil).

    Demography is the study of population.  At the very most basic level, demographers are concerned with births, deaths, and migration events, but they may study any characteristic that affects health, population size, or settlement patterns.  Demographers are interested in how many people live in a given area, why they live there, and who they are.  The area may be as small as a neighborhood or as large as a continent (or even the whole globe).  They look at age distributions, sex ratios, common causes of death, income distributions – anything that gives insight into a population.  Our lab exercise was most concerned with age and sex ratios, but we also reviewed some other common demographic parameters, which  helps elucidate the differences between demographic and paleodemographic work.  The parameters we discussed:
 

• Crude Birth Rate: Number of live births per 1000 per year; ignores sex and age composition of the population.  This rate may be high even if family size is small (for instance, in a very “young” country, where there are lots of young adults starting families).

• Sex Ratio: Ratio of males to females.  Typically around 105 or 106 at birth, dropping to 101 in your 30s, and finally to 44 in your 80s (in developed countries).

• Total Fertility Rate: Average number of children a woman will have during her reproductive years (ages 15-49).  A high number will have important consequences for generations to come.  This is not to be confused with fecundity – the biological ability to bear children.

• Crude Death Rate: Deaths per 1000 per year, again not considering age and sex.  All else equal, a nation with a large older population will have a higher death rate than a younger population.

• Infant Mortality Rate: Number of infants dying before age 1 per 1000 per year.  Often we also look at number dying before age 5, or between 1 month and 1 year.

• Life Expectancy at Birth: How long a person born in a given year can expect to live, assuming they follow the current mortality patterns of each age group upon reaching that age.  This changes for each individual over time – for instance, someone who is 30 has less chance of reaching 75 than someone who is 65, and in many developing nations where life expectancy at birth may be below 60, an infant surviving its first year immediately gains 3 or 4 years of expected life.

    Population Pyramids, the tool we used in the lab exercise, are graphical representations of a  group's population composition according to age and gender.  These are essentially stacked horizontal bar graphs, with the length of each bar representing the number of men or women in each age/gender category; they often indicate the percentage of the total population in each category as well.  Why do we care?  Population pyramids are very helpful in terms of identifying the type of population we're dealing with, as well as predicting its future growth and other demographic characteristics.  We also discussed how population pyramids can be used within the context of the Demographic Transition Model, a population dynamics model developed on the basis of European population trends over the past few hundred years.  There are four stages:
 

• Stage 1: Both birth and death rates are relatively high, with some fluctuation; the overall effect is that the population doesn’t grow much because effectively births equal deaths.  UK pre-1780; Ethiopia, etc. today.

• Stage 2:  For whatever reason (better health care, sanitation, etc.), death rates fall, while birth rates remain high.  This causes rapid population growth.  UK 1780-1880; Sri Lanka, Brazil today.

• Stage 3: Birth rates begin to drop, adjusting to lowered death rates; death rates may continue to fall, but not so precipitously as in Stage 2.  This takes time, as people must adjust to the idea that having large numbers of children is no longer necessary for security and may actually be disadvantageous.  The population total will often peak at this point, and by the end population growth is negligible. UK 1880-1940; Uruguay and China today.

• Stage 4: Birth and death rates are again close to equal, but much lower than in Stage 1.  This implies low or zero population growth.  UK after 1940; most other developed, industrial nations today.

    Paleodemography has existed by that name for about 30 years, if you trace the beginnings to Angel’s paper.  The goal is to obtain skeletal age and sex distributions, along with any helpful information on pathology and disease.  It is then possible to convert age at death distributions into mortality profiles.  The discipline really got underway when people began to construct life tables from cemetery samples, comparing with modern life tables.  The discipline assumes that any demographic disturbances stabilize quickly (i.e., effects of epidemics, etc., are short-lived).

    For paleodemography, of course, data on migrations (of the specificity needed here) are obviously not available, nor do we know the exact life spans or cause of death or number of children per woman.  Paleodemography is about assessing age at death, sex, health status in archaeological populations.  From this, as the Angel article states, we can get an idea of longevity, infant and child mortality, sex ratios, etc., although the effects of time are difficult to take into account – we often don’t know, for instance, how long a cemetery was used, so it is not possible to assume a single generation is being sampled (almost certainly this will NOT be the case).  Age and especially “fertility” indicators on skeletons are also not highly precise tools, so questions can be raised on that front as well.  Keep in mind, many of the inferences in the Angel article depend on the assumption that the skeletons available represent a valid sample of the total population.  The Angel article is an early (probably the first major) article on this topic, and I assigned it to give you an idea of what paleodemography has been intended to accomplish.  There have been a number of criticisms, even to the point of some critics calling the subject “dead.”

    Essentially, the criticisms are along two main lines: First, the uncertainty in assigning age to skeletal remains, and second, doubts about the methods used to obtain demographically useful information (such as rates of population increase, etc.).  Recently physical anthropologists have been devoting much thought to these issues, hoping to come up with a reasonable assessment of the utility of paleodemographic approaches.    They wish to identify questions that can be answered ONLY using bones, as well as questions that are in essence unanswerable, given that mortality samples are usually limited to a few hundred individuals in even the best circumstances.  Other variables that must be taken into account include time control (i.e., how much time does a cemetery or burial ground represent?) as well as preservation and recovery biases.

    These biases are serious stuff.  Cemetery samples, at the outset, are biased samples of the living, NOT random samples of everyone living at the time the cemetery was in use.  This is because even within age groups, you will find variation in frailty (susceptibility to dying) among individuals.  There are also biases inherent in skeletal material:

1. Infants are underrepresented because their bones are so fragile.
2. There is truncation at the upper end of the age range because of the limits of aging techniques.
3. Uncertainty and error in age and sex estimation.  This is one of the big criticisms.
4. The age distribution of the reference samples used in developing aging techniques has a confounding effect on age determinations – this is the other big criticism, that archaeological age distributions resemble the modern reference samples too closely.
5. Assumption of zero population growth.  This will influence age at death distributions – they may thus reflect fertility, not mortality.  This is also called the non-stationarity problem (i.e., we assume that populations were stable, but they were not).
6. Selectivity problem.  Individuals throughout a living population have an unequal risk of death; selective mortality can confound demographic and paleopathological interpretations.

    There has been a back-and-forth in the literature on these questions.  Paleodemography continues on, however, despite the criticisms, because we still can derive useful information – the problems/biases should be best read as a caution against over-interpretation of our results.