Science and the Study of Anthropology

            As a social science, anthropology is concerned with applying the principles of science to the study of social behavior in the same way that scientific principles are applied to other areas of human behavior.  However, before applying a scientific approach, it is necessary to understand both what constitutes a scientific explanation and what are the minimum requirements of doing science.

What is Science?

            Too often, science is both viewed and taught as an “organized body of knowledge.”  However, cookbooks, telephone directories and the Sears Catalogue are organized bodies of knowledge, but they do not represent science.  Rather than think of science as an object (a noun); it is more useful and appropriate to think of science as a verb (a process).  Science is, therefore a process, not a product.  Science is a method of investigation. In other words, “science is sciencing.”  The principal distinction between science and non-science, then, is not in the material examined but in the approach taken.  Science is the analytical method that attempts to test theories about the behavior of phenomenon and that rejects those theories that are either invalid or less accurate than competing theories.

It is not the case that some phenomena are inherently more scientific than others.  The study of the circulation of the blood or the motion of the planets is not a priori more scientific than the study of marriage, family, politics or religion.  All behavior can be studied scientifically.  The motion of stars and planets in the heavens can be approached scientifically, as in astronomy, or non-scientifically, as in astrology.  Similarly, the diversity of life on earth can be explained scientifically through the application of evolutionary theory or non-scientifically through recourse to Biblical creation.   We can also approach the detection and treatment of disease through the application of germ theory or by recourse to witchcraft.

            Most aspects of human behavior can also be examined scientifically.  Family structure, mating behavior, warfare, religious movements, infanticide, cannibalism and most other forms of human behavior can be examined and more precisely understood through the application of scientific research methods.  However, there are certain fundamental methodological requirements inherent to the scientific process, and those studies that do not adhere to these methods cannot be considered scientific.  Unfortunately, much of what passes for explanation in the social and behavioral sciences fails to meet the minimum requirements of science.  However, this reflects more a lack of rigor by many social and behavioral scientists than the inability of anthropology to be scientific.  The physical and biological sciences were themselves at one time all unscientific by today’s standards.  However, they all underwent radical changes in their methodology to become the sciences they are generally considered to be today.  The beginning of this shift took place in nearly 400 years ago with the change from an Aristotelian to a Galileian perspective in physics (see Aristotelian vs. Galileian).  This methodological shift occurred subsequently in other fields.  The Darwinian Revolution in biology may be seen as a shift from an Aristotelian to a Galileian explanation of biological diversity.  While some social and behavior sciences have moved in the direction of this shift, considerable resistance persists within the social and behavioral sciences regarding whether the application of scientific research methods to the study of human behavior is either possible or desirable.  The perspective taken in this course is that it is.  Indeed, in no field of inquiry have we witnessed a situation where a movement away from a scientific understanding of specific phenomena has led to a better, more precise understanding of those phenomena.

            One basic requirement for doing science is that the principal concepts used to explain behavior have both empirical referents and operational definitions:  that is, they must be definable in terms of empirically observable phenomena.  While we can define the brain and its functions as empirical phenomena, we cannot do the same for the mind.  Similarly, it is not possible to empirically define a culture, “Manifest Destiny”, Patriarchy, Id, Ego or Superego (Freud), Jungian archetypes, “the will of the people” “the Protestant ethic,” “the Weiteri complex” (the value of fierceness among the Yanomamo Indians) and many of the other explanatory concepts that are used in everyday speech, as well as in many professional social science explanations.  The relationship between phenomena must also be measurable, because scientific explanations make specific predictions about outcomes.  In order to evaluate scientific predictions, they must be testable and refutable.  Those predictions that are testable and refutable are considered scientific, while those that are not are rejected as either non-science or pseudo-science.      

            Another critical component of a scientific explanation is the Uniformitarian Principle: to the extent that a theory or principle is correct, it must be correct everywhere.  For example, if the law of gravity as it is currently understood applies within the Ettinger Building, it must also apply within the Moyer Building.  It, likewise, must apply on the African continent and on the moon, as well as during the Middle Ages in Britain.  There cannot be exceptions to the theory or principle; rather, exceptions cast doubt on the validity of the theory or principle.  The Uniformitarian Principle is fundamental to science; without it, there can be no science, because individuals could simply claim exception for those cases that did not fit the theory or for those whose exception were of importance to an individual’s worldview.  The belief in miracles, thus, violates the Uniformitarian Principle.  Romantic views of Native Americans, which see American Indians as living in balance with nature, also violate the Uniformitarian Principle.  They essentially claim that these populations were not subject to the same ecological principles that operate among all other human and non-human populations.  Similarly, those who claim that the rise of Christianity in Palestine during the first century C.E. was a unique event, different from the rise of all other religious movements throughout history, likewise violate the Uniformitarian Principle by claiming that sociological and anthropological explanations for the origin of religious movements can be applied to all but this one movement.

            As I just stated, science is based on making predictions.  The fundamental scientific statement is “given a, . . . then b;” that is, given one particular set of conditions defined as “a”, then a subsequent outcome, defined as “b” will occur.  Change the conditions in “a” and the outcome “b” will also change. Competing scientific explanations are evaluated in terms of which theory provides the most accurate and complete prediction of the various outcomes.  Galileo’s Law of Motion could better explain the various motions of a ball (or other object) than could Aristotle’s principles and, therefore, replaced an Aristotelian explanation.  Similarly, in time, Newton’s Law of Motion explained more different types of motion than did Galileo’s and eventually replaced the his law.

            A scientific theory is a deductive statement (arguing from the general to the specific) that begins with a set of assumptions, leads logically to a set of propositions, which in turn leads to specific testable hypotheses. (Hypotheses are simply propositions stated in testable –i.e., measurable- terms.)  A scientific explanation, thus, takes the following form:

Assumptions     →     Propositions     →     Hypotheses    →     Conclusions

            A simple example will illustrate the point. In constructing his Law of Motion, Isaac Newton began with the assumption that energy and matter are completely different phenomena.  He also assumed that only matter had mass (weight), and that only those objects containing mass could be affected by gravity.  This led him to propose that energy, which according to him had no mass, was not affected by gravity. Many years later, Albert Einstein presented his now-famous equation:  E=mc2, where E equals energy, m equals mass and c equals the speed of light.  Einstein’s theory, thus, began with a different set of assumptions than Newton.  In Einstein’s theory, energy did have mass; therefore, energy like other objects is affected by gravity.  These two theories, thus, lead to different predictions regarding the behavior of light in relation to a heavy object. The critical test of the two theories was performed in the 1920’s when light coming from distant stars was measured passing the sun during a solar eclipse.  Newton’s Law of Motion predicted that the light should not bend (because it would not be affected by the gravitational attraction of the sun), whereas Einstein’s theory predicted that it should bend.  Einstein’s prediction was the correct one, and this critical test marked the beginning of the replacement of Newton’s Laws of Motion by Einstein’s Theory of Relativity.  Other predictions were compared, and as a result of its better ability to predict the data accurately, Einstein’s Theory has come to be the mainstay of modern physics. 

Each test of a scientific theory constitutes an Independent Line of Evidence supporting or undermining the scientific community's confidence in that theory.  Einstein's theory does not explain everything, but it explains more than Newton's Theory; thus, physicists have more "confidence" in Einstein's Theory of Relativity than in Newton's Theory of Motion.  This same perspective underlies the acceptance or rejection of all theories in science.  Indeed, it is the only legitimate basis for evaluating competing theories and explanations.  Evolutionary theory dominates the biological sciences for this very reason.  There is simply no competing theory that systematically explains so many aspects of life on earth than does modern evolutionary theory (which is a combination of Darwinian and Mendelian principles).  If we begin with the assumption that all life on earth had a common beginning and "evolved" over time in response to adaptive pressures, we would make very different predictions about the data we would find than if we began with the assumption that all of the existing species were created separately and independently from one another.  The widespread existence of vestigial organs throughout nature would make logical sense from an evolutionary perspective; since change is gradual, we would expect to see some organs still present within individual organisms as those organs are only gradually lost to an evolving population through time.  However, the existence of vestigial organs makes no sense to (i.e., they would not be predicted by) a theory that begins with the assumption that all species were created separately and independently.  Such a theory cannot explain why vestigial organ would even exist.  Similarly, assuming that all species have changed through time and have a common origin in earlier species explains: (1) the existence of extinct species documented in the fossil record; (2) the pattern of genetic similarities and differences throughout the plant and animal kingdoms; (3) the minute similarities and differences in the chemical structure of chromosomes among such closely related species as humans, chimpanzees, gorillas, orangutans and gibbons; (4) the fact that the fossil record shows a clear sequence in the origin of fish, amphibians, reptiles and mammals, and many, many more observations.  It is for this reason that biologists have such high confidence in evolutionary theory and such low confidence in the Biblical version of creation.

This same method of evaluating alternative theories and explanations underlies the application of a scientific approach in sociology and anthropology.  Competing explanations need to be logically examined in terms of their underlying assumptions and of the propositions that logically derive from those assumptions.  Specific consideration also needs to be given to constructing testable hypotheses by which proposed or competing explanations can be evaluated.  This is no less possible in the social sciences than it is in the physical or biological sciences, nor is it any less necessary for determining the legitimacy of competing theories of human behavior.  In the absence of rigorous scientific methods, subjective interpretations prevail.

Yanomamo Warfare

A simple example will suffice.  Napoleon Chagnon, an anthropologist, studied the Yanomamo Indians living near the Brazil-Venezuela border. The Yanomamo have been actively engaged in warfare for many years, with approximately 25% of the men dying from warfare or warfare-related causes.   Chagnon has argued that the principal cause of Yanomamo warfare is a shortage of women.  The Yanomamo raid other villages, according to Chagnon, in order to steal women.  There are two glaring problems with Chagnon's explanation.  First, one of the reasons there is a shortage of women is that the Yanomamo practice preferential female infanticide.  Second, Chagnon states that the Yanomamo never raid another village for women that they are not already at war with.  He, therefore, cannot explain why these villages are already at war with one another that leads them to raid each other for women.   

Another anthropologist, Marvin Harris, argues that Yanomamo warfare has evolved as a population control mechanism.  According to Harris, increased warfare places a premium on having more males in a village and, thus, increases the desire of Yanomamo parents to give birth to boys.  This results in an increase in female infanticide; with less females in the village giving birth to children, the population growth rate declines.  One problem with Harris' explanation is that, according to Chagnon's data, the population of the villages he studied comprised about 200 people in 1900, compared to about 2,000 in 1970 when Chagnon studied them.  Their population growth clearly did not slow down, let alone come under control.

Besides the individual problems with their respective explanations, both Chagnon and Harris also fail to explain specific developments.   The level of warfare is very different among different villages.  According to Chagnon, some villages have not been involved in warfare for over 5 years, while others have been raided 25 times in a single year.  Similarly, the level of female infanticide varies in different villages.  In some villages the M/F sex ration for children under 12 is 157/100 while in others it is 121/100.

I have proposed an explanation that differs sharply from those offered by both Chagnon and Harris (see chart below).  My explanation begins with the assumption that, among other consequences, population growth leads to increasing competition over resources among neighboring populations.  As populations get larger, it becomes more difficult to obtain sufficient food by hunting, which enhances the prestige of a good hunter (who are males).   This increased competition also leads to greater conflict among neighboring villages and places, as has been suggested, a premium on the birth of males in a village.  Population growth also leads to the intensification of food production by males, which results in both patrilocal residence and patrilineal descent.  This further enhances the importance of males among the Yanomamo.  The shortage of women in a village now makes sense.  The availability of women declines as a result of preferential female infanticide and polygyny (one male having more than one wife).  Powerful men are able to make more alliances with other families through marriage.  My "theory" leads to specific testable hypotheses.  Those villages with the higher population density (P/R) should be the same villages that practice the highest rates of female infanticide and that are the most heavily involved in warfare ("given a, . . . then b").  Villages in the center of Yanomamo territory, according to Chagnon, are twice as large as those located on the periphery of this region (50-250 people in the central villages compared to 25-100 among the peripheral villages).  They are also located much closer to one another.  Population density is, therefore, much higher among the central villages than among the peripheral villages.  According to my model, the central villages should be the most actively engaged in warfare, practice more frequent female infanticide and have the highest sex ratio imbalance.  This is exactly the situation described by Chagnon.

Thus, examining the logic of Chagnon's, Harris' and my explanations, it is possible to determine in a scientific way which "theory" gives us the best explanation of Yanomamo warfare.  The fact that my model also explains the spiral of increasing warfare among the Plains Indians during the 19th century adds to my "confidence" in its validity.  The Plains Indian situation represents an "independent line of evidence" with which to test my explanation of the causes of warfare among contiguous populations.

Overpopulation

Another example, a little closer to home, will also illustrate the both the ability and the importance of applying careful scientific methods in social science research.  We hear and read constantly about overpopulation in developing countries and how population growth is the cause of these countries' underdevelopment.  However, this is only the belief of those demographers who accept what is known as neo-Malthusian population theory.  For neo-Malthusians, population growth results in more mouths to feed and, thus, less investment in economic development, as more energy must be spent just trying to feed the growing population.  Neo-Malthusian population theory has become accepted nearly as dogma in the West and forms the cornerstone of most environmental studies programs in the U.S.. There is a problem, however, with neo-Malthusian theory.  It does not explain the facts.  In fact, the facts largely contradict the theory.

For example, Sub-Saharan Africa had nearly 100 million people in A.D. 1500, about the same as Europe at that time (even though the African continent is perhaps 7-9 times the size of Europe, making population density in Africa only a fraction of what it was in Europe).  However, while Africa’s population declined precipitously following European contact, the populations of Europe and European derived countries (the U.S., Canada, Australia, etc.) increased dramatically.  Indeed, Africa’s population has only rebounded in the late 20th century, but it still comprises only 11% of the world population compared to 17% in 1500.  These facts raise several issues relating to the role of population growth in African (and other Third World countries') underdevelopment.  Who prospered more over the past 5 centuries:  the people living on the continent whose population size decreased (Africa), or the people living on the continent whose population size increased (North America and Europe)?  Africa's proportion of world population has actually decreased during the past 5 centuries.  Given the logic of neo-Malthusian population theory, we should expect that Africa as a region would have shown greater not lesser economic development than other world regions.  Also, since Europe’s population has increased from 100 million in 1500 to about 400 million today, shouldn't Europe be showing the negative signs of population growth?   Indeed, the population densities of Germany and Great Britain are about twice that of Nigeria, nearly 5 times that of either Ethiopia or Kenya and 30 times that of the Congo!  Moreover, per capita energy consumption in Germany and Great Britain is almost 20 times that in the Congo, 42 times that in either Kenya and Nigeria and 240 times that in Ethiopia.

The U.S. population has increased from 75 million in 1900 to nearly 300 million in 2000  --a phenomenal 300% between 1900 and 2000!  Yet the U.S. is perhaps the most prosperous country to have ever existed.  In addition, Japan (the second most prosperous society in the world) experienced a growth in its population from about 35 million in 1900 to over 128 million today  --an increase of over 250%.  Moreover, Japan contains a population 1/3 that of the entire U.S. living on a land surface smaller than California.  Why aren't the Japanese starving?  According to neo-Malthusian population theory, we should see widespread starvation in Europe, the U.S. and Japan.  What we have instead is a situation where the most prosperous countries in the world are those that have experienced some of the highest population growth during the period of their most rapid economic development and that have achieved among the highest population densities in the world, while the poorest countries in the world are those that have historically experienced population decline and today have among the smallest and least dense populations in the world.

An alternate non-Malthusian theory of population growth sees population growth, not as a detriment to economic development, but rather as a principle driving force contributing to economic development and social evolution.  Population growth, according to non-Malthusian population theory creates a demand for innovation  -- the development of new technology, new ways of organizing labor, developing new methods of food production, finding new sources of energy, etc.-- which together not only enable the population to survive and adapt but also to evolve socially and politically.  Non-Malthusian population theory certainly explains why the U.S., Europe and Japan have prospered while the countries of Africa and elsewhere have not better than the competing neo-Malthusian theory.  It also explains why China, the worlds largest population is currently experiencing the highest rates of economic growth ever recorded and why it is having to close hundreds of thousands of farms because they are producing more food than the country needs.  But it also explains why throughout history it has always been those regions that experienced increasing population growth that became the centers of civilization and that ultimately were the regions that prospered and dominated the peoples around them.  Civilization has always evolved among growing populations, not among societies whose populations stagnated or decreased in size.

Gender-Related Behavior

This same methodology can be applied to examine gender-related behavior.  Numerous conflicting accounts and explanations exist regarding the causes and consequences of gender-related behavior   It will be the purpose of this course  --and the responsibility of each student--  to evaluate the various competing claims and explanations of gender-related behavior in terms of the degree to which they possess scientific validity.  Some explanations are blatantly false, as with the American Association of University Women's (AAUW) fraudulent self-esteem study.  There are, however, many proposed explanations of gender-related behavior whose methodological problems are not so obvious or whose validity is not so clearly questionable.  Mary Pipher's book, Reviving Ophelia, is a case in point.  The following is a quote from Pipher's book (pp.27-28):

Our minds, which are shaped by the society in which we live, can oppress us.  . . .

Adolescent clients intrigue me as they struggle to sort themselves out.  But I wouldn't have written this book had it not been for these last few years when my office has been filled with girls:  girls with eating disorders, alcohol problems, posttraumatic stress reactions to sexual or physical assaults, sexually transmitted diseases (STDs), self-inflicted injuries and strange phobias, and girls who have tried to kill themselves or run away.  A health department survey showed that 40 percent of all girls in my midwestern city considered suicide last year.  The Centers for Disease Control in Atlanta reports that the suicide rate among children age ten to fourteen rose 75 percent between 1979 and 1988.  Something dramatic is happening to adolescent girls in America, something unnoticed by those not on the front lines.

Many of the pressures girls have always faced are intensified in the 1990s.  Many things contribute to this intensification:  more divorced families, chemical addictions, casual sex and violence against women.  Because of the media, which Clarence page calls "electronic wallpaper," girls all live in one big town  --a sleazy, dangerous tinsel town with lots of liquor stores and few protected spaces.  Increasingly, women have been sexualized and objectified, their bodies marketed to sell tractors and toothpaste.

Pipher displays a clear bias in her view of American society and in her analysis of adolescent emotional problems.  She believes that American society "oppresses" girls.  However, she never provides a meaningful (i.e., operational) definition of the word "oppress".  Without a workable definition of oppression, it is impossible to accept Pipher's claim that American society oppresses women.  Pipher's bias also leads her to make several unsubstantiated statements that are (1) based on a very weak argument and (2) lack even the minimum data needed to support them.  For example, in the above quote, Pipher makes three sequential statements that are made to substantiate her claim that girls are in trouble:  (1) She first states that "40 percent of all the girls in my Midwestern city considered suicide last year;" (2) She then states that "the suicide rate among children age ten to fourteen rose 75 percent between 1979 and 1988;" and she then concludes (3) that "something dramatic is happening to adolescent girls in America." 

Pipher's argument has several problems.  To begin with, she only gives us the percentage of girls in her town that "considered suicide".  Without comparable data on boys, the implication of this figure for the relative mental health of girls cannot be determined.  Furthermore, the logic of her three sentences does not follow.  She begins with a statement about girls considering suicide; follows this with a statement about the suicide rate of children (not girls) and then concludes with a statement about what this means for girls. Her concluding statement does not follow from her first two statements.

 Pipher first needs to present comparable data about boys and girls and must show that the data clearly indicate a worse situation for girls than boys if she is to be able to conclude that "something dramatic is happening to adolescent girls in America."  Furthermore, it is reasonable to ask why Pipher chose to present only the data she used when this data is clearly inadequate and when the data she needed to substantiate her claim is so readily available.  Why, for example, does she only present suicide data for children ten to fourteen when discussing adolescents?  Does adolescence begin at age 10 and end at age 14?  Why doesn't she give us suicide data up through, say 19 or older, and why doesn't she give us both male and female suicide data?  Could it be that this data completely undermines her whole argument?

The fact is that the suicide data completely contradict Pipher's argument.  As the table below clearly indicates, there are four times as many male suicides as there are female suicides.  Furthermore, the discrepancy has increased through time:  2.42   times in 1970 compared to 4.09 times in 1993.   In addition, the proportion of male suicides increases with age:  from 2.4 times as many male suicides as female suicides at age 10-14 to 6.1 times as many at age 20-24.  Similarly, the homicide rate for males is much greater than that for females.  In 1993 over 20,000 males were murdered compared to less than 6,000 females.  So the violence towards males (as measured by the homicide rate) is about 3.5 times the violence towards women.

It is not only reasonable to ask  --it is fundamentally important to ask--  to what extent Pipher's failure to present sufficient data was influenced by her biases.  She was very selective in the presentation of the facts.  She presented only data that supported her argument and ignored the data that did not, even though that data is both voluminous and readily available (I was able to obtain homicide and suicide data by age and sex from 1950-1993 from the Trexler library in 20 minutes.)  The data clearly shows that the increasing violence and suicide has nothing to do with America oppressing women, because it is primarily men being murdered and committing suicide.  Therefore, all of her statements about a hostile media exploiting girls and sexualizing them are simply irrelevant to her thesis that these pressures on girls has increased over the last 30 years resulting in a high rate of female suicide and emotional problems.  These statements simply don't explain the data.  If the factors she emphasizes were the cause of both violence and suicide, we should see more women dying and committing suicide than men ("given a, . . . then b."; see also Hill's  "Dose Response").  We should also see the relative number of women being killed and committing suicide increase over the past 40 years faster than the number of men.  However, the opposite is the case.  The male/female suicide rate in 1960 was 2.57, whereas the male/female suicide rate in 1993 was 4.09.  Similarly, the male/female homicide rate in 1960 was 3.17 compared to 3.56 in 1993.  In both cases, there has been a faster increase in the death of men since 1960.

Clearly, there has been an increase in the number of boys and girls, as well as men and women, dying prematurely, and it very likely is due to the increasing social pressures on both adolescents and adults that have occurred over the past 40 years.  However, the cause of this trend may have nothing to do with the kind of popular claims promoted simplistically by Pipher and others.   Social policies based on publications like Reviving Ophelia are, therefore, doomed to failure because they obviously target the wrong factors.  Consequently, it is reasonable to ask why such research continues to be popular despite the inability to explain the data.  Might this research persist and gain an audience for sociopolitical reasons rather than for scientific ones?

The above examples all illustrate the importance of adopting a careful scientific methodology to the study of human social behavior.  This is what thinking critically means.  The media, the airwaves, the Internet and even professional journals are filled with information and research related to contemporary issues.  Some of it is good, but unfortunately much of it is bad  (I would even argue that most of it is bad).  I have tried here to introduce you to some of the very minimum analytical methods you need to follow in order to evaluate the information that is available.  I have also tried to give you a several examples covering a variety of topics so that you can see that the scientific methodology is as applicable to the study of sociology, anthropology and the other social sciences as it is to physics, chemistry and biology.  You should keep these ideas in mind while you are doing the readings for the course and while writing any essays or research papers for the course  --as well as when you are reading newspapers, journals and magazines.  I have assigned many different readings throughout the semester, some of which are good, some of which are not so good, and some of which are really bad.  It is your ability to use critical thinking skills that I will be examining and testing throughout the semester; not your ability to memorize the information presented.