CROSS-CULTURAL ANALYSIS

CROSS-CULTURAL ANALYSIS by JUDITH L. GILLIES

Basic Premises

The basic premise of Cross-Cultural Analysis is that statistical cross-cultural comparisons are possible because cultures will, out of necessity of perpetuation, have some traits in common with each other within clusters of characteristic behavior or patterns of traits. The early basis for cross-cultural analysis was strongly based in the concept of cultural evolution. The premise of the cultural evolutionists was that all societies progress through an identical series of distinct evolutionary stages. Edward Burnett Tylor proposed that human cultures developed through three basic stages, savagery, barbarism, and civilization. Although this seems crude and ethnocentric, this was an advancement over the biological/theological belief that the more primitive societies of the world were at the stage of barbarism because they had fallen from grace. The hunters and gatherers, it was believed, had degenerated to their state, leaving them technologically and intellectually inferior to other cultures of the world. European society, especially Victorian England, which was seen as the top of the evolutionary scale. 

While Tylor (Primitive Culture 1871) was arriving at his concept of cultural evolution in England, Louis Henry Morgan was comparing cultures in America, to arrive at his own ideas of the levels of society. Morgan’s highest contribution to comparative studies was Systems and Consanguinity (1877). Morgan traveled and circulated questionnaires to collect information about kinship systems of Native Americans and other national groups in the United States. Morgan also used the terms savagery, barbarism, and civilization, but expanded on these to give us seven levels of cultural evolution. He determined his stages on the level of technological advancement, dividing each of the two lower stages into lower, middle, and upper stages. 

Morgan then classified cultures according to his system in his most famous book, Ancient Society (1877). 
lower savagery: began with earliest humanity—fruits and nuts subsistence 
middle savagery: began with discovery of fishing technology and the use of fire 
upper savagery: began with bow and arrow 
lower barbarism: began with pottery making
middle barbarism: began in Old World with the domestication of plants and animals / in the New World with the development of irrigation cultivation 
upper barbarism: began with smelting iron and the use of iron tools 
civilization: began with the invention of a phonetic alphabet and writing (Morgan 1877,12) 

Both Morgan and Tylor were more influenced by the ideas of social progress as asserted by Spencer than by evolutionary theories of Darwin. 

Cross-cultural analysts test hypotheses and draw statistical correlations based on the assumption of the existence of universal patterns. This process was greatly facilitated by the work of George Peter Murdock with the compilation of the ethnographies of over 300 cultures and 700 different cultural subject headings collected from ethnographies by Boas, Malinowski, and their students, among many others, who were not always professionals, into the Cross Cultural System, later known as the Human Relations Area Files. The trait lists of Cultural Universals, in “The Common Denominator of Cultures” in “The Science of Man in the World Crisis,” (Murdock 1945,123) were based on the HRAF (Ferraro 1992:74). Cross-cultural survey is “a comparative statistical study in which the “ tribe,” “society,” or “culture” is taken as the unit and samples from a world-wide universe are studied to test hypotheses about the nature of society or culture” (Naroll 1961, 221). The most famous example of this method is Murdock’s Social Structure (1949). 

Points of Reaction

The comparative method was used by early cultural evolutionists such as Morgan and Tylor in reaction against the degenerationists that placed hunter-gatherers and other less technologically advanced cultures in a class based on a supposed degeneration from perfection, which had made them less technologically and intellectually capable, inferior to the European societies of the 19th century. The development of the comparative method as used in Cross-Cultural Analysis was a reaction against the deductive reasoning of the Boasian tradition. Franz Boas was leading the majority of American anthropologists in the early 20th century. Boas had reacted against the comparative method as presented by Tylor before the turn of the century, and essentially, the comparative method had lain dormant in anthropology for 40 years. 

ADVANTAGES Levinson says that holocultural studies have six major advantages in the realm of theory testing concerning human culture and behavior: 
the sample covers “a much wider range of variation in cultural activities” than other studies based on single societies. 
this variation allows us to assume that it is more likely that “irrelevant variables” will not affect the results of these studies. 
range of variation allows researchers to measure degree and complexity of cultural evolution as variables in causal analysis.
certain variables, such as “language, religion, social structure, and cultural complexity” can only “be explained at the societal level.” 
holocultural studies are objective because the person who collects the data (ethnographer), and the theory tester (comparativist) are not the same individuals. This arrangement guards against the person collecting data consciously or unconsciously affecting the data in favor of a particular theory being tested. 
even the most rigorous holocultural studies are cost effective (1980,9).

DISADVANTAGES Levinson also points out four major disadvantages, although he states that these are outweighed by the six advantages listed above. They are as follows: 
holocultural studies often ignore the variability within a single culture for ease in coding uniformity, and the variation across cultures.
data used in holocultural studies is archival, and therefore lacks the sensitivity seen in case study work.
not all topics can be studied easily, and some perhaps not at all, because they are described poorly in the ethnographic literature. 
since the majority of the holocultural samples are compiled from small-scale societies, the large-scale societies are either unrepresented or under-represented (1980, 9-10). 

Leading Figures

Name and provide brief biographies of the principal scholars assoicated with the approach.

Sir Edward Burnett Tylor may be considered the father of the modern statistical cross-cultural approach to the study of culture for his paper, “On a Method of Investigating the Development of Institutions, Applied to Laws of Marriage and Descent” (1889). Tylor was born Oct. 2, 1832, into a well-to-do British Quaker family, and died. Jan. 2, 1917. He is considered the founder of social anthropology in Great Britain. Known for his research on culture, cultural evolution, and the origin and development of religion, Tylor never earned a university degree, but his position was earned through his research and writing. When he was 24, concern for his health led him to travel to America in 1856 and then on to Mexico. He returned to Great Britain and published his first book, “Anahuac: Mexico and the Mexicans, Ancient and Modern” (1861). Tylor’s unilineal view of progressive cultural evolution included the concept that earlier stages of development were exhibited by what he termed “survivals,” which were the single remnants of a paired set of ancient cultural traits that lingered on in more advanced cultures. He became keeper of the University Museum at Oxford in 1883, where he was a professor of anthropology from 1896 to 1909. His other major works include Primitive Culture (1871) and “Anthropology” (1881) (Kowalewski 1995). 

William Graham Sumner was born in Paterson, N.J., Oct. 30, 1840, and died Apr. 12, 1910 before the completion of his life’s major work, the four volume “Science of Society,” and the index for the volumes of comparative data. Sumner was a sociologist, economist, and Episcopal minister. As a Yale University professor (1872-1909), Sumner taught Keller and Murdock. Sumner introduced the classic concepts of Folkways and mores in Folkways (1906). William Graham Sumner was also the foremost publicist of the theory of Social Darwinism in the United States. Social Darwinists asserted that societies evolved by a natural process, like organisms. This theory contended that the most fit members of society survived or were most successful. This concept was roundly supported by political conservatism which argued that the most successful social classes also supposedly consisted of people who were obviously biologically superior. (Grolier Multimedia Encyclopedia 1995) The importance of this concept is that the basis for cross-cultural analysis was rooted in the concept of cultural evolution, and this was Sumner’s view of the process. 

George P. Murdock was born in Meriden, Conn., May 11, 1897, and died Mar. 29, 1985. Murdock, the most influential and important figure in 20th century cross-cultural analysis, was an American anthropologist known for his comparative studies of kinship systems and for his cross-cultural analyses of the regularities and differences among diverse peoples. During the time he was teaching at Yale (1928-1960) he developed the Cross Cultural Survey, in the 1930s-1940s, later known as the Human Relations Area Files. The HRAF is an index of many of the world’s ethnographically known societies. The HRAF is now available at over 250 institutional libraries worldwide, including a limited collection in Gorgas Library. Murdock’s publications include “Social Structure” (1949), “Africa: Its People and Their Culture History” (1959), and “Culture and Society” (1965) (Kowalewski 1995). Murdock descended from an anthropological ancestry opposing the traditional anthropological school of thought in America at the turn of the century headed by Franz Boas. Murdock hailed from the line descending from Tylor, Morgan, Spencer, Sumner, and Keller. Murdock was taught by A. G. Keller, and earned his Ph.D. under William Graham Sumner at Yale in 1925 (Levinson and Ember 1996:262). Sumner wished to create a comparative social science based on a “centrally located cross-cultural sample” (Tobin 1990:473). Murdock accomplished that, based on the original idea of Sumner’s central index. Sumner had begun the work of several volumes, most influential to the eventual work of Murdock in compiling the HRAF was the index completed posthumously by Sumner’s successor, A.G. Keller . 

Alfred Louis Kroeber was born in Hoboken, N.J., June 11, 1876, and died Oct. 5, 1960. He is often considered the most influential American cultural anthropologist after Franz Boas, who was one of his professors. He held tenure (1901-46) at the University of California at Berkeley. Kroeber was involved in regional cross-cultural study, comparing cultures to each other, not abstracted cultural traits, which he opposed. He advanced the study of California Indians and developed important theories about the nature of culture. Kroeber believed that human culture could not be entirely explained by psychology, biology, or related sciences, but required a science of its own. He was a major figure in the emergence of anthropology as an academic discipline. Kroeber published prolifically until the time of his death at the age of 85. His major works include “Anthropology” (1923; rev. ed. 1948); “Handbook of the Indians of California” (1925); “Configurations of Culture Growth” (1944); “Culture; a Critical Review of Concepts and Definitions” (1952), which he co-authored with Clyde Kluckhohn; and “Style and Civilizations” (1957) (Kowalewski 1995). 

Harold E. Driver, born 1907 -, was a Professor of Anthropology at Indiana University. His field research was concentrated in California and New Mexico. Comparative statistical methodology and culture area classifications were his areas of specialization. There is an excellent article by Driver in Readings in Cross-Cultural Methodology, entitled, “Introduction to Statistics for Comparative Research, which looks at such methods as chi-square and phi for the correlation between culture features. This article is written for the fairly unsophisticated statistician and is useful for comparative studies with other applications than just cross-cultural analysis. 

Clellan Ford, born 1909 -, was a professor of Anthropology at Yale and President of the HRAF. He took over the Human Relations program from Murdock. His field research areas were in the Northwest Coast of the United States, and the Fiji Islands. Comparative studies and human sexual behavior were his focus areas. 

David Levinson, born 1947-, and a prolific producer of anthropological encyclopedias, as well as cross-cultural work. Levinson has edited guide books for the use and understanding of the HRAF as well as books and articles that explain the studies that have been done utilizing the HRAF.

Other leading figures include many students of Murdock’s at Yale such as John and Beatrice Whiting, who conducted “The Six Cultures Project” with Irvin L. Child and William Lambert, and Melvin Ember, who is co-editor with Levinson and a major contributor to the “Encyclopedia of Cultural Anthropology” (1996). 

Key Works

Listed chronologically 

Morgan Louis Henry 1871 Systems of Consanguinity. Washington, D.C.: Smithsonian Institution. 
Kinship research based on interviews and questionnaires distributed across America to Native Americans and people of European descent. 

Morgan, Louis Henry 1963 Ancient Society. New York: World (orig. 1877).
In this book Morgan detailed the seven stages of society. The text contains a system for classifying cultures to determine their position on the cultural evolutionary ladder. 
Edward Tylor 1889 On a Method of Investigating the Development of Institutions: Applied to Laws of Marriage and Descent. Journal of Royal Anthropological Institute 18:245-269. 
Tylor was the first to attempt a statistical cross-cultural analysis with this paper, delivered to the Royal Anthropological Institute. 
Sumner, William Graham, and Albert Galloway Keller 1927 The Science of Society. New Haven: Yale University Press; London, H. Milford, Oxford University Press. 
Three volumes of entries of societies catalogued by Sumner. Volume 4 is the index of the entries. The fourth volume index had a great influence upon Murdock. 

Murdock, George Peter. 1945 The Common Denominator of Cultures. In The Science of Man in the World Crisis, Ralph Linton, ed. P. 123. New York: Columbia University Press. 
This is a listing of common traits among cultures, what Murdock called “cultural universals,” which could be used to determine what is common or variable among cultures in a holocultural study. 
Murdock, George Peter 1949/1968- Human Relations Area Files Microfilms International. Ann Arbor: University. 

The Cross Cultural System, which later became the Human Relations Area Files, was compiled by George Peter Murdock and colleagues at Yale in 1930s-1940s. It is a coded data retrieval system, which initially contained the ethnographies of over 300 cultures and 700 different cultural headings collected by the 1940s from ethnographies of Boas, Malinowski, and their students, among others, who were not always professionals (Ferraro 1992:74). The HRAF was originally produced on index cards, the HRAF Paper Files (1949), available on microfiche since 1968, and more recently available in a CD format. The entries to the HRAF increase annually, and subscriptions are bought by institutions on a yearly basis. Murdock wrote “The Common Denominator of Cultures” (1945). The cultural headings in the HRAF are partially based on the Cultural Universals Murdock sets forth in this work. 

Murdock, George Peter 1949 Social Structure. New York, Macmillan Co. 
In 1949 Murdock used the HRAF as the foundation for his book “Social Structure” in which he correlated information on family and kinship organizations around the world (Ferraro 1992:28-29). 
Murdock, George Peter 1957 World Ethnographic Sample. In “American Anthropologist” 59:664-687. 

Murdock, George Peter 1967 Ethnographic Atlas. Pittsburgh: University of Pittsburgh Press. 
Book. Classification of ethnographies. 
Whiting, Beatrice, and John W. M. Whiting 1974 Children of Six Cultures Cambridge , Mass.: Harvard University Press. 
This project was a far-reaching concept of the effect of child-rearing practices on adult behavior, which utilized cross-cultural analysis, but was based in the school of Culture and Personality. This project resulted in a book by the same name, but it really did not add to anthropological knowledge and exposed some problems concerning the use of inappropriate methodology for research that is not specific enough in its hypothesis. 
Rohner, Ronald P.1975 They Love Me, They Love Me Not: A Worldwide Study of the Effects of Parental Acceptance and Rejection. New Haven: HRAF Press. 
Levinson considers this book to be one of the important cross-cultural contributions of this century. 
Levinson, David, ed. 1977 “A Guide to Social Theory: Worldwide Cross-Cultural Tests” volume I - Introduction, New Haven, Connecticut, Human Relations Area Files. 
This is Guide Number One for the HRAF Theoretical Information Control System. In the Introduction to the Guide, Levinson states that it “is a new kind of information retrieval tool—an analytical propositional inventory of theories of human behavior that have been developed or tested by means of worldwide cross-cultural studies” (1977:2). There are five volumes of the Guide. This introductory volume contains a description of the Guide and tells one how to use it, including copies of the codebook that were used in the process of compiling the Guide. 
Levinson, David, and Martin J. Malone 1980 Toward Explaining Human Culture: A Critical Review of the Findings of Worldwide Cross-Cultural Research. New Haven, Connecticut: HRAF Press. 
Kinship, marriage, descent patterns, incest taboos, residence patterns, settlement patterns, religion, and aggression, among other cultural subjects, based on results obtained from holocultural studies. A bibliography and index are included. Levinson states that “this is a book about theories of human culture that have been tested holoculturally” (1980:5). 
Murdock, George Peter 1980 Atlas of World Cultures. Pittsburgh:University of Pittsburgh Press. 
Book. Includes an index. 

Principal Concepts
Regional comparison, is well represented by the works of Kroeber and Driver. This approach is an attempt to define classifications of cultures and to then make inferences about processes of diffusion within a cultural region (Levinson and Ember 1996:263). This is a question of how cultures relate to each other as whole cultural units, and comes more from the Boasian tradition.
Holocultural analysis, the more recent term, or worldwide cross-cultural analysis, has developed out of the ancestry from Tylor to Sumner and Keller, and then to Murdock. Levinson says that a holocultural study “is designed to test or develop a proposition through the statistical analysis of data on a sample of ten or more nonliterate societies from three or more geographical regions of the world” (1977:3). In this approach, cultural traits are taken out of the context of the whole culture and are compared with cultural traits in widely diverse cultures to determine patterns of regularities and differences within the broad base of the study. Both of these approaches are comparing cultural units, but their point of departure is the determination of what constitutes a unit of analysis. The comparative method as utilized in the worldwide approach presents a basic problem to anthropology, and to anthropologists. There is a conflict between the basic holistic approach of anthropology, with the evaluation of societies within cultural context, and the abstraction, comparison, and generalization of cultural traits in the comparative method as applied by Murdock (Winthrop: 44). 
Controlled Comparison is the approach toward smaller scale comparative studies. Eggan suggests the combination of the anthropological concepts of ethnology with structure and function, allowing the researcher to pose more specific questions on a broader range of subjects (1961, 125-127). Analysts are attempting to answer more specific questions in these research situations such as Spoehr’s study which examined the changes in kinship systems among the Creek, Chickasaw, and Chocktaw, and other regional tribes of the Southeast after their removal to the Oklahoma reservations. Spoehr detailed these changes with an analysis of the historical factors responsible for them and the resulting processes (Eggan 1961,125-126). � Holonational study is the study of universal traits within a national framework. 
Coding: data can be collected in two ways. Data can be coded directly from ethnographic sources, or from ethnographic reports in the HRAF files. The second method entails using previously coded data from coded ethnographic sources or from previous holocultural studies. Levinson and Malone suggest coding the dependent variables from HRAF files or ethnographic sources, and collecting independent variables from compendia of coded data. 

Methodologies

Not all Cross-Cultural analysts agree on the same methodology, but there are two main concepts: 
comparison is essential to anthropological research. To understand culture, societies must be compared. 
all theories, despite fads or current trends require testing. Without comparison there is no way to evaluate if presumed cause and effect are related. This relates to the logical “if” ---“then” inductive process. If cause is not present then the effect should not be present (Levinson and Ember 1996:262). 

The comparative method is a search for comparable culture patterns in multiple societies, particularly the comparison of cultural traits taken out of cultural context (Winthrop 1991: 43). There are two main goals of cross-cultural analysis. 
The first goal is to describe the range and distribution of cultural variation existent in the ethnographies recorded. 
The second goal is to test the hypotheses and theories that are proposed to explain the variation recorded (Levinson and Ember 1996:261). 

General requirements, that are stringently applied to the comparative method are: 
Scientific principles, method, and research design must be used. 
Explicit theory or hypothesis must be stated. 
Detail involved in study must be shown, allowing others to replicate study. 
Research must show measures are valid and reliable. 
Sampling procedure must be objective and clearly specified. 
Data must be made available to other researchers. 
Appropriate statistical tests must be employed. 
Results must be displayed for verification (Levinson and Ember 1996:261). 

Methods that are specific to Cross-Cultural Analysis are: 
Cases must be chosen from different cultures. 
Research aims must represent the entire ethnographic record or geographic region. 
Research must compare cases that agree with hypothesis with and without the presumed causes to verify if the presumed effect is associated with causes. This is a Static-Group Comparison (Levinson and Ember 1996:261). 

Accomplishments

Edward B. Tylor made the move into modern cross-cultural analysis with his statistical methodology explained in the school’s modern premiere paper, “On a Method of Investigating the Development of Institutions, Applied to Laws of Marriage and Descent” (1889). 

William Graham Sumner compiled and wrote most of the massive four-volume “The Science of Society” (1927) which was completed after Sumner’s death, including the index, by A.G. Keller (Harris 1968:607). 

George Peter Murdock developed the Cross Cultural Survey in the 1930s-40s at Yale, as head of the Human Relations Program. This beginning grew into the Human Relations Area Files, which is now available in over 250 institutional libraries both here and abroad. 

George P. Murdock combined the modern statistical method with modern ethnography, and statistical cross-cultural comparative method to create the HRAF. Murdock compiled the “Ethnographic Atlas, “ which was published in “Ethnology,” a journal that Murdock founded in 1962. This is an atlas of the 600 societies described on the basis of several dozen coded features in Murdock’s “World Ethnographic Sample.” 

Driver (1967) reanalyzed Murdock’s “Ethnographic Atlas” using the two basic approaches of statistical analysis for anthropology—the cultural traits as units of analysis, as proposed by Tylor and Murdock, and the approach suggested by Boas and Kroeber, by using societies or tribes as the units of analysis. Driver combined the concepts of these two approaches and came up with a more sophisticated method by inductively determining culture areas or “sets of strata” (Seymour-Smith 1986:61). 

Criticisms

“Galton’s Problem.”

When Tylor delivered his paper, “On a Method of Investigating the Development of Institutions, Applied to Laws of Marriage and Descent” (1889) to the Royal Anthropological Institute, Francis Galton, skilled in research design, was the presiding officer. Galton voiced what he saw as obvious flaws in the comparative methodology. This has ever since been known as “Galton’s Problem.” 
Galton observed. quite precisely, that since societies could acquire customs by borrowing them, it is possible that the number of cultural adhesions could be fewer than assumed. 
Galton asserted that the circumstances in which the adhesion occurred, whether by diffusion or by independent emergence, would effect the interpretation of the cases. 

Solutions proposed 
Not using multiple cultures within the same geographic region for worldwide cross-cultural analysis. 
More recently, mathematical anthropologists have devised a set of tests for “spatial autocorrection” based on language and distance in multiple regression analysis (Levinson and Ember 1996:263). 

Problems with the Comparative Method have been discussed by many anthropologists, including Murdock (1949), White (1973), Eggan (1954), Driver and Chaney (1973), and Hobhouse, Wheeler, and Ginsberg (1915).From these and other authors have emerged four major problem areas: 
identification and classification of the cultural items to be compared. What determines the scale of the items? 
the scope of the comparison temporally and spatially, or, generally, what is the scope of the degree of expected difference between the pairs of social units compared. 
the aims of the comparison. Is the intent of the comparison the formulation of scientific ‘laws’ of functional relationship, or is it the reconstruction of history from subsequent materials? Are the comparisons made for descriptive or analytic purposes? Is the style of argument inductive or deductive? 
the design of the comparison. How much control can be exercised over exogenous variation? How much attention is paid to sampling and statistical reliability?” (Hammel 1980:147-148). 

Additional criticisms of a more general nature were voiced by Marvin Harris. 
The ethnographies in the HRAF are not all of equal quality. 
the ethnographies are chosen for a higher quality, which may cause there to be a built-in bias toward certain areas or traits, limiting the value of statistical measures derived from the HRAF (Harris 1968:615). 
how can an outsider write with an emic view of the culture? He or she may not comprehend what is actually happening in any given situation. 

Addressing the inconsistencies in the quality of data in the HRAF, Murdock is said to have commented that there was a “robustness” in Cross-Cultural method. He was unconcerned about errors occasionally occurring in data because he did not think that they would harm the validity of a study. Naroll was more concerned with this problem and thought that errors would threaten validity. He proposed a process of analyzing data quality of the ethnographies already in use. Naroll suggested that researchers should rate ethnographies for certain qualities, such as the author’s command of the native language, and time spent in the field. This suggestion was carried through in an organized study and data quality of the ethnographies was found to effect results obtained in cross-cultural analysis in only a very few cases (Levinson and Ember 1996:263). 

Sources 
Bourginon, Erika 1973 Diversity and Homogeneity in World Societies. New Haven Connecticut: HRAF Press. 
Ember, Carol R. 1988 Guide to Cross-cultural Research Using the HRAF. New Haven: HRAF Press. 
Ferraro, Gary 1992 Cultural Anthropology: An Applied Perspective. St. Paul: West Publishing Company. 
Ford, Clellan Stearns 1967 Cross-cultural Approaches: Readings in Comparative Research. New Haven: HRAF Press. 
Hammel, E. A. 1980 The Comparative Method in Anthropological Perspective. In Comparative Studies in Society and History. Cambridge: Cambridge University Press 22:145-155. 
Harris, Marvin 1968 The Rise of Anthropological Theory: A History of Theories of Culture. New York: Crowell.
Kowalewski, Stephen A. 1995 George P. Murdock. In Grolier Multimedia Encyclopedia, version 7.0.2. Grolier Electronic Publishing, Inc. 
Levinson, David, ed. 1977 A Guide to Social Theory: Worldwide Cross-Cultural Tests volume I - Introduction, New Haven, Connecticut, Human Relations Area Files. 
Levinson, David, and Martin J. Malone 1980 Toward Explaining Human Culture: A Critical Review of the Findings of Worldwide Cross-Cultural Research. New Haven, Connecticut: HRAF Press.
Levinson, David and Melvin Ember, editors 1996 Comparative Method. In Encyclopedia of Cultural Anthropology. 1: New York: Henry Holt. 
Moore, Frank W. 1970, c1961 Readings in Cross-cultural Methodology. New Haven: HRAF Press. 
Naroll, Raoul 1961 Two Solutions to Galton’s Problem. In Readings in Cross-Cultural Methodology, edited by Frank Moore, pp.221-245. New Haven: HRAF Press. 
Rohner, Ronald P. 1996 From Conception Through Birth: Origins of the Society for Cross-Cultural Research. In Cross-Cultural Research, Vol. 30 No. 3, 265-274. August 1996, Sage Publications, Inc. 
Seymour-Smith, Charlotte 1968 Macmillan Dictionary of Anthropology. London: Macmillan. 
Winthrop, Robert H. 1991 Dictionary of Concepts in Cultural Anthropology. New York: Greenwood Press.