Human Relations Area Files Research Paper

Human Relations Area Files, Inc. is the nonprofit membership organization that produces and distributes the HRAF databases. It also organizes and edits encyclopedias and sponsors the quarterly journal Cross-Cultural Research: The Journal of Comparative Social Science. HRAF is an international consortium, with nearly 200 member institutions in 25 countries; it is also a financially autonomous research agency of Yale University, where it was founded in 1949. The HRAF databases, which are produced now only in electronic format, are accessible only through HRAF member institutions. The first and larger database is the HRAF Collection of Ethnography, which has been building in paper, microfiche, and electronic format for more than 50 years. It includes indexed full-text materials on more than 365 cultures around the world. The electronic version, eHRAF: Ethnography, has been building since 1994 and as of 2000 covers a geographically stratified random sample of 60 cultures, plus assorted ethnic groups. The second data-base is the HRAF Collection of Archaeology, which has been building solely in electronic format since 1999. As of 2000, eHRAF: Archaeology includes indexed full-text materials on 20 randomly selected major archaeological traditions, and many more subtraditions and sites, around the world.

1. Uniqueness Of The HRAF Databases

The task of finding particular kinds of information in the HRAF databases is facilitated by the unique indexing system that HRAF has developed and refined since the 1950s. The HRAF indexing system (the Outline of Cultural Materials or OCM) helps the user find information on more than 750 subjects, covering just about anything that might be described in an ethnographic or archaeological document. Searching by one of the categories goes right to the relevant passages, rather than to a pointer to the passages. The efficiency of the indexing derives from the fact that a particular subject may not be identified unambiguously in a source, and even if there is an index (if the source is a book) there may not be an entry for the kind of information sought. And, of course, articles and dissertations do not usually have indexes at all. So a particular kind of information may not be easy to find just by looking through a document, because different documents may refer to what is being looked for in different ways, using different words or under different headings. The HRAF indexing system solves the problem of searching for data in texts that do not employ a standard vocabulary. Using the OCM index categories enables one to find just about any kind of information in seconds, because the categories bring the user to the relevant passages even if the original authors used different words and headings. There is also an index to the OCM categories; this large list of words points to the relevant OCM categories for finding particular kinds of information.

For example, if you were interested in assessing the degree to which various cultures depend on stored foods, you might search the index to the OCM, where you would find ‘Preservation, of food, 251.’ This directs you to OCM category 251, ‘Preservation and Storage of Food.’ Searching the database by OCM 251 will bring you to all of the passages that describe dried, smoked, pickled, refrigerated, frozen, canned, and irradiated foods, and whatever other ways are used to store or preserve food. The analysts at HRAF, who have read through and indexed every page of every text that goes into the HRAF files, have made it possible to find the relevant information, even when you cannot anticipate the particular words (including untranslated native words) the original authors may have used.

The HRAF collections are indexed by subject, but not coded by variable. This is deliberate. There are many kinds of potential comparisons and therefore no one system of coding would suit every user. The OCM indexing system gets the user directly to the text passages that are relevant, whatever the comparative purpose and whatever the information of interest. The eHRAF collections can also be searched by the words that actually appear in the texts. In eHRAF, word searches (including proximity and Boolean searches) are possible with or without specifying OCM subject categories. In any search, the context of the hits is never lost. The sections before and after the retrieved paragraphs can be read easily.

The OCM indexing system is a flexible search tool. The user is always in control, because it is the user who decides exactly how to search and exactly how to deal with the information retrieved. Although it is frequently believed that the qualitative nature of most ethnographic description precludes coding and quantitative measurement, this is not true. It is not difficult, after a little practice, to develop at least ordinal scales that can allow the user to code the words of ethnography into measures, and once done it is easy to use available software to test hypotheses, and com-pare, combine, and model rival hypotheses. Needless to say, the indexed texts in HRAF are also amenable to qualitative cross-cultural comparisons.

Installment 48 of the HRAF Collection of Ethnography (available in 2000) is the sixth annual installment of eHRAF: Ethnography. It completes the updating and electronic conversion of the 60-culture Probability Sample Files ( PSF) that were built in the 1970s and 1980s. Raoul Naroll planned how the PSF was built. Quality control considerations were used to select a large list of eligible preindustrial and peasant cultures, and then one culture from each of 60 culture areas was randomly selected. Thus, researchers can use the PSF to test hypotheses on a fairly large and unbiased sample of the world’s preindustrial and peasant cultures. Correlations and other statistical results are likely to be trustworthy and functional, not due to duplications in the sample because of random diffusion or common ancestry. From 2000 on, eHRAF: Ethnography will grow by adding other randomly selected cases (as well as additional con-temporary ethnic groups in the USA and elsewhere). The cases will be drawn by simple random sampling from a new sampling frame for the world that has been constructed by HRAF with the aid of external experts on the various regions. The new sampling frame will eventually include national cultures as well as immigrant cultures all over the world.

The results of comparative studies using ethnography may suggest causality, but social scientists would like to go beyond cross-sectional validation. In particular, they would like to see if temporal sequences validate causal theories. A cross-cultural study using ethnography may be able to discover predictors, but it cannot by itself confirm the temporal sequence assumed in a causal theory. Of course, one could try to measure the ethnographic cases for two different points in time. Measuring each case twice may be increasingly possible as the ethnographic record grows and includes more and more cultures that have been re-studied. But for many questions about cultural evolution, the ethnographic (or ethnohistorical) record is unlikely to provide the required time-series data. This dilemma particularly applies to the classical questions about human cultural evolution. How could we test answers to questions about the emergence of agriculture, the rise of social inequality and the first cities, and the origins of the state?

Now there is a research tool that allows us to escape this dilemma. Investigators of cultural evolution can use eHRAF: Archaeology to study and model causal sequences. Cross-cultural (comparative ethnographic) studies can provide archaeological indicators of cultural and other (e.g., physical and social environ-mental) features. Using those indicators, researchers can test many causal ideas about the major events in cultural evolution on the time-series data in the archaeological record. Thus, the data in eHRAF: Archaeology can allow researchers to determine whether evolutionary patterns in one region are repeated in others, and to determine whether the presumed causal factors in one region are also important, and antecedent, in other world regions. Comparative ethnography can tell us about cultural statics, what is associated with what in the ethnographic record. Comparative archaeology can tell us about cultural dynamics, what precedes what in the archaeological record.

Compared with conventional library research, retrieval of information from the electronic (eHRAF) collections is like a day compared to a month. When comparing or measuring things about a number of cultures, it is not necessary to devote weeks to building bibliographies for each of your sample cases; to search for the relevant books and articles and dissertations, which might have to be obtained by interlibrary loan; to read every page of a source that does not have an index to find all of the passages with relevant in-formation; or to bring any sources back to the library. If the researcher’s institution belongs to the HRAF consortium, and the culture(s) of interest are included in eHRAF, it is easy to access the text passages that contain the particular kinds of information sought.

2. HRAF And Useful Social Science

The usefulness of social science, the possibility of social engineering or applying knowledge to solve social problems, depends on the validity of social science findings and theories. If a finding or theory is not true under some circumstances, it would be foolhardy to think of applying it to real world situations and problems. Hence there seems to be a growing awareness that social science risks being useless, even parochial, if it does not apply to many times and places. The more a finding or theory fits the world generally, the better for understanding and application. Culture really does make a difference. People in different times and places behave differently at least partly because they have different repertoires of customary ideas and practices. So cultural variation cannot be ignored by social scientists if they want to discover principles and relationships that are universal, which is probably why introductory textbooks in psychology, sociology, political science, and international relations, as well as in anthropology, are now paying more attention to the results of cross-cultural research.

In addition, more and more types of investigators— not just anthropologists—are now doing worldwide cross-cultural research. Psychologists, political scientists, sociologists, and evolutionary biologists now use HRAF to test universal hypotheses. And researchers are now using HRAF to test hypotheses about many new kinds of human variable—not just the standard cultural observables such as where couples typically live after they marry or how people get their food. The new kinds of variable include: psychological variables that are measured from projective materials such as folklore; population (biological) variables that are measured using information found in ethnographic and archaeological reports; and comparative archaeological variables that are measured on the basis of ethnographic data that have archaeological correlations. There are many kinds of information in the ethnographic and archaeological records that can be used to construct and test theories.

Of all the research strategies for testing theory, worldwide cross-cultural research is the strategy that most ensures the generalizability of results. To be sure, not all theories can be tested on data from the ethnographic and archaeological records. If something is not widely described, we cannot measure it. Hence social scientists also have to do comparative field studies, cross-historical and cross-national studies, and simulations to test theory. But if we want to maximize the effectiveness of applied social science, we need to try to see if our theories survive worldwide cross-cultural tests. Why should theories that have been cross-culturally tested and supported be more likely to suggest effective solutions to social problems? The logic of statistical inference tells us. A theory that seems to fit a particular region or a sample of nations may not be true for human societies generally. There is no way to tell without cross-cultural testing. A study in one or a few countries may suggest an explanation or application. So may computer simulations and modeling. But ultimately we have to see if the data in the real world are consistent with what we hypothesize. This is why HRAF was invented in the first place, to enable scientists to test their ideas about cultural variation and evolution on a worldwide basis. How often could an explanation that is not universally valid suggest an application or solution that is generally effective?

Bibliography:

1. Ember C R, Ember M 1998 Cross-cultural research. In: Bernard H R (ed.) Handbook of Methods in Cultural Anthropology. AltaMira, Walnut Creek, CA, pp. 647–87
2. Ember C R, Ember M 2001 Cross-cultural Research Methods. Alta Mira, Walnut Creek, CA
3. Ember M 1997 Evolution of the Human Relations Area Files. Cross-Cultural Research 31: 3–15
4. Ember M 2000 HRAF at the Millennium: Blueprint for the Future. Human Relations Area Files, Inc., New Haven, CT
5. Ford C S 1970 Human Relations Area Files: 1949–1969, a Twenty-Year Report. Human Relations Area Files, Inc., New Haven, CT
6. Murdock G P, Ford C S, Hudson A E, Kennedy R, Simmons L W, Whiting J W M 1987 Outline of Cultural Materials, 5th rev. edn. Human Relations Area Files, Inc., New Haven, CT
7. Naroll R 1967 The proposed HRAF probability sample. Behavior Science Notes 2: 70–80