Etic Approach

Definition

An etic approach describes a society using categories the researcher brings from outside, chosen so that one case can be compared against another. The account it produces may be one that the people described would not recognize, and it is not required to be.

Its counterpart is the emic approach, which works from the categories participants themselves use.

Origin

Kenneth Pike coined both terms in 1954, from the endings of phonetic and phonemic. Phonetics is the etic case in its original form. The International Phonetic Alphabet can record any sound a human vocal tract produces, in any language, whether or not speakers of that language treat the sound as meaningful. The notation is fixed in advance and applied uniformly, which is what makes it possible to compare two languages that share no sounds at all.

Pike's proposal was that behavior admits the same treatment. A researcher can build a set of categories for kinship, subsistence, or ritual that is defined before fieldwork begins and applied identically everywhere, and thereby generate data that can be lined up across hundreds of societies.

Marvin Harris took the terms into anthropology and changed what they were asking. Pike's question was whose categories the description uses. Harris changed the question to who decides whether the description is correct. An etic account is correct if it holds up to other researchers, whether or not the people described would accept it.

Method

Etic work depends on the categories being fixed before observation, applied consistently, and defined precisely enough that a second researcher applying them to the same case reaches the same result. Counts of calories, hours of labor, household size, rainfall, and protein intake are the classic materials, because they can be recorded the same way in Bolivia and in Nepal.

[The Human Relations Area Files https://hraf.yale.edu/], begun at Yale in 1949, is the largest project built on this principle: ethnographic material from hundreds of societies indexed under a single coding scheme so that hypotheses can be tested across the sample. Cross-cultural statistical work of this kind is only possible on etic data.

Harris used the Hindu prohibition on cattle slaughter to show what an etic account looks like next to an emic one. The emic account is the religious one that Hindu farmers give. The etic account Harris offered is an ecological argument about the value of live cattle for traction, dung, and milk in a smallholding economy. He held that both accounts can be correct at once, and that neither substitutes for the other.

Criticism

One objection to etic categories is that they not neutral. They come from somewhere, usually from the researcher's own society, and applying them elsewhere can impose a framework that distorts what it measures. Counting households requires a definition of household. Counting religious practice requires a definition of religion. Both definitions carry assumptions that may not hold in the case being studied.

Clifford Geertz argued for interpretation over comparison on these grounds, holding that the density of local meaning is the thing worth having and that it does not survive coding. The response from Harris and others is that any description uses some framework, that leaving the framework implicit does not remove it, and that abandoning comparison surrenders the ability to explain why societies differ.

Use in technology research

Product analytics is etic almost by definition. Event logs, retention curves, session length, and conversion funnels are recorded in categories an analyst specified in advance, applied identically to every user, and comparable across releases, cohorts, and countries. Standardized instruments such as the System Usability Scale work the same way, which is what allows a score of 72 on one product to mean something next to a score of 68 on another.

The strength is scale and comparability. A million sessions can be counted; a million interviews cannot.

One weakness is that the categories were chosen before anyone knew whether they mattered. For instance, a conversion funnel measures progress toward a step the designer defined as progress. Time on page cannot distinguish absorption from confusion. Optimizing an etic measure moves the measure, which is only useful if the measure was pointed at something realistic.

Thus, one working arrangement is to gather emic material first, identify the distinctions people actually make, build etic measures around those distinctions, and then track them at scale.

Related Reading

• Emic Approach
• Research Methods in Anthropology: Qualitative and Quantitative
• Ethnomethodology
• Material Culture and Technology in Everyday Life: Ethnographic Approaches
• Suicide, a Study in Sociology
• Beyond Culture
• The Silent Language
• Careers in Digital Anthropology

Key texts

• Pike, Kenneth L. Language in Relation to a Unified Theory of the Structure of Human Behavior. 2nd ed. The Hague: Mouton, 1967.
• Harris, Marvin. Cultural Materialism: The Struggle for a Science of Culture. New York: Random House, 1979.
• Headland, Thomas N., Kenneth L. Pike, and Marvin Harris, eds. Emics and Etics: The Insider/Outsider Debate. Newbury Park: Sage, 1990.
• Geertz, Clifford. The Interpretation of Cultures. New York: Basic Books, 1973.