Russell Bernard

Introduction

Harvey Russell Bernard (born 12 June 1940) is an American cultural anthropologist and social scientist, known for three largely separate bodies of work: the development of the network scale-up method with the oceanographer Peter Killworth, a sustained effort to make anthropological research methodologically rigorous and teachable, and the use of computers to help native speakers document their own languages. He is professor emeritus of anthropology at the University of Florida and, since 2015, director of the Institute for Social Science Research at Arizona State University.

He is included in a psychology collection because his work on social networks and on the reliability of informant recall underpins methods now standard in social and personality psychology. He is not a psychologist, and his training and intellectual commitments are those of an anthropologist.

Early Life and Education

Bernard was born in New York City and majored in anthropology and sociology at Queens College, New York, taking his BA in 1961. He then went to the University of Illinois, where he took an MA in anthropological linguistics in 1963, work he later said shaped much of the rest of his career. He stayed at Illinois for the doctorate, doing graduate work in quantitative data analysis under Edward Bruner, and received his PhD in cultural anthropology in 1968 with a dissertation on economic and cultural change on the Greek sponge-fishing island of Kalymnos.

Academic and Professional Career

Bernard taught at Washington State University from 1966 to 1972 and at West Virginia University from 1972 to 1979, with periods as a Fulbright lecturer in Athens and as director of the Aegean School of Cultural Anthropology on Naxos. The University of Florida recruited him as professor and department chair in 1979; he chaired the department until 1990, was professor of anthropology until 2007, and has been professor emeritus since.

A single nine-month appointment redirected his research. In 1972 he was a research associate at the Center for Marine Affairs at the Scripps Institution of Oceanography, where he met Killworth. Bernard was studying how the people aboard an oceanographic vessel knew one another; Killworth proposed an algorithmic approach to the problem, and the collaboration that followed lasted until Killworth's death in 2008.

Bernard's editorial and institutional work shaped the discipline as much as his research did. He edited Human Organization from 1976 to 1981, was editor-in-chief of the American Anthropologist from 1981 to 1989, and founded Cultural Anthropology Methods Journal in 1989, which became Field Methods, and which he still edits. From 1987 to 2016 he ran the annual National Science Foundation summer institute on research methods, through which a large share of the field's methodologists passed.

He received the Franz Boas Award for Exemplary Service to Anthropology from the American Anthropological Association in 2003 and was elected to the National Academy of Sciences in 2010 (NAS member record), a rare distinction for a cultural anthropologist.

Major Contributions to Psychology

The network scale-up method

With Killworth and colleagues, Bernard developed the network scale-up method, an inexpensive way to estimate the size of populations that cannot be counted directly. Respondents are asked how many people they know in each of a list of subpopulations, most of known size. Because the known sizes calibrate the responses, the answers yield an estimate of each respondent's total network size, which can then be used in reverse to size the unknown groups.

The method was first used to estimate deaths from a recent earthquake, and was later applied to HIV prevalence, homelessness, and rape in the United States, and by the World Health Organization and other bodies to humanitarian problems including stigmatised disease and conflict casualties.

The reverse small-world experiment and network size

Bernard and Killworth's 1978 "reverse small world" experiment inverted Stanley Milgram's design. Where Milgram had asked people to pass a message toward an unknown target, Bernard and Killworth asked people to name the person they would first approach if they were setting out to reach a stranger, on the theory that stated first choices would be more accurately reported than the chains Milgram relied on.

Their work also produced an estimate of average personal network size of about 290, against the roughly 150 associated with Dunbar's number. The discrepancy between the two figures has never been resolved and remains a live methodological disagreement.

Informant accuracy

In a series of studies beginning in the mid-1970s, Bernard, Killworth, and Peter Sailer established that informants' reports of who interacts with whom correspond poorly to observed behaviour. The finding is one of the most consequential in the methodology of social networks, because it calls into question any study that relies on people's reports of their own ties.

Methods training and language preservation

Bernard's textbooks, particularly Research Methods in Anthropology, treat qualitative and quantitative approaches as a single toolkit rather than as opposed camps, and insist that method is a craft learned by doing. Separately, from 1962 onward he worked with the Otomi speaker Jesús Salinas Pedraza on an ethnography written in Otomi rather than about it, producing four volumes in the language, and he has spent decades helping native speakers use computers to document endangered languages.

Major Publications

Research Methods in Anthropology: Qualitative and Quantitative Approaches (first edition 1988; sixth edition, Rowman & Littlefield, 2018) — the standard graduate methods text in the discipline.
Social Research Methods: Qualitative and Quantitative Approaches (2nd edition, Sage, 2012) — the same argument extended beyond anthropology.
Analyzing Qualitative Data: Systematic Approaches (with Gery Ryan and Amber Wutich, 2nd edition, Sage, 2016) — systematic approaches to coding and analysing textual data.
Native Ethnography: A Mexican Indian Describes His Culture (with Jesús Salinas Pedraza, Sage, 1989) — ethnography written in Ñähñu by a native speaker, and an argument about who is entitled to describe a culture.
Handbook of Methods in Cultural Anthropology (1998), as editor — a comprehensive reference on ethnographic, observational, and survey methods.
Killworth, McCarty, Bernard, Shelley, and Johnsen (1998), "Estimation of seroprevalence, rape, and homelessness in the United States using a social network approach," Evaluation Review 22(2): 289–308 — the canonical statement of the scale-up method.
Bernard, Johnsen, Killworth, and Robinson (1991), "Estimating the size of an average personal network and of an event subpopulation," Social Science Research 20: 109–121.

Impact and Legacy

Bernard trained a large share of the anthropologists and methodologists now working, both through the NSF summer institute and through textbooks used by tens of thousands of students. The network scale-up method has become a standard instrument in demography, epidemiology, and official statistics, and is used by the World Health Organization and national statistical offices to size populations that no register captures. His election to the National Academy of Sciences reflects influence extending well beyond anthropology into sociology, political science, public health, and epidemiology.

The criticisms cluster around three points. The first is methodological. A 2021 review in the Journal of the American Statistical Association, surveying thirty years of the network scale-up method, concluded that the approach has produced promising results across many real-world studies but still requires substantial improvement. The source of the problem is structural: the method assumes respondents know the relevant facts about their contacts, that members of hidden populations are as visible to those who know them as anyone else, and that people can count large categories of acquaintance accurately. Transmission error, barrier effects, and recall error each violate those assumptions, and the resulting bias is not captured by the method's confidence intervals. Later estimators, including the network reporting method and the generalised scale-up estimator, were designed specifically to relax the conditions the original model required.

The second is that his most famous negative finding cuts against his own instrument. The informant accuracy work demonstrated that self-reported social ties are unreliable, and the scale-up method depends on self-reported ties. Bernard would treat this as a reason to model the error rather than to despair of the data, but critics note that the two halves of his research programme sit in tension.

The third is disciplinary and more pointed. Bernard has been a prominent and unapologetic advocate for "the science in social science," arguing that anthropology needs systematic sampling, explicit coding, and quantitative discipline. Interpretive and postmodern anthropologists have criticised this stance as imposing a positivist template on a discipline whose object, they argue, does not permit it, and his textbooks' presentation of qualitative and quantitative work as one continuous toolkit is read by some as flattening genuine epistemological differences rather than reconciling them. Bernard's reply, in substance, has been that exposing false ideologies such as racism requires science, while deciding what matters in human life requires humanism, and that the discipline needs a great deal of both.

Learn More

Peter Killworth — his collaborator of more than thirty years, and co-developer of the network scale-up method.
Stanley Milgram — whose small-world experiments set the problem Bernard and Killworth reformulated.

References

Bernard, H. R. (1988). Research Methods in Cultural Anthropology. Sage.