Reflections on the value of ethnographic re-studies: learning from the past

Abstract

This article investigates the value of ethnographic re-studies, and some of the theoretical and methodological issues that carrying out a re-study can generate. It is argued that the disputes that arose around some classic examples in anthropology and community studies are particularly illuminating. For example, Mead’s work on Samoa and Redfield’s study of Tepoztlán, and the re-studies of these, generated considerable discussion, and this still offers lessons for today. Three key functions of re-studies are discussed: replication, mapping change, and producing a fuller portrait of a community or institution. Also explored is the role of the personal characteristics and the theoretical orientations of researchers. It is concluded that re-studies reveal threats to validity that are present in all social research. At the same time, there is a danger that the discrepancies in findings between study and re-study will be exaggerated, indeed that the methodological problems involved will be mistakenly treated as representing a logical impasse.

Keywords:

• re-study
• replication
• objectivity
• mapping change
• anthropology
• community study

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. For an early but still illuminating discussion of re-studies in anthropology, see Lewis (1970, p. 30). On the field of community studies more generally, see Arensberg and Kimball (1964), Bell and Newby (1971), Crow and Allan (1994), Crow (2002, 2012). It should be noted that the value of this kind of work has by no means gone uncontested. For early discussions, see Vidich, Bensman, and Stein (1964), Glass (1966), and Stacey (1969). Interest in the community study tradition of re-study has recently revived, see Burawoy (2003), International Journal of Social Research Methodology, 11, 2, 2008, and Sociological Review, 60, 3, 2012.

2. See Collins (1985), but also Franklin (1994) and Collins (1994). For a recent discussion of the need for, but also problems involved in, replication in biology, see http://andrewgelman.com/2013/12/17/replication-backlash/.

3. There are a few examples of attempted replications taking place in natural settings, see for instance Hardyck and Braden (1962).

4. On the distinctions between descriptive, explanatory, and theoretical claims I am using here, see Hammersley (1998, Chapter 2, 2014, Chapter 2). While there are important differences between the concepts of replication and re-study, these are sometimes exaggerated and distorted, for an example see Burawoy (2003, pp. 646–647).

5. It is worth noting that Redfield prefaced his study with the comment that the materials he collected were ‘no more than a small sample of materials which might be obtained. They are far from constituting an ethnographic monograph on the people of Tepoztlán. […] Eight months [the time he was able to spend there] is not even time enough to correct all one’s initial blunders. That this report still contains some positive mistakes the writer knows is likely. The materials are now published only because it seems improbable that he will soon be able to return to Tepoztlán to continue the study here only introduced’ (Redfield, 1930, p. v). Redfield (1950) himself later carried out a re-study of Chan Kom, another Mexican village that he had studied. Bock (1980) carried out a further re-study of Tepoztlán in the 1970s.

6. Indeed, he has been accused of strategically orchestrating the wide publicity it received; though, of course, this is precisely what social researchers are now frequently encouraged to do in order to maximize the ‘impact’ of their work.

7. Freeman had communicated his initial findings to Mead in 1964 and in subsequent private correspondence with her, indicating the disagreement. Mead had responded to Holmes’ earlier re-study, leading to a subsequent relationship that Holmes describes, ironically, as ‘one of storm and stress’ (1987, p. x). However, he was one of the strongest critics of Freeman’s work. Mead had carried out a re-study of one of the places included in her earlier book Growing Up in New Guinea, but not of Samoa (Mead, 1930, 1956).

8. Though Lewis (1970, p. 24) insisted that ‘except for a few glaring exceptions, most of Redfield’s descriptive data were confirmed by my findings’, the divergence lay in a few ‘matters of small factual detail’ and ‘broad interpretations and total impressions of village life’.

9. Interestingly, Redfield was one of the first to reject this orientation, insisting that social change (conceptualized in terms of ‘the folk-urban continuum’) should be the focus for anthropological inquiry (Wilcox, 2004). For a more recent discussion of the value of re-studies for investigating social change, see McLeod and Thomson (2009).

10. This is a problem that also arises with Geertz’s advocacy of ‘thick description’, see Hammersley (2008, Chapter 3).

11. For an indication of Lewis’s methodological orientation, see Lewis (1953b/1970), and also Rigdon (1988).

12. It is an issue that was addressed a long time ago by Radin (1933), and provided the basis for his early critique of the work of Margaret Mead (see pp. 178–179). In the debate over Freeman’s (1983/1984, 1998) re-study of Samoa, one finds an interesting range of explanations offered by anthropologists for the discrepancies in findings, most of these preserving the validity of Mead’s account (see Bryman, 1994). Redfield (1960, p. 136) offers an interpretation of the problem that emphasizes the role of the research questions addressed. Lewis (1953b/1970, p. 467) specifically proposed a comparative analysis of re-studies as a means of developing theory about the threats to objectivity, and Holmes (1957, Chapter 1) attempted this by comparing multiple studies of the same peoples. Such work could usefully be located within a broader discussion of the use of multiple research strategies: see Burgess (1982, Chapter 22).

13. Muncie is the most re-studied community on the planet: the Lynds carried out their own re-study, and there have been further ones by other researchers since then: see Caccamo (2002) and Lassiter, Goodall, Campbell, and Johnson (2004).

14. This relates to a recent emphasis in the methodological literature on the importance of ‘case-focused’ analysis, rather than treating cases as collections of features whose variation across cases can be used to detect causal patterns: see Byrne and Ragin (2009).

15. Another function that re-studies can serve is that a researcher returning to a setting gains insight into the reception of the initial study by participants in that setting, and perhaps also opportunities to remedy any misconceptions or misuse of the study (see Crow, 2013). There are important ethical issues associated with case studies, as with other kinds of research, and a key one is of course whether the research does harm. But equally important for some commentators is the notion of reciprocity. Thus, Crow argues that re-studies provide us with an opportunity to get a clearer sense of what participants hope and want to get from research. Burawoy (2003, p. 673) similarly argues for the value of what he calls ‘valedictory revisits’.