“Geographical Distribution Patterns of Various Genes”: Genetic studies of human variation after 1945

Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

Contents lists available at ScienceDirect

Studies in History and Philosophy of Biological and Biomedical Sciences journal homepage: www.elsevier.com/locate/shpsc

“Geographical Distribution Patterns of Various Genes”: Genetic studies of human variation after 1945q Veronika Lipphardt a, b a b

Max Planck Institute for the History of Science, Boltzmannstr. 22, D-14195 Berlin, Germany Free University, Berlin, Germany

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 19 July 2014

After WWII, physical anthropologists and human geneticists struggled hard to demonstrate distance from ‘racial science’ and ‘eugenics’. This was a crucial factor in the ‘revolution’ of physical anthropology in the 1950s, as contemporary accounts referred to it. My paper examines the apparent turn during this period from anthropometric measurements to blood-group analysis, and from ‘races’ to ‘small endogamous populations’, or ‘isolates’, as the unit of study. I demonstrate that anthropometry and blood-group analysis were used simultaneously and in the same research projects until the 1960s. Isolated populations were the new target groups of human population geneticists, from large continental groups to small village populations. Colonial infrastructures provided suitable conditions for these kinds of transnational research projects. I argue that this new framework helped to translate much of the content of earlier racial studies into a less attackable approach to human variation. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: Population genetics Physical anthropology Anthropometry Isolation Human variation Evolution

When citing this paper, please use the full journal title Studies in History and Philosophy of Biological and Biomedical Sciences.

1. Introduction “Since World War II, and especially in the past dozen years”, anthropologist Stanley M. Garn commented on the status of physical anthropology in 1962, “[.] anthropometry is virtually gone. Typology is gone. Craniology, with its indexes and skull types, is gone too. And gone for good is old-fashioned anomaly-anatomy [.]”.1 What a story of success: a dozen years before, in 1950, anthropologist Sherwood Washburn and geneticist Theodosius Dobzhansky had gathered experts of both disciplines at a Cold Spring Harbor Symposium, titled “Origins and Evolution of Man”, in order to reform physical anthropologists’ understanding of human variation. Simultanously, Dobzhansky and other geneticists had launched a number of anti-racist campaigns and shaped the outcome of the UNESCO initiative that would result in two “Statements on Race” (1950 and 1951) (Barkan, 1996; Brattain, 2007; Gannett, 2001; Gormley, 2009; Reardon, 2005). The aim of their

activities had been explicitly to introduce population genetics to the study of human variation.2 In accordance with scientists like Ernst Mayr, numerous historical accounts have claimed that race science and anthropometry, together with typology, were abandoned soon after WWII.3 As Jenny Reardon (2004, 2005), Lisa Gannett (2001) and Nancy Stepan (2003) have argued, the shift from race science to population genetic studies of human variation was neither complete nor as fundamental as others have suggested. These scholars base their analysis on political statements and activities of scientists, most notably in the context of the UNESCO statements, and on single scientists’ theoretical work, along with their popularizing books and pamphlets.4 The newly adopted term of ‘population’, they argue, was a reformulation of race concepts and not a break away from ‘race’. My paper adds to this literature with an analysis that focuses on research programs, research designs, conceptual tools, empirical approaches

2

Reardon (2005), pp. 17e44. For example, see Stepan (1984); Barkan (1992). Stepan, though, later endorsed a more nuanced view: Stepan (2003). 4 See Gannett (2001); Reardon (2004, 2005); Brattain (2007). See also: Gayon (2003). As an exception, see Sommer (2008a). 3

q The title cites physical anthropologist Nigel Barnicot (1965), p. 86. E-mail address: [email protected]. 1 Garn (1962), p. 917. http://dx.doi.org/10.1016/j.shpsc.2014.05.006 1369-8486/Ó 2014 Elsevier Ltd. All rights reserved.

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

and pragmatic decisions. I argue that a closer look into research practices reveals even more continuities and inconsistencies. With regard to sampling practices, group labels, narratives and the concept of the isolate, scientists held on to many of the crucial components of race concepts in their empirical work. Therefore, I have examined mainly publications of empirical results from transnational research endeavors based on population genetics. This paper aims to contribute to the history of race concepts in the second half of the 20th century. Against recent diagnoses of a “reemergence of race” in the life sciences, I argue that ‘race’ did not reemerge, because it never disappeared.5 I attempt to frame this process more broadly, as a longue-durée history of attempts to understand human variation. Concerns about ‘human diversity’ and ‘human variation’, with explicit usage of these very terms, date back to the 18th century. Human variation stood for that particular kind of difference between human groups that was (and is) perceived to be geographically patterned, transmitted from generation to generation and not easily changed in an individual’s life time (Müller-Wille & Rheinberger, 2012). ‘Race’ was but one of several ways to speak about human variation, and yet the most dominant one in this history. The term ‘race’ already implies a certain pattern of human variation, as the much less prominent terms ‘gradient’ and ‘isolate’ do as well. All of these terms superimpose specific patterns of diversity, and each has its own history; in some cases they have been used as opposites of one another, and in others as complementing each other. One could point to endless complexities of human variation, and further that it is a moving target. But most scientists concerned with this issue believed they were capable of determining the patterns of human variation with the help of those terms. Seen from this perspective, the study of human variation entailed broader moral and epistemic concerns, stakes and curiosities than simply ‘racism’, or the classification and discrimination of certain human groups called ‘races’. The study of human variation could serve many purposes. This was particularly the case if researchers focused not on the entirety of humankind, but on the diversity of only one region or on the differences between a small number of (locally sampled) groups. Interest in human variation sometimes arose as a side-product of a medical study on certain diseases or demographies. For some researchers, studying human variation could be a useful preparation for genetic studies, for the manifold epistemic benefits of working with supposedly ‘stable’ differences. In other instances, the study of human variation went hand-in-hand with an interest in origin, history and evolution. For researchers in this field, even the smallest genetic difference between human groups might potentially hint at a major insight into human evolution. Digging for difference, and stabilizing it for further studies, was one of the main preoccupations of that field. Hence, what looked like a medical genetic study of some striking local pathological differences could have profound implications for evolutionary biology; conversely, mere narratives of longue durée group history could have profound consequences for the medical treatment of contemporary populations.6 This is not to say that this broader field was innocent, harmless or free of racism, biological determinism, political influences and eugenic beliefs7dquite the contrary. The whole field has always

5 For “re-emergence of race” see Rose (2007). For an extended discussion, see Reardon (2005); Müller-Wille & Rheinberger (2008). 6 As in the case of Sickle Cell Anemia: Tapper (1999); Wailoo & Pemberton (2006). 7 There is a great deal of literature on the history of human genetics in connection with the legacies of eugenics, which addresses the implications of ‘race’ for human genetics, mostly with respect to its entanglements with ‘racial hygiene’. Not so many studies touch upon the question of how human genetics was involved in studies of human variation before and after WWII. See Marks (1996); Mazumdar (1996); Mendelsohn (2001); Gannett & Griesemer (2004); Pogliano (2005); Spörri (2012).

51

been soaked with judgmental power, independent of whether the term ‘race’ was praised or condemned. To better understand how such judgments played a role in the empirical work of scientists studying human variationdin a supposedly innocent, harmless and objective waydmy paper focuses on the continuities of conceptual tools, methods, biohistorical narratives and intellectual curiosities linked with human variation. The 1950s are a particularly important period in this regard, but I will also briefly allude to a transformation of that research field occurring in the 1960s. 2. Global institutions and networks The early 1950s saw the emergence of new research designs and programs, as well as the foundation of new institutes and departments for the ‘study of human variation’. Pioneering empirical case studies were carried out, which helped to formulate and refine those new programs, approaches and institutionalizations. This section briefly outlines some of the relevant networks and institutions in this field. After WWII, studies of human variation were more often than not transnational collaborative endeavors. Several intertwined research objectivesdmedical, evolutionary, biochemical, genetic and physiologicaldcould all be pursued at once, in one and the same research expedition, with a multidisciplinary set of methods, and play out in many publications aiming at different audiences. The teams did not follow one main research question, but many diverse questions at the same time. The ambitions of collaborative research teams were global in their dimension; the rationale they followed was to visit as many populations as possible around the globe, and they were all-encompassing in their scope, as they sought to collect as much data and as many samples as possible. There were some preferred locations, often in contexts with a colonial past or present, that seemed to offer particularly interesting patterns of diversity: for example, the Pacific islands as a pattern of many small isolates; Africa; Latin America; or India with its caste system. Most of the groups that scientists chose to study constituted social minorities in politically tense situations, and researchers gratefully drew on the infrastructure provided by administrations and health services. The case study by Edna SuárezDíaz in this issue offers an example: Mexican anthropologist Ruben Lisker approached indigenous people through institutions, networks and political programs which sought to modernize the nation state.8 The Rockefeller Foundation, as well as other funding bodies, allocated enormous financial sources for genetic research on populations around the globe. New research institutions, such as the “Institute for the Study of Human Variation” at Columbia University, the “Laboratory for Human Genetics” at the Federal University of Paranà in Brazil, and the “Laboratory for the Study of Human Variation” in Bombay, all founded in the early 1950s, helped to foster transnational exchange (Gormley, 2009).9 The latter two institutions were founded by researchers born in Brazil and India, respectively. Prominent geneticists promoted human variation studies in their preferred host countries, including Haldane in India and Dobzhansky in Brazil. Gradually, as Luigi Luca Cavalli-Sforza, who partook in the new paradigm as a young researcher, later put it: “a body of data began to accumulate”, one from which many

8 To name some examples: the Duncker community 1952 (Glass, Sacks, Jahn, & Hess, 1952); some African “tribes” (by Jean Hiernaux, see below); the Western Apaches (Kraus & White, 1956); Australian Aborigines (Phillips, 1928; Simmons, Graydon, & Semple, 1954); the Walser and Romansh in Switzerland (MoorJankowski & Huser, 1956); Brazilian Indian Tribes (Kalmus, 1957); the Amish and the Basques. 9 Gormley (2009).

52

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

correlations could be followed up.10 In the 1960s, the WHO initiated a transnational program in population genetics.11 Various international conferences in the 1950s and early 1960s served as arenas for exchange, debates and struggles around theoretical and methodological concerns.12 In addition, they provided an ideal stage for young researchers to present their results from research expeditions. Far from silencing discussions on ‘race’, these conferences indicated that the interest in ‘racial variation’ was as great as ever.13 How the various disciplines, actors and institutions were linked or opposed to each other remains an understudied aspect of this history and cannot be discussed in length here.14 But from the discussions in journals of physical anthropology and genetics, it becomes clear that tensions and cooperations cut across disciplinary boundaries. 3. Physical anthropology versusdor hand-in-hand withdpopulation genetics? As many analysts have pointed out, claims about what counts as legitimate ‘knowledge’ have played a crucial role in racist and antiracist discourse. In discussions on race and racism that reached beyond unreflected daily actions and experiences, these knowledge claims were articulated and made accessible to many. In these debates, science was attributed a crucial role, for it was continuously evoked as the ultimate referee to authoritatively settle the ‘race question’. The general trust in science to ultimately resolve the issue has seen only minor tremblings since WWII; this trust remained one of the main driving forces for a broad consensus for the support, pursuit and funding of human variation studies. Yet, the trust in certain disciplinary approaches to human variation changed considerably after WWII, albeit not in the way suggested by the actors cited in the beginning of this paper. Anthropometric methods had been criticized since the late 19th century from various angles, but remained a well-respected set of methods until the 1960s. Nevertheless, after WWII, general genetics enjoyed more trust for pointing a way out of the entanglement of human variation research with racism. Accordingly, the post-war era saw a considerable number of prominent scientists from general genetics and animal genetics, such as Theodosius Dobzhansky, John Burdon Sanderson Haldane and Leslie Clarence Dunn, participating in anti-racist activities.15 Geneticists propagated the introduction of population genetics into the study of human variation, and, willingly or not, scientists pursuing empirical studies of human variation integrated population-genetic methods into their toolboxes.16 Although some well-established physical anthropologists in a number of countries refused to accept genetic methods for the study of human variation, younger researchers increasingly endorsed genetics in their fieldwork and theoretical explanations, and rejected older methods.17 This negative attitude towards older

10

Cavalli-Sforza (1994), p. 3. 11 Radin (in this issue). 12 Cold Spring Harbor Symposium on Quantitative Biology 15: Origin and Evolution of Man (1950), “Studies on Isolates” (1956), Proceedings of the Second International Congress of Human Genetics (1961). See also Smocovitis (2012). 13 Reardon (2005), pp. 66e69 has analyzed one of those conferences in depth. 14 Reardon (2005), p. 41f. has given a brief account of some debates. 15 Gannett (2001); Reardon (2004, 2005). Furthermore, see Marks (1996, 2008). 16 See Cold Spring Harbor Symposium on Quantitative Biology 15: Origin and Evolution of Man (1950). 17 Some well-established physical anthropologists also adopted population genetics along with typological notions, for example William Boyd and later Alice Brues. However, as Smocovitis (2012) points out, in the 1950s these scientists were not integrated into the geneticists’ successfull attempts at extending the ‘Evolutionary Synthesis’ onto human evolution.

research practices may have entailed a good deal of distancing for political reasons, and could also be attributed to moral embarrassment over recent political entanglements of the field.18 In any case, in the early 1950s, ‘racial variation’ was still seen as a fascinating research object, but a methodology solely consisting of anthropometry seemed too narrow an approach. On the basis of a handful of genetic traitsdor, as they were now called, ‘markers’, such as blood groups, PTC taste sensitivity and color blindnessdresearchers hoped to study human variation in a more precise way than anthropometric methods would allow for. In contrast to stature, height or pigmentation, these genetic markers remained the same over the life span of an individual, and their mode of inheritance was well established.19 Already by 1955, the Belgian physical anthropologist, Jean Hiernaux, noted that “the great value of gene frequencies in physical anthropology is now widely accepted”.20 This trend was also reflected in the growing number of publications on blood group distributions in journals of physical anthropology. A lively debate accompanied this trend, maintaining a certain tension between physical anthropology and population genetics; many theoretical and conceptual discussions and publications followed in conference proceedings and in nearly every journal dealing with genetics or anthropology.21 And yet, even though genetic markers were increasingly favored over anthropometric measures, as late as 1964 a WHO expert group recommended scientists to take both blood samples and measurements in the course of their studies of “population genetics of primitive groups”.22 But this does not mean that the line between ‘old’ and ‘new’ paradigms simply is to be drawn at a later date. Rather, on two different levels, there were remarkable connections and overlaps between those approaches at least until the mid-1960s. Firstly, there clearly was common ground on the conceptual level. In an editorial from the 1950 issue of the American Journal of Physical Anthropology, geneticist Herman Strandskov, and physical anthropologist Sherwood Washburn, outlined the common approach: Recently there has been a tendency to contrast genetics and physical anthropology as if they were necessarily opposed. Some deny the utility of anthropological methods for racial analysis, while others find the genetic concept of race unrealistic. We think that the basic idea of race is the same no matter whether races are described in terms of anatomical traits or gene frequencies. Races are large groups which differ in heredity. If all mankind mated in random manner, there would be no races. Partial isolation is caused primarily by geographic factors, reinforced by custom. Since there is partial isolation, genetic distinction may develop. These distinctions between the varieties of mankind may be described either in terms of anatomical traits (phenotypes) or gene frequencies. Both may be used together [...]. There is no conflict between genetic and anatomic methods of studying races, provided the descriptions are based on populations and the same concept of evolution is used.23 Accordingly, both approaches could well be employed together in the eyes of some of their proponents. In addition, by drawing on concepts of isolation, populations and evolution, Strandskov and Washburn perfectly framed what was to become the new theoretical basis of research programs of human variation studies, which I will discuss later in more detail.

18 19 20 21 22 23

See Lindee & Santos (2012); Reardon (2005). See Schneider (1995); Marks (1996); Mazumdar (1996). Hiernaux (1955), p. 455. The debate began as early as 1944. See Science (1944). World Health Organization (1964). Strandskov & Washburn (1951).

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

Secondly, on the level of empirical work, both approaches went hand-in-hand. In 1950, several dozen population genetic studies were underway. Analyses of blood groups and other ‘polymorphisms’24 were carried out in connection with anthropometry, particularly by young scientists who were trained in both methodologies. To answer questions of human evolution and diversity, research teams would take all relevant methods to the field: they took measurements and blood samples from ‘unique populations’ all over the world, and some of them still described and classified the groups they studied with typological terms. For example, a Belgian research team, headed by Hiernaux, turned to the Belgian colonial territory of Rwanda, Burundi and Congo in the 1950s to examine various ethnic groups living there. The overall aim was to distinguish between Hutu, Tutsi and Twadwhich the researchers classified typologically as “hamitic”, “ethiopid” and “pygmoid”dby genetic and anthropometric methods, although the publications from these studies later contributed to a much wider range of topics.25 In India, a research team had been making similar research efforts since the late 1940s, focusing on Indian castes. The team leader, L. D. Sanghvi, had completed a Ph.D. with Dobzhansky and established the “Laboratory for Studies of Human Variation” in Bombay. This team also published widely on many topics, including a methodological comparison of the results from Sanghvi’s genetic and anthropometric analysis.26 An American team, headed by anthropologist Joseph Birdsell, studied Australian Aborigines as a racial group and as a continental isolate, as part of Birdsell’s attempt at “revitalizing racial anthropology”.27 In the light of these examples, it seems incontrovertible that the “revolution” or break away from anthropometry to population genetics was anything but complete.28 Both practices were generally acknowledged, regardless of scientists’ engagements in political battles waging around racism. As British serologist Arthur Mourant explained, “I wish to stress that there can be no real incompatibility between the information supplied by blood groups and that yielded by morphological characters. Each kind of evidence must add to the information yielded by the other”.29 Mourant’s statement of 1956 and the WHO expert group’s recommendation (1964) both point to the importance of accumulating data; it was obviously difficult to abandon a well established form of data collection, and it was still believed that the combined accumulation of both could yield more clues in the future. The large data collections based on anthropometric measurements were still retained and consulted for decades after the measurements were made. Notably, anthropometric data was not only used for racial studies, or studies in human genetic variation (geographically patterned), but was also important for nutrition and growth studies, for example. And if data could not be interpreted for the migration history of groups, it could potentially be useful for making claims about issues of child development. And yet, physical anthropology gradually lost its traditional authority on questions of human variation in the 1960s, whereas human population genetics was increasingly trusted to be the most objective way to investigate human diversity and evolution. After

24 Polymorphism means that a certain gene is found in several versions or alleles. Other genes only occur in a single version (allele). 25 E.g. Hubinot, Hiernaux, Massart-Guiot, & Massart-Guiot (1953); Hiernaux (1954, 1955, 1963, 1965a, 1965b, 1966). 26 Sanghvi (1953). See also Sanghvi & Khanolkar (1950); Sanghvi, Varder, & Master (1956); Sanghvi (1954, 1966). 27 Birdsell (1950), p. 259. 28 Thompson (1967), among others, uses the term ‘revolution’. 29 For a detailed account of Mourant’s own research practices, see Jenny Bangham (in this issue).

53

decades of debating anatomical or morphological traits under the lead of physical anthropologists, geneticists now defined ‘good markers’ or ‘valid markers’ anew. Only a few anatomical characters remained accepted, and only if they fit the criteria of being inherited in a reproducibly stable way. Geneticist Theodosius Dobzhansky summed up all known traits in a hierarchical “list of traits that might be useful in raciological studies”. Only for nine genetic markers did the “mode of inheritance” seem “well established” (such as blood groups), while it was “approximately established” for thirteen anatomical markers (such as ear lobe or eye fold) and “poorly established” for fourteen markers, among them the classical anthropometric ones (such as eye and hair color) (Fig. 1).30 Strikingly, no pathological markers appear on Dobzhansky’s list. Scientists studying geographically patterned and heritable human variation during this time typically searched for normal rather than pathological markers. Genetic diseases based on single mutationsdpathological alleles, so to speakd, so central to the work of medical geneticists, were of little use for Dobzhansky’s purposes, as they were presumably under selection and thus did not fit the category of being non-adaptive. But there was a long-standing history of cooperation between geneticists and medical research. Since blood typing was part of many medical services at the time, blood group research was often carried out in the framework of a medical institution and by physicians.31 Both sides shared some epistemic interests and sometimes also eugenic concerns. One obvious link between these fields was inbreeding, or consanguinity. Medical geneticists assessed ‘genetic loads’ of populations, or accumulations of harmful mutations by inbreeding or by radiation exposure. The interest in inbreeding/consanguinity can be viewed as an interface where medical, genetic and eugenic research interests overlapped, and many geneticists, such as James Neel and Cavalli-Sforza, studied the effects of consanguinity in all its dimensions in the early years of their careers.32 More new methodologies came with technical advancements, and seemed to revolutionize anthropometry in a different way: “Today’s physical anthropology leans heavily upon today’s technology. Measuring is donedbut on radiographs, using automatic print-out, motordriven equipment”, Stanley Garn noted. “Reflection colorimeters and spectrophotometers have replaced colored tiles and dyed-hair samples. With paper chromatography and starch-gel electrophoresis, urinary constituents and serum components are separated into individual amino acids and protein fractions”.33 Nevertheless, in what follows, I will mainly concentrate on genetic methods that were used, regardless of whether they were employed by geneticists or physical anthropologists, to explain how the new approach was conceptualized and implemented in empirical studies of population genetics.

4. Populations, selection, evolution: translating old terms into innovative concepts As Reardon has demonstrated, geneticists claimed that only population genetics could help to overcome the errors associated with the race science of the interwar period.34 They claimed that the ‘natural groupings’ of mankind were not everlasting divisions

30

Dobzhansky (1950a), p. 391. Bangham (2013). See, for example, Cavalli-Sforza (1956); Barrai, Cavalli-Sforza, & Moroni (1962); Schull & Neel (1963). 33 Garn (1962), p. 917. 34 Reardon (2005). See also Schneider (1995); Marks (1996, 2008); Gannett & Griesemer (2004). 31

32

54

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

Fig. 1. In his 1950 paper ‘Human Diversity and Adaptation’, Theodosius Dobzhansky reproduces a list, originally compiled by Boyd and Spuhler, “which may prove useful in raciological studies”. No anthropometric measurement is listed in the A section of characters with the “mode of inheritance well established” (1950a, p. 391).

fixed in stable races, but the ‘product of evolution’ as a dynamic process. Though framed as absolutely new, this understanding of human variation was anything but new. Even German race scientists of the interwar period (today rightfully deemed to have been most racist) had explicitly viewed races as ‘products of evolution’, and the process of their formationdthen called ‘racial history’das dynamic.35 However, evolutionary thinking was now based on a different temporality of evolutionary change; and it was now applied much more systematically, with a coherent framework and language and with some mathematical precision of key terms, such as migration, separation, isolation, mutation, variation, selection and random drift.36 All these terms carried large semantic fields with them, and evoked associations that scientists were eager to control through the greatest possible terminological exactness. In addition, many allegedly new concepts, such as ‘reproductive isolation’, ‘non-adaptive characters’, or ‘gene frequencies in

35 36

See Lipphardt (2012). See Stepan (1984), p. 178; Reardon (2005), p. 40.

populations’ had been at the core of studies of human variation long before.37 In particular, ‘human variation’ had long been imagined as structured in two seemingly contradictory patterns: either in gradients, or in isolates. Whereas some conceptual elements had previously functioned as self-evident, unquestioned assumptions, and thus not in need of thorough theoretical explication or empirical testing, now they had to be proven and accounted for in systematic ways. They were re-framed and emphasized differently, and, for the first time, empirically accounted for.38 This section will revisit some of the curious continuity-by-reformulation cases at stake in the above-mentioned shift. Geneticists were well aware that the technical language of the new synthesis of evolution theory would not go unchallenged, for it inevitably carried values and notions associated with radical political views. Without explicitly discussing them, the memory of WWII atrocities, genocide and euthanasia cast a long shadow on this discourse, and practitioners sought to avoid any explicit reference to the most discredited terms associated with that history. In a rare moment of candor, Birdsell alluded to the historical baggage of the concept of selection and why anthropologists would hesitate to use it: “The pain of the idea is such that it is by no means universally accepted”.39 Perhaps to avoid such painful associations, scientists used the attribute “nonadaptive” for characters that were supposedly not subject to selection, and “adaptive” for characters under selection.40 Accordingly, there was much discussion around the term ‘race’ and whether it should be abandoned, both as a term and as a concept. As Reardon has shown, a crucial step following the new theoretical premises of population genetics was to adopt the term ‘population’ as the principal unit of mankind’s division.41 In many forms and meanings, ‘population’ was the buzz word of the day: in academia, politics and the public. In genetics, humans were no longer divided into races with typical traits, but into populations that differed in gene allele frequencies.42 Yet the substitution of ‘race’ with ‘population’ necessitated many further innovations, as well as recourses to older ‘racial’ knowledge. The philosopher of science Lisa Gannett, in a close reading of Dobzhansky’s texts, has demonstrated that his very influential population concept was still a concept of race, but a populational one.43 Indeed, the new race conceptdpopulations differing in allele frequenciesdallowed for a very flexible use of ‘population’ for groups of all sizes: small ones such as islands or religious communities, but also large populations that mapped closely onto what had been conceptualized as ‘races’ in the pre-war period.44 The new approach of population genetics aimed at quantification, and thus also required mathematical innovations. With new mathematical models, simplifying assumptions, approximations and complex systems of variables, geneticistsdoften supported by mathematiciansdaimed at further empirical control of the population under study. A significant number of such innovations had already been published in the pre-WWII era.45 Ever since, but particularly since 1950, scientists discussed mathematical formulas

37

Lipphardt (2012). This had happened several times before in the history of human variation research, as, for example, with the re-consideration of ‘race mixing’ and ‘pure lines’ of humans after the rediscovery of Mendel’s laws. 39 Quoted from Glass (1953), p. 104. 40 This entails a certain irony: the term “adaptive” had long been associated with Lamarckism, as opposed to Weismann’s theory, which strongly emphasized selection. “Negative selection” was used for characters that were selected against. 41 Reardon (2005), pp. 32e36. 42 The same understanding was expressed by German physical anthropologist and geneticist, Eugen Fischer, already in 1930: Fischer (1930). 43 Gannett (2001). 44 Cf. Reardon (2005), p. 65. 45 Bernstein (1928); Wright (1932); Dahlberg (1943); Fisher (1949). 38

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

for calculating the “effective breeding population size” for assessing the impact of drift, population growth, selection, mutation rate, geographic distances etc. on the so-called genetic structure of populations.46 Implementing this population concept from theoretical evolutionary biology to empirical studies in human population genetics was fruitful and prompted many new research projects, but it also proved quite a challenge. One needed to rule out, account for, or study in a controlled way, mutation, selection, endogamy, drift and migration. Geneticists had to find good genetic markers, and answer the question of why a particular marker would be useful. They needed to decide which populations to study, and which requirements a good study population would have to meet. Moreover, if these theoretical requirements were fulfilled, the question was how to delineate a population empirically? These challenges led to a dynamic interplay of markers, populations, mathematic operations and biohistorical narratives, each being the object of scientific curiosity as well as the tool for studying other objects.

5. Biohistorical narratives Equipped with only a few theoretical and practical tools, biologists believed they were now able to understand what happened to genes in evolving populations.47 This approach helped to align well established classifications of humankind and a mass of empirical data on human variation with abstract evolutionary mechanisms (namely, variation, selection, drift or isolation), and thereby transduced these mechanisms into new research designs, which allowed for the empirical proof of many hitherto implicit assumptions. One can thus hardly overestimate the great impact of population genetics on empirical studies of human variation. Notably, however, human population genetics also aligned two very different traditions. In the case of homo sapiens, the evolutionary mechanisms mentioned above were merged with older narrative and mythical accounts of group history, typically with very concrete and specific details about the geographic and historical cohesion of groups.48 I call these narrations ‘biohistorical narratives’, because they provide specific group remembrances of descent, belonging, displacement, settlement, aggressive and friendly encounters, love and family, and resemblance and relationships, all of which could be interpreted as relevant for the biological study of a group. Biohistorical narrations are embedded in all kinds of narrative accounts of communities, nations or even families; many have been translateddeven if only incompletelydinto the realm of biological theories ever since Darwin’s publications. In this biologized form, single group narrations, such as those of the Basques as a cultural and biological isolate, for example, have been integrated into the ‘grand narration’ of humankind’s history and pre-history: a grand story continuously told by scientists and academics of various disciplines since the 19th century.49 These narratives, as the

46

For example, Fisher (1949); Wright (1932, 1952). For a differentiated account of the so-called Modern Evolutionary Synthesis, see Cain (2009). See also Mitchell & Dietrich (2006); Bowler (2009). On the application of the Evolutionary Synthesis onto human evolution, see Smocovitis (2012). 48 One of the most classical sources of such narrations was the Bible, with its stories of migrations and encounters of peoples. An exemplary analysis of biblical narratives re-interpreted scientifically in archeology is provided by Wiedemann (2012). 49 Lisa Gannett and Jonathan Marks have developed terms such as “biohistory” and “biohistorical narrations”: Marks (1995); Gannett & Griesemer (2004). See also Sommer (2008b). For more about how I understand biohistorical narratives, see Lipphardt & Niewöhner (2007). 47

55

following section shows, continued to inform population genetics, as they had informed race science.

6. Patterns of diversity: isolates As mentioned above, the gradient, or cline, was one of the patterns of diversity that scientists continued to be interested in. In 1962, anthropologist, Frank Livingstone, claimed “There are no races, only clines”, although this view was hardly new.50 But the cline concept did not sit easily with the imperative to investigate populations: clines were difficult to investigate. In contrast, many scientists endorsed the view that isolates, that is, relatively small isolated populations, were the unit to study. The isolated populations of islands, castes or other closed communities were considered highly relevant. Reardon has already pointed out how geneticists made this a strong claim in their theoretical and programmatic publications since the 1980s.51 But even before, and simultaneously, studies of isolates were already underway. Many publications in the 1950s centered or drew on small isolates as the most promising ‘populations’ to study.52 Isolates seemed manageable from the research pragmatics point of view: large populations were difficult to study empirically. Furthermore, the concept of population as the principal unit of human diversity was based on an evolutionary understanding of population, and thus, as a theoretical and empirical prerequisite, rested on the assumption of periods of ‘reproductive isolation’dthat is, time periods during which a human group had been reproductively isolated from other human groups. As the longue durée history of these populations was crucially important for the proof of continued isolation, scientists were in need of well-established sources for the (biological and reproductive) history of a population. Contextualizing the empirical work on evolutionary isolates in its historical background shows that this was hardly a new concept. Notably, the idea that races might be conceived as isolated populations has 18th and 19th century precursors in public debates about certain groups, such as the Basques, the Jews or the Sami. It had found its way into scientific discourse in the late 19th century as one of those ‘self-evident’ assumptions about human variation. Older scientific race concepts had already employed the notion of human races as reproductively isolated groups, within which specific genetic traits were transferred from one generation to the next. Hence, the scientific interest in ‘reproductive isolation’ and in ‘crosses/mixing’ of (supposedly) previously isolated groups dates back to the 19th century.53 With the specific turn to temporary isolation, it had been formulated systematically by Felix Bernstein in 1928: [.] in the past of the human species, at certain time layers, true inbreeding communities existed which were disposed to the same influences. Such inbreeding communities, existing in certain time layers and certain local settings, we can call races, if this would not cause a severe misunderstanding; namely, that these communities were the ultimate unit and don’t need further inquiry into their original composition.54 The crux was that ‘temporarily’ could mean very different things: it all depended on how long isolation had worked on a population;

50 Livingstone (1962). For a discussion of the debate on clines, see Reardon (2005), pp. 32e36. 51 Reardon (2005), pp. 67e69. 52 See, e.g. “Studies on Isolates” (1956). See also Goldschmidt (1963). 53 Lipphardt (2012). 54 Bernstein (1928), p. 438.

56

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

and this entailed quite fundamental assumptions about the history of the group under consideration. Contrary to the later claims of population geneticists, therefore, isolates had already been an empirical preoccupation of those race scientists from whom they were so eager to distance themselves. Apart from scientists’ preoccupation with purity in a eugenic sense, ‘isolated’ races were also important methodological prerequisites for empirical studies of human variation. Whereas the expression ‘pure race’ was used predominantly in the late 19th century, and aimed to describe a stable condition since prehistoric times, the term ‘isolation’, used more often after 1900, evoked notions of processes and temporality, which was more in concordance with Darwinian and Mendelian ideas. The empirical turn to ‘isolates’ (as well as to previously ‘isolated populations’ that had presumably mixed under ‘controllable conditions’) had already begun in the 1920s as an international trend. Scientists turned to supposedly isolated populations in a very Darwinian fashion, namely, to island inhabitants, people separated by geographic barriers such as mountains or rivers, that could be studied abroad, in remote places and in colonial contexts. But they also turned to isolates ‘at home’: minorities and outsider communities seemed to offer themselves as ready examplesdand occasionally, they literally ‘offered themselves’, as for example was the case with some Basque emigrants in London (Bangham, in this issue). Migrant groups were seen as perfect subjects for such studies of isolates: upon their arrival, migrants were understood as remaining isolated from the ‘autochtone’ populations surrounding them, and their difficulties with adaptation seemed to confirm the notion that they were genetically ‘adapted’ (by selection) to their place of origin. Hence, isolates under study were often groups that, historically, were blamed for social tensions within certain national or imperial contexts. Many of the isolates studied before WWII were still considered scientifically and morally appropriate objects of study after the war, particularly the notorious ‘isolates’ from Europe: the Jews, the Roma, the Basques, the Sami and the Fins. Within the new approach, many small ‘races’ of past studies were simply reconceptualized as ‘isolates’. How, then, was ‘isolation’ conceptualized by geneticists in the 1950s? In their eyes, evolution was an experiment, and the test unit submitted to evolutionary forces was “the breeding unit”, “the isolate”, “the Mendelian population” or “a reproductive community of sexual and cross-fertilizing individuals which share in a common gene pool”.55 A species would be a Mendelian population, albeit “differentiated into complexes of subordinate Mendelian populations”; and these, again, were equated with races, subspecies or local populations, separated by more or less effective reproductive barriers.56 Reproductive barriers would, through centuries, lead to genetic distinctness of populations, and this could be measured in allele frequencies of genes that were polymorphic (i.e. had different allelic forms). As Joseph Birdsell put it in 1950, a “genetical concept of race” addressed “a genetically more or less isolated division of mankind possessing a corporate genic content which differs from that of all other similar isolates”. He concluded that “it is an explicit imperative of such a definition that the proper unit of study is a population, or an isolate”.57 The conceptual turn to isolates brought about new programmatic directions and requirements for empirical work. Most

importantly, in the 1950s, scientists were for the first time required to account for the isolation status of groups they wanted to sample. They did so by importing a concept from social science: endogamy (Lipphardt, 2010). This theory of social reproduction referred to communities in which marriages were mainly contracted among group members and not with outsiders (yet some scientists were eager to point out that it was not to be confused with incestuous relationships between close family members). For the geneticist, ‘endogamy’ easily translated into ‘inbreeding’, or into the notion of communities that experienced no reproductive mixing with other groups. So, the conceptual turn to isolates went along with the empirical turn to endogamous groups that could be delineated by social scientists (although it remained a challenge to pin down the nature of particular relationships between group and family members in small isolates). Geneticists now considered very small ‘endogamous group’ to be the unit that best represented a genetic and evolutionary population, and hence provided the soundest empirical basis for population genetic investigations. As the Indian geneticist Sanghvi put it, “The ultimate racial units of importance are the endogamous groups”.58 Endogamous groups offered the “invariable and necessary framework [...] for a study of genetic characters”.59 Sanghvi also agreed with Dobzshansky’s appreciation of India as a great place for genetic studies: “The Indian caste system is the grandest [...] genetic experiment ever performed on human populations”.60 “People of India”, Sanghvi wrote in 1950, “are almost under an experimental environment, broken up into a large number of mutually exclusive groups, whose members are forbidden, by an inexorable social law, to marry outside their own group. [.] The people of India are not just a collection of individuals, but a mass of corporate entities, whose numbers, names, characters and functions are infinitely diverse”.61 Together with V. R. Khanolkar and his students, Sanghvi investigated castesdamong them immigrant groupsdas endogamous groups in India.62 They took blood samples, tested PTC taste sensitivity, color blindness, and took body measurements. Within the six groups studied, Sanghvi and his colleague Khanolkar found significant genetic differences, especially between two particular groups, the Koknasth Brahmans and the Chandrasenyia Kayasth Prabhus. These differences, they said, were as large as differences between ‘Whites’ and ‘Blacks’ established by studies from the US; and yet, these two very different endogamous Indian groups had hitherto always been lumped together into the same racial category. Hence, they explicitly challenged the older racial classifications of the people of India. At the same time, they drew new lines between human groups alongside socialecultural notions that might have seemed more plausible to Indian-born researchers than the racial divisions developed by Western scientists. Thus, in spite of people of India being “infinitely diverse”, Sanghvi believed in their evolutionaryegenetic groupness, and that it could be clustered and classified, even if along very different lines than racial science had previously maintained. On the other side of the population size spectrum, even the ‘autochthonous’ population of a continent could be rendered an isolate. In 1950, Joseph Birdsell published research on Australian populations that was considered path-breaking due to its

58

Sanghvi & Khanolkar (1950), p. 62. Sanghvi & Khanolkar (1950), pp. 52e53. Dobzhansky (1962), p. 234. 61 Sanghvi & Khanolkar (1950), pp. 52e53. 62 Endogamy, in these cases, was not clearly defined, but from the scientists’ account it becomes clear that it meant to denote inbreeding but not incest. At the time, how endogamy could be defined biologically was debated among geneticists, mathematicians and others. 59 60

55

Dobzhansky (1950b), p. 405. Dobzhansky (1950b), p. 405: “By far the most complex system of Mendelian populations exists in the human species”. 57 Birdsell (1950), p. 259. For a genetic definition of race, see Stern (1949), p. 558. Cf. Reardon (2005), pp. 67e69. 56

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

meticulous methodology and innovative visual approaches. He collected data on the genetics of blood groups as well as morphological traits such as stature, teeth and facial height. Birdsell’s work also stands as a rare example for how researchers sought to integrate the two seemingly contradicting patterns of variation in their research designs; even within local isolates, Birdsell maintained, ‘clines’, or gradients of traits, could be observed.63 Vice versa, one could account for large continental gradients by sampling isolates that lived along the gradient and by printing their gene frequenciesdor trait frequenciesdinto isoline maps that would display a cline. And yet, ‘non-autochtonous isolates’dmigrated populationsdwould pose a challenge to this approach. Notably, with regard to geographical isolation, the new conceptual tools had been taken from animal and plant genetics.64 Complaints about the human species as a test object were frequent: “The study of human genetics, as everyone knows, is severely hampered by such inevitable obstacles as the unfeasibility of experimentation, the small number of progeny, and the great length of each generation”.65 The animal and plant geneticists had conceptualized reproductive barriers as, for example, geographical distance, mountains, rivers or predators. But in the case of humans, more specific factors of isolation needed to be considered. As Hermann Strandskov wrote: “Relative to the human species it may be a psychological, social or cultural factor. In the latter event two or more incompletely closed intrabreeding populations may occupy the same spatial range without interbreeding to any appreciable extent”.66 His colleague, geneticist Leslie Clarence Dunn, named “language, custom, religion, economic organization, social class” the isolating factors specific to humans: “In modern societies social isolation tends to replace other isolating factors of the natural environment”.67 7. Identifying isolates by non-genetic methods Geneticists took all of these factors into account, but finding a true isolatedan endogamous groupdproved difficult. They obviously needed indicators leading them to the right kind of population. In several respects, the concept of the ‘isolate’ as an ‘inbreeding unit’ seemed to help with the practical task of sampling. However, demarcating a population for an empirical investigation in population genetics proved tricky and required ever more conceptual innovations. A good population to study would have to fulfill a number of requirements. First of all, the population needed to possess what was conceptualized as ‘evolutionary cohesion’; that is, long-lasting reproductive isolation beyond any doubt. Second, the population needed to be presently available for examinations; but often populations were difficult to find and to examine. Third, researchers were in need of detailed and reliable documentation of the population’s reproductive history. Fourth, and importantly, the population needed to be of what was understood as the optimal size for analysis, as mentioned above. Although, for example, Africans were seen as having been reproductively isolated from Europeans for a long time span, empirically it proved impossible to

63 Biologists define a “cline” as a gradient of morphological or physiological change in a group of related organisms, usually along a line of environmental or geographic transition. 64 See similar research veins on empirical population genetics of zoological species (e.g. Ford’s “ecological genetics” and Dobzhansky’s “genetics of natural populations”). There was not only laboratory research into animal genetics, but also field studies in the variation of wild animal species. 65 Taku Komia, cited in: Demerec (1955). 66 Strandskov (1950), p. 2. 67 Dunn (1959), p. 109.

57

render either population an isolate for research purposes. It seemed easier to postulate isolation for small populations of ‘Aboriginal Australians’ or ‘Native Americans’, as long as they could be declared to have been isolated until the day of the study. Hence, in light of the new theoretical requirements, but also with regard to empirical requirements, it seemed unwise to concentrate on large scale ‘racial units’, or ‘races’, such as ‘the Europids’ or ‘the Africans’. William Laughlin, physical anthropologist at the University of Wisconsin, maintained that the focus should be on the comparison of small, geographically neighboring groups: Strictly speaking, we must rule out such a question as ‘Are the Lapps Mongoloids or Europeans?’ and ask to what degree they are similar to other groups. Such a question takes on more biological or evolutionary meaning when we ask ‘Are the Lapps more similar to western Finns than they are to northern Swedes?’69 Significantly, this turn to an empirical account of relations, relatedness, and relative distances between small, well defined groups shaped the work of scientists studying human variation in the decades to follow. One assumed characteristic of these kinds of populations was that isolates were not simply isolated, given that young people would strive to find partners elsewhere. Isolates were, so to speak, starting points for one-way-routes for genes. An isolate was assumed to only experience genetic “efflux” and no “influx”. That is, people would not marry into the isolate; they would only marry out. The remaining gene pool was therefore losing gene carriers; no intermarriage would introduce new alleles. By the same token, geneticists often worked on isolates they considered extremely unattractive communities for a great proportion of their younger members. They often lived in unattractive places, with slim chances of social advancement and high rates of poverty and unemployment. At the same time, such poor communities were often seen as providing better conditions for research, as the members would presumably be more willing to act as test subjects if incentives were offered. Island populations, at least when they had never been in contact with travelers, seemed to fit the notion of ‘primitive isolates’, figured as backward, underdeveloped and poor, and therefore apt for isolate studies.68 Sandra Widmer’s paper in this issue provides an example, describing how scientists approached niVanuatu as genetically coherent populations. For continental isolates, most often the notion of their being isolated rested on a history of political and social persecution and discrimination. However, as many researchers believed, the contact between populations had recently increasedda worrying situation for population geneticists, as isolates might quickly lose their isolated status. Thus, further requirements needed to be fulfilled; a group isolated by cultural or social means would have to display well documented, strict endogamous marriage rules in the past and present, as allegedly in the case of the Indian caste system, for example. Furthermore, for the comparison of two isolated populations with respect to their allele frequencies, no ‘racial crossing’ should have occurred in the past or present.70 Finally, empirically useful isolates were to be larger thandand thus not to be confused withdinbreeding groups, as one might find in large families or villages in which inbreeding took place. The latter was precisely the unit of interest to medical geneticists, who worked on

69

Laughlin (1960), p. 78. The above cited WHO report offers a classical research program designed for studying “primitive isolates” in any respect. 70 The problem was tackled as a quantitative one by Lasker (1952). 68

58

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

consanguinity and set up pedigrees of hereditary abnormal conditions.71 But for population geneticists, it was necessary to rule out too close inbreeding, which was understood to occur more frequently in smaller isolates and larger families. In their quest to secure an endogamous group for research, population geneticists sought out what was understood as a ‘proper’ degree of relatedness in a population: not too much, not too little. Notably, population geneticists did not require that these units should have been isolated since prehistoric times, as older race concepts had maintained. A time span of a few thousand, even of some hundred years of isolation was understood as enough to bring about differing allele frequencies between two populations.72 Following from these conceptual imperatives and premises, the identification of the endogamous group in practical terms remained the greatest methodological challenge for population geneticists. Thus far in race science, there had been no reason to provide empirical proof of reproductive isolation. By the 1950s, this had changed; now, a considerable part of scientists’ time had to be spent identifying isolates, endogamous groups and their reproductive behaviors. William Laughlin described the course of this research methodology as follows in 1960: The first step in the study of a group by a physical anthropologist has nothing to do with reagents or callipers. Upon wading ashore at a new village his first task is to find out who belongs there, i.e. to define his sample. His only method for delimiting the group consists in defining its members and this is a genealogical endeavour. Cooperation with an ethnographer or linguist is frequently of great advantage to both researchers at this point.73 But beyond taking genealogies, there were more techniques to be employed than Laughlin’s account suggests. Before blood tests and measurements could be carried out, much work was done to stabilize the group under study as an evolutionary coherent isolate. Most of all, scientists were in need of history; that is, they sought out biohistorical narratives that would narrate the supposedly socialdand hence reproductivedisolation of the group. To meet this challenge, they had to rely on experts from other disciplines and even on ‘non-scientific’ knowledge. They conducted expert interviews and interviews with test subjects, had questionnaires filled out and asked for birthplaces or mother tongues of family members. Researchers collected narrations of group history and identity from written or oral mythical sources.74 They browsed through community records, church registers (for birthplaces), and demographic data, and analyzed the distribution of surnames. They sought the help of linguists, ethnographers, historians, cultural and social anthropologists, ethnologists and sociologists, particularly with regard to marriage systems and social isolation.75 For example, Khanolkar, Sanghvi and his students noted as a selfevident fact that their test populations were endogamous, pointing to these populations’ marriage laws, which could be accounted for by interviews. For the historical demarcation of some isolates, they cited mythical traditions, as in the case of the Bene-Israel Jews, a ‘report’ by

71

Lindee (2005), pp. 58e90. Sanghvi & Khanolkar (1950), pp. 52e53. 73 Laughlin (1960), p. 80. 74 The “WHO Scientific Group on Research in Population Genetics of Primitive Groups” noted in 1970: “All groups have learned individuals, e.g., experts on oral traditions and those with systematized knowledge and interpretations of natural phenomena. [.] Such information is pertinent to their cultural and therefore biological history” (World Health Organization, 1970, p. 233). 75 In Congo and Ruanda, geneticist Jean Hiernaux collaborated with linguists, ethnographers, cultural anthropologists (mediating access to the population), demographers, economists and social scientists (Hiernaux, 1966, p. 288). Dunn employed surname analysis, a sociological technique later taken up by CavalliSforza as telephone book analysis. On Dunn’s study, see Lipphardt (2010). 72

David Rahabi, a partly mythical figure who had presumably lived 1000 A.D. or later, that narrated the group history as that of a group of Jewish immigrants. The population’s ongoing status as an isolated one following the arrival, as described by Rahabi, was inferred from their present, unwritten but orally confirmed marriage rules. The primary criterion for sampling was therefore self-identification with the caste and its historicalecultural cohesion. Thus, accounts of isolation and reproduction from all kinds of sources entered the research design, mostly by a priori classifications and biohistorical narratives, and turned out to be essential for studies of human genetic diversity. Just as in race science, ‘purely scientific’ explanations of ‘genetic diversity’ heavily relied on other, non-genetic knowledge. Notably, scientists studying human variation struggled with overtly messy realities; it was very difficult practically to isolate isolates.76 The various actors involved often did not share one concept of endogamy, isolation or group cohesion. Data might have been incomplete, ambiguous, non-reliable or simply not compatible with biologists’ demands. Research literature from the humanities might also have simply reformulated older myths of origin, migration, isolation or miscegenation. Inferring definite evolutionary narratives from these sourcesdthat is, a single narrative per groupdremained problematic. Some scientists were aware of this problem. Already in 1927, Franz Boas’ student, Melville Herskovits, had noted that it might be “hazardous” and risk “jeopardizing any results” to base genealogies on the accounts of the test persons, but he regarded the problem irrelevant in the light of his research design. In the 1950s, the difficulties were registered, but not understood as dire enough to prompt scientists to reflect upon the problematic self-evidence of their assumptions about human variation, as a quote by Leslie Clarence Dunn shows: The other difficulties inherent in the genetical study of human populations do not call for special comment here. Identification of family members, of the families composing a community, determination of the mating pattern and of the continuity of endogamy are problems which each investigator must face and overcome as best he may. They are, so to speak, the price he pays for dealing with animals over which he has no control.77

8. Doubts and critiques in the 1960s In the 1960s, critical awareness began to grow: “The difficulties of defining geographical groups in man have been recognized since the eighteenth century at least and are not diminished by modern data”, Nigel Barnicot wrote. He continued, “the grouping problem is evidently complex and it need hardly be added that no precise definitions have yet been offered, still less agreed [.]”.78 And yet, many of the groups under study as isolates in the 1950s remained on lists of scientifically significant populations kept by institutions such as the WHO Program on Population Genetics of Primitive Groups in the late 1960s, the International Biological Program in the 1970s and the Human Genome Diversity Project in the 1990s. From the 1970s, the notion of the primitive isolate, important in international organizations and academic fields in the 1960s, was heavily critiqued in cultural anthropology, but not in population genetics. Nevertheless, even in population genetics, human variation was increasingly viewed as more complex than isolate studies

76 77 78

Lipphardt (2010); Widmer (2014). Dunn (not dated), p. 6f.; Herskovits (1927), p. 76. Barnicot (1965), p. 86.

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

had accounted for. Clines complicated the picture, and in the late 1960s, Hiernaux, who had investigated ‘racial variation’ on the basis of a priori classifications, began to advocate ‘numerical taxonomy’ as a way to overcome the shortcomings of a priori classification (Hiernaux, 1966). Moreover, in spite of the continuities outlined above, significant changes were underway in population genetics in the 1960s, resulting from new mathematical and biochemical techniques, and from the growing number of genetic markers, rather than from an explicit re-conceptualization of human variation.79 Paradoxically, the boom of markers did not result in clear answers to questions about human variation. Whereas in the 1950s, on the basis of only a handful of markers, racial variation still seemed to be the most intriguing sort of human difference, in the 1960s, when basing their studies on a growing number of proteins, researchers gained a novel sense of the enormous complexities of human genetic variation.80 ‘Racial variation’ was now only seen as one aspect of ‘genetic variation’, and lost its central and defining status for ‘human variation’. And yet, in the eyes of geneticists some decades later, even this paradigmatic shift in understanding human variation was dwarfed by the extent of variation revealed by DNA analysis that came later in the 1980s.81 New technologies of linking datadestablishing correlationsdfurther increased the flexibility of studies in human variation. Once enough data from many populations all over the world were collected, one coulddaccording to geneticistsdcorrelate gene alleles with life habits, diseases or anthropometric data, and thereby gain, for example, insights into the nutritional status of populations, how it might be influenced by inheritance, and how it might link to epidemic concerns. Lisa Gannett’s paper in this issue deals with these interconnected interests surrounding genetic variation. Thus, aside from political reasons, population geneticists’ amazement in response to their new discoveries of variation offers an explanation for why they saw their work as disconnected from the work of pre-WWII scientists and physical anthropologists. However, the earlier predominant search for racial variation left its stamps, or watermarks (Widmer, 2014), in the practises and data of the field. 9. Conclusions In this paper, I have tried to highlight some historical continuities and ruptures in both the concepts and practices through which human variation has been studied. I have argued that, in spite of terminological, conceptual and political adjustments, a biological concept of human variation remained alive and well. I have also sought to demonstrate a previously overlooked disciplinary overlap, namely, between physical anthropology and human genetics in the decades following WWII. At the same time, the scientific curiosities of the physical anthropologists and the human geneticists perfectly complemented one another. On the one hand, the origins and evolution of humankind, along with that of specific human groups, remained an epistemic object of interest to which practitioners of both fields applied new methods and theoretical tools. On the other hand, many turned their scientific attention to how evolutionary factors work in populations; here, the concrete biohistory of certain groups became an epistemic instrument for

79 Significantly, markers pointed to new research directions. Studying their inheritance, and the distribution of their varieties, could yield clues to the pathology or normal function of genes. In this vein, human diversity could function as an epistemic tool for the calibration of a new marker and for the investigation of its function and inheritance. Finding, stabilizing, calibrating and correlating markers became a field of its own. 80 Thompson (1967), p. 68. 81 Cavalli-Sforza (1994), pp. 3e4.

59

explaining and quantifying the emergence of difference. For many of them, now at stake was not just evolution in its abstract formdfor example, how did dark skin color evolve as a trait under selection by sun radiationdbut how these factors had worked in the biohistory of certain groups, located in space and time. Looking at scientists’ attempts to ‘isolate the isolate’, it seems that what they ideally hoped to find was a population that would allow for a precise measure of the genetic baseline of variation, or what was understood as the ‘natural degree of variation’ that seemed to occur in any isolated population. This ideal condition would then also have allowed for a more rigorous study of the factors that led to these differences. However, as no universal standards emerged from those studies, one wonders whether the fervent attempts to document population differences also aimed at establishing specific baselines for diverse populations. Human populations, dispersed, adapted and selected in different environments, seemed to reveal meaningful patterns of diversity, which could be used for the study of many further epistemic objects, including modes of inheritance, pathological versus normal conditions of certain human features, and demographic developments. Variation did a great deal of work for scientists who sought out the most stabledthat is, heritablediterations of variation in their research. Assuming the emergence of variation in evolution also helped them with practical tasks, taking the outcome of evolutiondthat is, diverse populationsdas the research object was a convenient way of demarcating populations in practice. What remaineddand persists in human population genetics to some degree even todaydwas a certain preoccupation with specific groups, and especially with presumably isolated ones.82 This is remarkable if one considers that at the same time the political situation of most of these populations had changed dramatically. Until decolonization, many of these groups were subject to colonial rule, and the specific form of scientific research these colonial situations made possibledand how it shifted during and following decolonizationddeserves more attention. Given that the Cold War also heavily influenced forms of transnational investigations of populations (Radin, in this issue), the continuities are ever more striking.83 Furthermore, to investigate human diversity meant to engage in dialogue with humans in order to elicit information from individuals about their origins and the communities to which they belonged. Eventually, this also meant intervening into their private lives. In any case, investing much empirical work into identifying an isolate had not been necessary earlier. This was increasingly so after the 1950s, and with this shift came multiple new challenges. Notably, scientists pursuing empirical research rarely discussed the underlying assumptions of temporality in their research on variation. However, the temporalities of human biological evolution underwent a thorough re-consideration in this period. Throughout the late 19th and early 20th centuries, scientists partaking in debates about ‘race’ and ‘human variation’ had rather implicitly drawn on very different notions of changeability in time. While some assumed that races had been stable since prehistoric times, others claimed that what their colleagues considered as fixed traits were indeed characteristics that were malleable within only one or two generations. Then in the 1950s, most scientists assumed the temporality of evolutionary change to work within a range of several hundreds to several thousands of years.84

82

Cf. Reardon (2005). On the other hand, new political situations also brought about new questions for human population geneticists: for example, the question of a potential natural adaptation to radioactivity of some geographic populations. I am grateful to Alexander von Schwerin and Susanne Bauer for having brought this to my attention. 84 Cf. Reardon (2005). 83

60

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61

In this transformative period of cooperation, scientists studying and reflecting on human variation added data to data, anthropometrics to population genetics, from all continents and regions, and thereby helped to pave the way for a new paradigmatic practice in the human sciences, that is, the practice of correlating data. Data, even if collected using methodologies that were later contested, could not simply be discarded; any kind of difference could gain importance if strong correlations with other differences could be established. The data accumulated by anthropometric methods over decades in the early 20th century was considered an extremely valuable resource for potential correlations with differences established by new methodologies. Acknowledgments Many thanks to Jenny Bangham and Tania Munz for their helpful comments on this paper, and to Frederike Heinitz and Katrin Kleemann for their support. Early versions of this paper were discussed at two consecutive workshops and in a research colloquium at the MPI for the History of Science; many thanks to all participants who have commented on the paper. References Bangham, J. (2013). Between the transfusion services and blood group research: Human genetics in Britain during the Second World War. In E. Ramsden, & S. Muller-Wille (Eds.), A cultural history of human heredity. London: Pickering and Chatto. Bangham, J. (2014). Blood groups and human groups: Collecting and calibrating genetic data after World War Two (in this volume). Barkan, E. (1992). The retreat of scientific racism: Changing concepts of race in Britain and the United States between the world wars. Cambridge: Cambridge University Press. Barkan, E. (1996). The politics of race in science: Ashley Montagu and UNESCO’s anti-racist declarations. In L. T. Reynolds, & L. Lieberman (Eds.), Race and other misadventures: Essays in honor of Ashley Montagu in his ninetieth year (pp. 96e 105). New York: General Hall. Barnicot, N. (1965). The scientific problem of race in the light of recent knowledge. International Social Science Journal, 17(1), 85e87. Barrai, I., Cavalli-Sforza, L. L., & Moroni, A. (1962). Frequencies of pedigrees of consanguineous marriages and mating structure of the population. Annals of Human Genetics, 25(4), 347e377. Bernstein, F. (1928). Über mendelistische Anthropologie. In H. Nachtsheim (Ed.), Verhandlungen des 5. Internationalen Kongresses für Vererbungswissenschaft, Berlin 1927 (pp. 431e438). Leipzig: Bornträger. Birdsell, J. B. (1950). Some implications of the genetical concept of race in terms of spatial analysis. Cold Spring Harbor Symposium on Quantitative Biology: Origin and Evolution of Man, 15, 259e314. Bowler, P. J. (2009). Evolution: The history of an idea. Berkeley, CA: University of California Press. Brattain, M. (2007). Race, racism, and antiracism: UNESCO and the politics of presenting science to the postwar public. American Historical Review, 112(5), 1386e 1413. Cain, J. (2009). Rethinking the synthesis period in evolutionary studies. Journal of the History of Biology, 42(4), 621e648. Cavalli-Sforza, L. L. (1956). Some notes on the breeding patterns of human populations. In Proceedings of the first international congress of human genetics (Vol. 2, pp. 395e399). Basel & New York: Acta Genetica et Statistica Medica. Cavalli-Sforza, L. L. (1994). The history and geography of human genes. Princeton: Princeton University Press. Cold Spring Harbor Symposium on Quantitative Biology 15: Origin and Evolution of Man. (1950). Dahlberg, G. (1943). Race, reason and rubbish: An examination of the biological credentials of the Nazi creed. London: Allen and Unwin. Demerec, M. (1955). Foreword. Cold Spring Harbor Symposia on Quantitative Biology: Population Genetics: The Nature and Causes of Genetic Variability in Populations, 20. v. Dobzhansky, T. (1950a). Human diversity and adaptation. Cold Spring Harbor Symposium on Quantitative Biology: Origin and Evolution of Man, 15, 385e400. Dobzhansky, T. (1950b). Mendelian populations and their evolution. The American Naturalist, 84(819), 401e418. Dobzhansky, T. (1962). Mankind evolving: The evolution of the human species. New Haven: Yale University Press. Dunn, L. C. (not dated). A genetical study of a Jewish community, the old Ghetto Community of Rome. Manuscript. Dunn, L. C. (1959). Heredity and evolution in human populations. Cambridge, MA: Harvard University Press.

Fischer, E. (1930). Versuch einer Genanalyse des Menschen: mit besonderer Berücksichtigung der Anthropologischen Systemrassen. Berlin: Bornträger. Fisher, R. A. (1949). The theory of inbreeding. Edinburgh and London: Oliver & Boyd. Gannett, L. (2001). Racism and human genome diversity research: The ethical limits of “population thinking.” Philosophy of Science, 68(3), 479e492. Gannett, L., & Griesemer, J. (2004). The AB0 blood groups: Mapping the history and geography of genes in Homo sapiens. In H.-J. Rheinberger, & J.-P. Gaudillière (Eds.), Classical genetic research and its legacy: The mapping cultures of twentiethcentury genetics (pp. 119e172). London: Routledge. Garn, S. M. (1962). The newer physical anthropology. American Anthropologist, 64(5), 917e918. Gayon, J. (2003). Do biologists need the expression “human race”? UNESCO 1950e 1951. In J. Rozenberg (Ed.), Bioethical and ethical issues surrounding the trials and code of Nuremberg (pp. 23e48). New York: Edwin Melon Press. Glass, B. (1953). Genetic changes in human populations, especially those due to gene flow and genetic drift (reprint from Advances in Genetics). Yearbook of Physical Anthropology, 9, 161e207. Glass, B., Sacks, M. S., Jahn, E. F., & Hess, C. (1952). Genetic drift in a religious isolate: An analysis of the causes of variation in blood group and other gene frequencies in a small population. The American Naturalist, 86(828), 145e159. Goldschmidt, E. (Ed.). (1963). The genetics of migrant and isolate populations: Proceedings of a conference on human population genetics in Israel held at the Hebrew University, Jerusalem. Baltimore: Williams & Wilkins. Gormley, M. (2009). Scientific discrimination and the activist scientist: L.C. Dunn and the professionalization of genetics and human genetics in the United States. Journal of the History of Biology, 42(1), 33e72. Herskovits, M. (1927). Variability and racial mixture. The American Naturalist, 61(627), 68e81. Hiernaux, J. (1954). A new find of dimple-based pottery in Africa. Man, 54, 164. Hiernaux, J. (1955). Physical anthropology and the frequency of genes with a selective value: The sickle cell gene. American Journal of Physical Anthropology, 13(3), 455e472. Hiernaux, J. (1963). Heredity and environment: Their influence on human morphology. A comparison of two independent lines of study. American Journal of Physical Anthropology, 21(4), 575e589. Hiernaux, J. (1965a). Introduction: The Moscow Expert Meeting. International Social Science Journal, 17(1), 73e84. Hiernaux, J. (1965b). The application of the concept of race to the human species. International Social Science Journal, 17(1), 112e114. Hiernaux, J. (1966). Human biological diversity in Central Africa. Man, 1(3), 287e 306. Hubinot, M. D., Hiernaux, J., Massart-Guiot, M. D., & Massart-Guiot, T. (1953). Blood groups of the ABO, MN and CDE-cde systems in the native populations of RuandaeUrundi territories. Annals of Eugenics, 18, 13e21. Kalmus, H. (1957). Defective colour vision, P.T.C. tasting and drepanocytosis in samples from fifteen Brazilian populations. Annals of Human Genetics, 21, 313e 317. Kraus, B. S., & White, C. B. (1956). Micro-evolution in a human population: A study of the social endogamy and blood type distributions among the Western Apache. American Anthropologist, 54(3), 433e436. Lasker, G. W. (1952). Mixture and genetic drift in ongoing human evolution. American Anthropologist, 54(3), 433e436. Laughlin, W. S. (1960). Aspects of current physical anthropology: Method and theory. Southwestern Journal of Anthropology, 16(1), 75e92. Lindee, S. (2005). Moments of truth in genetic medicine. Baltimore: John Hopkins University Press. Lindee, S., & Santos, R. V. (2012). The biological anthropology of living human populations: World histories, national styles, and international networks: An introduction to supplement 5. Current Anthropology, 53(S5), 3e16. Lipphardt, V. (2010). The Jewish community of Rome: An isolated population? Sampling procedures and bio-historical narratives in genetic analysis in the 1950s. BioSocieties, 5(3), 306e329. Lipphardt, V. (2012). Isolates and crosses: Human evolution and population genetics in the mid-twentieth century. Current Anthropology, 53(S5), 69e82. Lipphardt, V., & Niewöhner, J. (2007). Producing difference in an age of biosociality. Biohistorical narratives, standardisation and resistance as translations. Science, Technology & Innovation Studies, 3(1), 45e65. Livingstone, F. B. (1962). On the non-existence of human races. Current Anthropology, 3(3), 279e281. Marks, J. (1995). Human biodiversity: Genes, race, and history. New York: Aldine de Gruyter. Marks, J. (1996). The legacy of serological studies in American physical anthropology. History and Philosophy of the Life Sciences, 18(3), 345e362. Marks, J. (2008). Race across the physicalecultural divide in American Anthropology. In H. Kuklick (Ed.), A new history of anthropology (pp. 242e258). Malden, MA: Blackwell. Mazumdar, P. M. H. (1996). Two models for human genetics: Blood grouping and psychiatry in Germany between the World Wars. Bulletin of the History of Medicine, 70(4), 609e657. Mendelsohn, A. (2001). Medicine and the making of bodily inequality in twentiethcentury Europe. In J.-P. Gaudillière, & I. Löwy (Eds.), Heredity and infection: A history of disease transmission (pp. 21e79). London: Routledge. Mitchell, S. D., & Dietrich, M. R. (2006). Integration without unification: An argument for pluralism in the biological sciences. The American Naturalist, 168(S6), 73e79.

V. Lipphardt / Studies in History and Philosophy of Biological and Biomedical Sciences 47 (2014) 50e61 Moor-Jankowski, J. K., & Huser, H. J. (1956). Sero-anthropological investigations in the Walser and Romansh isolates in the Swiss Alps and their methodological aspects. In Proceedings of the first international congress of human genetics (Vol. 6, pp. 527e 531). Basel & New York: Acta Genetica et Statistica Medica. Mourant, A. E. (1956). Blood groups and human evolution. The Advancement of Science, 13(50), 91e103. Müller-Wille, S., & Rheinberger, H.-J. (2008). Race and genomics: Old wine in new bottles? NTM e Zeitschrift für Geschichte der Wissenschaften, Technik und Medizin, 16(3), 363e386. Müller-Wille, S., & Rheinberger, H.-J. (2012). A cultural history of heredity. Chicago: Chicago University Press. Phillips, G. (1928). The blood groups of full-blood Australian Aborigines. Medical Journal of Australia, 15(2), 296. Pogliano, C. (2005). L’ossessione della razza: Antropologia e genetica nel XX secolo. Pisa: Scuola Normale Superior. Proceedings of the first international congress of human genetics, Part II, Section: Studies on isolates. (1956). In Acta Genetica et Statistica Medica (Vol. 6). Proceedings of the second international Congress of human genetics. (1961). Rome: Istituto G. Mendel. Radin, J. (2014). Unfolding epidemiological stories: How the WHO made frozen blood into a flexible resource for the future (in this volume). Reardon, J. (2004). Decoding race and human difference in a genomic age. Differences: A Journal of Feminist Cultural Studies, 15(3), 38e65. Reardon, J. (2005). Race to the finish: Identity and governance in an age of genomics. Princeton: Princeton University Press. Rose, N. (2007). Race in the age of genomic medicine. In Politics of life itself: Biomedicine, power and subjectivity in the twenty-first century (pp. 155e186). Princeton: Princeton University Press. Sanghvi, L. D. (1953). Comparison of genetical and morphological methods for a study of biological differences. American Journal of Physical Anthropology, 11(3), 385e404. Sanghvi, L. D. (1954). Genetic diversity in the people of Western India. Eugenics Quarterly, 1(4), 235e239. Sanghvi, L. D. (1966). Inbreeding in India. Eugenics Quarterly, 13(4), 291e301. Sanghvi, L. D., & Khanolkar, V. R. (1950). Data relating to seven genetical characters in six endogamous groups in Bombay. Annals of Eugenics, 15, 52e76. Sanghvi, L. D., Varder, D. S., & Master, H. R. (1956). Frequency of consanguineous marriages in twelve endogamous groups in Bombay. Acta Genetica et Statistica Medica, 6, 41e49. Schneider, W. H. (1995). Blood group research in Great Britain, France and the United States between the World Wars. Yearbook of Physical Anthropology, 38, 77e104. Schull, W. J., & Neel, J. V. (1963). The effect of inbreeding on mortality in Japan. In Proceedings of the Second International Congress of Human Genetics (Vol. 1, pp. 60e70). Rome: Istituto G. Mendel. Science, 100, (1944). Simmons, R. T., Graydon, J. J., & Semple, N. M. (1954). A blood group genetical survey in Australian aborigines. American Journal of Physical Anthropology, 12(4), 599e 606.

61

Smocovitis, V. B. (2012). Humanizing evolution. Anthropology, the evolutionary synthesis, and the prehistory of biological anthropology, 1927e1962. Current Anthropology, 53(suppl. 3), 108e125. Sommer, M. (2008a). History in the gene: Negotiations between molecular and organismal anthropology. Journal of the History of Biology, 41(3), 473e528. Sommer, M. (2008b). Bones and ochre: The curious afterlife of the Red Lady of Paviland. Cambridge: Harvard University Press. Spörri, M. (2012). Reines und gemischtes Blut: Zur Kulturgeschichte der Blutgruppenforschung, 1900e1933. Bielefeld: Transcript. Stepan, N. L. (1984). The idea of race in science: Great Britain 1800e1960. Basingstoke: Macmillan. Stepan, N. L. (2003). Science and race: Before and after the Human Genome Diversity Project. Socialist Register, 329e346. Special Issue on “Fighting identities: race, religion and ethno-nationalism”. Stern, C. (1949). Principles of human genetics. San Francisco: Freeman. Strandskov, H. H. (1950). Genetics and the origin and evolution of man. Cold Spring Harbor Symposia on Quantitative Biology, 15, 1e11. Strandskov, H. H., & Washburn, S. L. (1951). Genetics and physical anthropology. American Journal of Physical Anthropology, 9(3), 261e264. Tapper, M. (1999). In the blood: Sickle cell anemia and the politics of race. Philadelphia: University of Pennsylvania Press. Thompson, L. (1967). Steps toward a unified anthropology. Current Anthropology, 8(1/2), 67e91. Wailoo, K., & Pemberton, S. G. (2006). The troubled dream of genetic medicine: Ethnicity and innovation in Tay-Sachs, Cystic Fibrosis, and Sickle Cell Disease. Baltimore: Johns Hopkins University Press. Widmer, A. (July 2014). The imbalanced sex ratio and the high bride price: Watermarks of race in demography, census, and the colonial regulation of reproduction. Science, Technology & Human Values, 39, 538e560. Widmer, A. (2014). Making ‘Melanesian’ blood: Fieldwork and isolating techniques in genetic epidemiology (1963e1976) (in this volume). Wiedemann, F. (2012). Völkerwellen und Kulturbringer. Herkunfts- und Wanderungsnarrative in Historisch-archäologischen Interpretationen des Vorderen Orients um 1900. Ethnographisch-Archäologische Zeitschrift, 1(2), 105e128. World Health Organization. (1964). In Research in population genetics of primitive groups: Report of a WHO Scientific Group (Vol. 279). Geneva: World Health Organization. World Health Organization. (1970). Research on human population genetics: Report of a WHO scientific group. Current Anthropology, 11(2), 225e233. Wright, S. (1932). The Roles of mutation, inbreeding, crossbreeding and selection in evolution. In Proceedings of the sixth international congress of genetics (Vol. 1, pp. 356e366). Wright, S. (1952). The genetics of quantitative variability. In E. C. R. Reeve, & C. H. Waddington (Eds.), Quantitative inheritance: papers read at a colloquium held at the Institute of Animal Genetics Edinburgh University under the auspices of Agricultural Research Council, April 4th to 6th, 1950 (pp. 5e41). London: H.M.S.O.