The comparative method in nineteenth-century anthropology was born of four elements, none of which was peculiar to anthropology: one, an approach to collecting information; two, an established style of dividing intellectual labour; three, a methodological orientation; four, an aesthetic of generalization and explanation.

Collecting information: There was a standard way to collect information about any given specific object, creature, or practice that was found in many places — the questionnaire. The most important of anthropologists’ questionnaires was the volume jointly published by the British Association for the Advancement of Science and the Royal Anthropological Institute at irregular intervals from 1874 to 1951, Notes and Queries on Anthropology (it was originally entitled Notes and Queries on Anthropology for the Use of Travellers and Residents in Uncivilized Lands).^[13] This was not the only anthropological questionnaire circulated; individuals such as J.G. Frazer, for example, drew up and distributed their own questionnaires. But it is important to recognize that Notes and Queries on Anthropology merely elaborated a general form. This had taken shape no later than the late seventeenth century, when some Fellows of the Royal Society printed and circulated a questionnaire to elicit information about the natural and built features of the environments of England and Wales.^[14]

Consider the periodical entitled simply Notes and Queries, a vehicle for inquiring minds who were curious about virtually anything. On March 18, 1854, for example, it published a query from a reader who seemed bent on acquiring confirmation of his judgement that mackerel were blind (his explanation for their lack of interest in the flies he cast in the waters in which he wished to catch them).^[15] On November 28, 1857, Notes and Queries published a request for information about the career of a small child who had at some time been exhibited in London, in the irises of whose eyes was said to be visible the name “Emperor Napoleon”.^[16] And on February 1, 1862, it published readers’ responses to a question about the number of societies past and present in which human corpses were ceremonially buried in the foetal position. One of these responses came from Sir John Lubbock (a baronet, later Lord Avebury), a scientific polymath with special expertise in insects, who knew Charles Darwin as a neighbour, family friend and intellectual mentor, and who was a long-serving member of Parliament as well as a wealthy banker. Lubbock’s many accomplishments included being the first president of the Anthropological Institute of Great Britain and Ireland (later the Royal Anthropological Institute), formed in 1871 from the union of the previously warring Ethnological Society of London and the Anthropological Society of London.^[17] That Lubbock was among the readers of Notes and Queries, along with the incompetent mackerel fisherman, surely indicates that this publication reached a broad audience.

Then, there existed the informal questionnaire, such as the letters of inquiry that Charles Darwin mailed to his global network of naturalist-informants, without whose assistance he could never have gathered the wealth of data from which he produced such works as the On the Origin of Species in 1859.^[18] Last, but hardly least, there were information-gathering kits, such as those for collecting insects that would-be purchasers of specimens distributed to travellers to foreign parts (such as naval officers and traders); these included operating instructions, as well as the materials necessary to kill, preserve and store insect specimens without mutilating them.^[19] Of course, individual scholars supplemented the material they gleaned from systematic inquiries with information they happened upon in various ways, but it is the existence of a variety of routinized forms of deliberate solicitation of information that is significant.

Division of labour: There was a clear division of labour between collectors of information and analysts thereof, which represented a division along class lines — roughly, the division between players and gentlemen, respectively. Those who received compensation for their scientific activities, as collectors often did, were not considered capable of judging evidence dispassionately — and also were not, in any event, “clubbable”, almost invariably being considered unfit for leadership roles in the societies of enthusiastic amateurs who dominated many spheres of British scientific life until the twentieth century. Consider the case of Paul du Chaillu, a French-born traveller to equatorial Africa in the middle of the nineteenth century (who had various financial supporters, including, for example, the Academy of Natural Sciences in Philadelphia). When he made public appearances in London, his descriptions were greeted with scepticism and he himself was widely believed to be something of a cad: his reports of African peoples and their natural environment, which included the first eyewitness observations of the gorilla, were accepted only after authentication by gentlemen-scientists. The gentlemen-scientists had no empirical basis for their judgements, basing them on their assessments of du Chaillu’s character. Evidently, their high social status counted for more in scientific circles than du Chaillu’s research experience.^[20]

I must emphasize that to report this standard division of labour is not to suggest that we must identify with the view of the gentlemen-amateurs, or that we must believe that the collectors who served gentlemen-scientists regarded their role as that of mere servants. As Anne Secord has ingeniously documented, working-class collectors who supplied nineteenth-century gentlemen-botanists with specimens had considerable scientific expertise, as well as genuine commitment to scientific inquiry.^[21] But no matter what were the satisfactions working-class collectors derived from their scientific labours, their voices were not audible in prestigious scientific circles.

Methodological orientation: The collection of evidence from thither and yon somehow had to be rationalized as a reasonable procedure. Intellectual historians must begin with the working principle that the figures we study were at least as smart as we are, and so we should not be surprised to learn that many nineteenth-century anthropologists were aware that their informants were not presenting them with what the philosopher of science calls “brute facts” — as she customarily does, immediately prior to declaring that there are no such things. At least some of anthropologists' disciplinary ancestors recognised that, as the philosopher Mary Hesse has memorably stated, theories are always “underdetermined by facts” — a point especially relevant in accounts of paradigm change (and to which we will return).^[22] In 1870, for example, John Lubbock observed that the particular perspectives of individuals shaped their reports, saying, “Whether any given writer praises or blames a particular race, depends at least as much on the character of the writer as on that of the people.”^[23] Indeed, attention to observer bias had a considerable lineage in science, dating at least to the astronomers’ definition of the so-called personal equation in the 1820s, a phenomenon of which Sir John was surely aware, since his father was a distinguished astronomer.^[24] That is, by the end of the eighteenth century, astronomers had noted that individual observers varied in their reaction times to, say, the transit of a star, and astronomers subsequently undertook to calculate patterns of variations among individuals, specifying individuals’ differences, so as to achieve inter-subjective measures by which vital matters such as the setting of clocks could be resolved.^[25]

It is no accident, to use historians’ once-canonical locution, that Francis Galton delivered descriptions of some of his most important statistical innovations to the Anthropological Institute, of which he was President from 1885 to 1888, as he did when he described the normal frequency distribution (also known as the bell curve) which should be observable for any variable, ranging from, say, height to life expectancy; he intended his statistics to be useful in describing the phenomena anthropologists examined, including cultural and biological traits. In his history of the development of statistics in Britain, Donald MacKenzie emphasizes that Galton and others, most notably Karl Pearson, who developed such statistical measures still in use as the correlation coefficient, were motivated by their desire to document patterns of human heredity so that they could promulgate eugenic strategies for improving the physical and mental condition of the British population.^[26] But Galton was surely aware of anthropologists’ problem in reconciling disparate reports in order to formulate generalizations. And if he did not say explicitly that he expected his statistical techniques to solve anthropologists’ methodological problem, another sometime president of the Anthropological Institute, E. W. Brabrook, did, stating that anthropologists need not trouble themselves with inconsistencies in the reports they received from “every direction”, but could trust “rather to the general laws of numbers than to the skill of individuals to eliminate errors”.^[27]

An aesthetic of generalization: The comparative method in anthropology was born in an era in which it was assumed that all satisfactory explanations were historical ones. It was not merely the case that historical analogies were frequently made, and history examined for its moral lessons, as I have already discussed. The general idea of evolution, once called “transformism”, antedated Darwin’s conceptualization of it and was applied to description of various natural phenomena. Consider the practice of embryology in the last quarter of the nineteenth century, when it was a highly prestigious area of research in the biological sciences. Embryological research was animated by the recapitulation hypothesis, which could be applied more or less strictly. Compatible with Darwinism (although also reconcilable with other developmental schemes), this was the idea that ontogeny recapitulated phylogeny, that the development of each individual recapitulated the history of the development of the species to which it belonged. Thus, examination of embryonic growth patterns was expected to yield answers to the fundamental questions of species’ evolution. For example, a development in an embryo that proved just a transient phase in the progress of an organism to its infant stage could indicate whether degeneration was just as important a feature of its evolutionary history as progress — a possibility much debated at the end of the century. (In the developmental scheme so influentially promulgated by Herbert Spencer in the late-nineteenth century, progress generally meant a movement from simplicity to complexity.)^[28] To give another example of the use of the recapitulation hypothesis, Sigmund Freud's Lamarckian explanation of the structure of the individual psyche conjured up real human experiences in the remote historical past that would be recapitulated in childhood maturation — which could constitute degeneration (or at least arrested development) if the full course of the species’ development was not followed.^[29]

In the first decades of the twentieth century, all manner of varieties of historical explanations would be dismissed as such, as expressions of what philosophers termed the “genetic fallacy” — a pejorative phrase expressing the idea that the origins of any given institution in the past had no necessary relevance to understanding the operations of that institution in the present. Nowadays, facile dismissals of assertions as expressions of the genetic fallacy are rarely heard, although we can easily conjure up illustrations of arguments that might be dismissed as based on the genetic fallacy. For example, we should not devalue Galton and Pearson’s statistics because the two men were motivated by eugenic objectives, since their statistics can be used in all manner of research projects. At the same time, we can also see that the historical antecedents of some present practices may be meaningful in contemporary contexts. And historical explanations are back in fashion. Regardless, appeals to historical explanations are themselves historical phenomena; they have seemed more plausible in some eras than in others.