1. A career and a thesis

The many interests and achievements of André Leroi-Gourhan (1911–1986) are becoming increasingly familiar to various audiences across the social sciences and the humanities, in France and internationally as well.² The later phases of Leroi-Gourhan’s career are particularly well known, including his documentation and interpretation of Palaeolithic parietal art in its Franco-Cantabrian heartland from the late 1950s,³ his pioneering excavation methodologies at the open-air Magdalenian site of Pincevent in the Paris basin from 1964 onwards,⁴ and his major publication, Le Geste et la Parole (1964–1965), whose technological and paleoanthropological propositions were soon embraced by proponents of postmodern philosophical anthropology and ‘French theory’ from Jacques Derrida to Bernard Stiegler, en passant by Tim Ingold.⁵

During the first years of his scientific activities, however, Leroi-Gourhan’s disciplinary affinities were mostly with ethnology, museology and orientalism, in the footsteps of his early mentors, the sociologist Marcel Mauss (1872–1950), nephew and heir to Émile Durkheim, and the physical anthropologist, Americanist and museum director Paul Rivet (1876–1958). Returning in 1939 from a two years long research bursary in Japan, Leroi-Gourhan spent the war-years in occupied Paris, ensconced in its oriental art museums and in the newly-inaugurated cutting-edge ethnographic Musée de l’Homme. This enabled him to research and to complete the two volumes of Evolution et techniques (L’Homme et la Matière and Milieu et techniques).⁶ Following his pioneering article on ‘L’Homme et la nature’ in the 1936 Encyclopédie française,⁷ Leroi-Gourhan drew on his observations and documentation of ‘traditional’ material practices, in Japan and elsewhere, to further develop his highly original classification and conceptualisation of techniques. This classification was built around what he called the ‘elementary means of action on matter’ (most notably ‘percussions’), enriched with the philosophy-inspired notions of tendances (general, historical tendencies) and faits (particular, ethnographically observed facts). At the same time, Leroi-Gourhan begun working on his doctorate, notably in order to be eligible for a newly-created lectureship in ‘colonial ethnology’ at the University of Lyon. To this end, he assembled together a series of lectures he had delivered in 1941–1942 as a guest lecturer at the Collège de France. Titled Archéologie du Pacifique-Nord. Matériaux pour l’étude des relations entre les peuples riverains d’Asie et d’Amérique, his thesis was directed by Paul Rivet, then in self-imposed exile in South America in protestation at the collaborationist Vichy regime. Soon after the successful defence of his thesis in June 1944, Leroi-Gourhan secured the university position in Lyon, where he lectured for the coming decade. In 1956, he took over the chair of ‘General ethnology’ at the Sorbonne, from which he conducted his archaeological and prehistoric art investigations. The crowning academic and institutional achievement of his career came in 1969, with his election to the chair of ‘Prehistory’ at the Collège de France which he held until his retirement in 1982 (Figure 1).

What these professional milestones also attest to is a decades-long process of conceptual and disciplinary mobility, broadly stretching from ethnology (or ethnography) towards prehistoric archaeology, and from ‘comparative’ to ‘prehistoric technology’. Leroi-Gourhan’s growing interests in prehistory, including its distinctive excavation methods and finds analysis, essentially stone artefacts, certainly reflected his changing research orientations and his academic opportunities in Lyon, where his chair was officially dedicated to ‘Ethnologie coloniale’. At the same time, at least in part, this turn to prehistory also resulted from a gradual (and barely acknowledged) disengagement from ethnology within the post-war scientific landscape, with the overtaking of the traditional museum-based approach that had long been his own by more enticing and proactive Marxist, Structuralist and Structural-Marxist schools of socio-cultural anthropology. At the same time, Leroi-Gourhan’s disenchantment was also due to the tacit repositioning or reassessment of his own theoretical and interpretative ambitions. Some of the shifts and challenges at stakes can be perceived in his archival holdings, the bulk of which are kept in a dedicated archive service at the Maison des sciences de l’Homme – Mondes (MSH-M, formerly MAE) in Nanterre. More specifically, there is much for us to learn and reflect on in a file labelled ‘Arch.MSH-M. ALG 70’, which includes the printers’ proofs of Leroi-Gourhan’s doctoral dissertation, published in 1946 as volume 47 of the Travaux et Mémoires de I’Institut d’Ethnologie.⁹

2. Back and forth proofs: 3 May to 14 June 1946

In line with the academic conventions of the times, Leroi-Gourhan’s doctoral research was an essentially descriptive and documentary undertaking. His main aim was to identify and assess the range of relevant evidence that could serve to reconstruct long-term connections and exchanges between Asia and America, interactions which resulted in a distinctive Trans-Pacific civilisation commune. In considering the relevance of a whole array of material items, Leroi-Gourhan singled-out three useful criteria: these items were to be (a) small in size and therefore easily transportable, (b) made from unperishable or at least durable materials, and finally (c) of sufficiently straightforward and identifiable function. This enabled him to highlight such selected ‘themes’ as axes, adzes, hammers, knives, pottery, stone lamps, pipes, spear-throwers and harpoons… all successively described, drawn, located and mapped across the chapters of his thesis. Leroi-Gourhan could not be oblivious to the culture-historical and diffusionist setting in which his research was conducted, and he did at times acknowledge some of this paradigm’s central questions; namely, whether these were people, or ideas, or objects that moved over space and time. Likewise, he stressed that eventual similarities or ‘kinship’ between material items found on either side of the Pacific could be explained by either population migrations, borrowing between groups, or the convergence of technical tendances.¹⁰ Without ignoring then these questions, Leroi-Gourhan remained adamant throughout this work that speculations on these matters were still far too ‘premature’, and that researchers should aim for ‘security’ and seek to be ‘bridled by the facts’, all the more so that ‘Philosophical speculations are at the ends of the human sciences; in no way can they pass as an instrument of investigation’¹¹ – all this, again, well in tune with the positivist empiricism prevailing in his research milieu.

The theory, when it came (in the published version of this thesis) appeared as a bold and even daring epistemological claim – and it came about, quite literally, as an aesthetically-motivated space-filling afterthought on the very last page of his book. This can be confirmed through a brief and targeted foray into Arch.MSH-M. ALG 70 and the proof-correcting process it attests to. Starting in May 1946, Leroi-Gourhan and Protat Frères, the typesetters-printers of the Institut d’ethnologie monograph series based in Mâcon (Saône-et-Loire), engaged in a sustained back-and-forth correspondence concerning these proofs – as indicated by the ‘expedition’ and ‘reception’ dates stamped on these pages. The proofs showing the concluding sentences of the book, namely page number 513, were thus expedited by the printers on 3 May 1946, and they generated two distinct kinds of corrections.

The first rectifications concerned the occurrence on that page of several spelling errors, as well as typographic misplacements, including some inverted or upside-down characters. For example, the last paragraphs displayed the words ‘oonsisterait’ instead of ‘consisterait’, but also ‘qu _i’ instead of ‘qui’, ‘documəntation’ for ‘documentation’ or again ‘spéɔulation’ for ‘spéculation’. These particular ‘typos’ actually deserve a closer reading (see below), but Leroi-Gourhan himself merely indicated their rectification with conventional marginal notations. Received back by Protat Frères on 24 May 1946, these requested corrections were duly integrated into the next set of proofs, sent out on 4 June.

Another series of corrections to these proofs of p. 513 was however of quite a different order. They no longer concerned some straightforward ‘objective’ mistakes at the level of individual typographic characters, but rather, so it seemed, the aesthetics of the page layout, its topographic ‘character’ as a whole. As this layout appeared (on 3 May), the final sentence of the concluding chapter and of the whole book – ‘Ainsi peut-on véritablement construire’ [Thus can we effectively build] – found itself tucked at the very end of the printed space of p. 513. What is more, this ultimate sentence was immediately overlying what appeared to be a running title in smaller font-size (‘A. Leroi-Gourhan. – Archéologie du Pacifique Nord’) followed by the number ‘33’ (also in smaller font) (Figure 2).

This configuration (more on which below) apparently jarred Leroi-Gourhan’s sense of harmony or propriety; he consequently took measures to have his book end on a further (new) page, where the ‘terminal’ status of his last lines of text could clearly stand out. As he put it in the (encircled) comment jotted on the proofs he returned to the printers: ‘Si la chose est possible, il vaudrait mieux détacher les deux derniers paragraphes et placer les dix dernières lignes en haut de page’ [If it is possible, it would be better to detach the last two paragraphs and to place the ten last lines on top of page]. Indeed he imagined, in his typesetting ‘thought-experiment’, that widening the space between the paragraphs on p. 513 would achieve the desired ‘sliding’ of its last lines of text into the next page. The printers Protat Frères clearly understood this request, but they doubted with their professional nous that this gap alone would suffice: they therefore fatefully replied (on 24 May) to Leroi-Gourhan, ‘Il faudrait pour cela ajouter quelques lignes de composition’ [It would then be necessary to add some lines of composition].

Thus it was then, acting as-per-instructions in order to occupy and to make available blank space, that Leroi-Gourhan turned to ‘add a few lines’ in the next set of proofs of p. 513 he sent out 4 June 1946 (Figure 3). One such addition, in the middle-of-the-page paragraph, was a 32 words long sentence (amounting to 3 lines or so of manuscript text) written on the left hand margin, in which he granted the provisional character of his work – not only for staple positivist reasons, but also as an acknowledgment of the unique historical circumstances of its writing: ‘Par sa nature même, [ce travail] est condamné à être dépassé rapidement, d’autant plus rapidement qu’il a été édifié dans des années où les contacts avec la vie du Pacifique nous étaient interdits’ [By its very nature, this work is condemned to be soon overtaken [by new data], all the more quickly that it has been produced during years when contacts with the life of the Pacific were forbidden to us]. A cue to the second addition was placed (encircled) on the bottom right margin: ‘voir au verso’ [See on verso]. Turning then to this still unprinted blank verso, Leroi-Gourhan first inserted a further printers’ indication (encircled), requesting to ‘ajouter au verso pour faire tomber les 8 ou 10 dernières lignes en haut de page’ [Add [the following] to the verso in order to have the 8 or 10 last sentences fall on the top of [next] page] (Figure 4). To achieve this ‘positioning‘, he wrote a paragraph – 62 words in 3 sentences and 8 lines of manuscript text – in which he reiterated some of his research aims. Then, last but not least, he indicated an addition in the final paragraph. To the initial first sentence of this paragraph – ‘L’ethnologie n’est pas un jeu qui consisterait à remplir harmonieusement par des symboles, les vides de la documentation’ [Ethnology is not a game which would consist in filling up harmoniously with symbols the gaps in the documentation] – Leroi-Gourhan indicated the following addition: ‘L’Ethnologie (et l’Archéologie n’est que de l’Ethnologie au passé) suite: n’est pas un jeu …’ [Ethnology (and Archaeology is but Ethnology in the past tense). (followed by: is not a game…’. [Underline added].

These manually corrected sheets reached Protat Frères on 12 June 1946: with commendable efficiency they sent back to Leroi-Gourhan two days later another set of proofs, realising the changes he had envisioned. The last two paragraphs of p. 513 were separated by a double space (the only paragraphs of the whole book to be so spaced), and this p. 513 was duly followed by the newly created page 514 – its verso – into which have ‘fallen’ a total of 11 lines of printed text. Thus, the last sentence of the book, ‘Ainsi peut-on véritablement construire’, was followed by a satisfyingly solemn or closure-marking blank space occupying the final three-quarters of the page – a configuration reproduced in the printed and marketed version of the book, and thus fixed for posterity in the public domain (Figure 5 a, b).

3. On actors and theories…

Time now to capitalise on this archival foray. Brief and clinically-targeted as it has been, our excursion has still managed to touch on the processes of accretion, reshuffling and distillation prevalent within archival holdings – including those of André Leroi-Gourhan.¹³ Indeed we have from the onset touched on the potential of archives for securing a biographical appreciation of the actor(s) implicated in their creation. This in turn has underscored the contribution of archives-based biographical research to the historical and anthropological understandings of archaeology, as a research undertaking carried out within broader scientific, cultural and political milieus. Biographies of ‘great scholars’ (more often than not white, male and deceased) have long generated justified suspicion, tending as they occasionally do to veer into personal aggrandisement or to serve for academic model-setting or score-settling. More recent research, however, methodologically far more sensitive and explicitly reliant on the critical and contextual use of archival sources, has done much to revitalize and legitimate the genre, as notably exemplified by the work of Marc-Antoine Kaeser.¹⁴

To be sure, the history of archaeology writ-large is no longer designed simply to keep us steady on the route valiantly cleared up by our predecessors from unfounded superstitions to positive certainty. Among other implications, this hard-earned maturity calls for a reassessment of the relations between the historical and the theoretical developments of the discipline. Leroi-Gourhan’s last-page addition to his 1946 publication, whereby ‘Archaeology is but Ethnology in the past tense’, is in several respects quite a radical claim. Just as we need to historicise its background and its motivations, so we could further seek to theorise it – all the more so that, as suggested earlier, it sheds light on its author’s gradual disciplinary shift from ethnology towards prehistoric archaeology. Coined adventitiously and almost on the spur of the moment as it may have been (see more below), never to recur in this formulation in his writings, this bold slogan did not for that come out of the blue – and nor did it simply reprise evolutionist-cum-colonialist ideological equivalences between ‘primitive’ and ‘prehistoric’ peoples. From the 1920s onwards, the Musée de l’Homme in-the-making had been preaching, and to a large extent also practicing, a ‘totalizing’ ethnology (physical anthropology, ethnography, prehistoric archaeology) predicated on the ‘fundamental unity’ of humankind in all its dimensions – a view that, so far as Mauss and Rivet were concerned, was motivated by a strong moral opposition to the fascist and racist ideologies then in ascendency.¹⁸ To this ‘fundamental unity’ was however attached or even entailed an epistemological correlate regarding the indivisibly of humankind in space and in time as well: ‘Given that prehistory differs from ethnography only by the antiquity of the population studied’, so Rivet had offhandedly put it, ‘it is evident that the study of prehistoric collections can only be done by following the same methods applied to ethnographic ones’.¹⁹

These arguments, just like Leroi-Gourhan’s endorsement, are intelligible within the research traditions and institutional strategies in which they emerged. It is worth noting that the equation of archaeology with ethnology proposed here by Leroi-Gourhan antedates his familiarity with archaeological research (of the non-museological or art-historical kind): consequently, this claim has only remote links with the ‘ethnologie préhistorique’ or ‘paléoethnologie’ approach which he promoted with his collaborators at the Magdalenian open-site of Pincevent from the mid-1960s onwards.²⁰ It remains that the discussion of these claims today, in their context, can resonate with and even contribute to contemporary theoretical or epistemological debates. Indeed, both the feasibility and the utility of such past-present connections has been well demonstrated in recent history-of-archaeology scholarship, for example by Tim Murray or by Oscar Moro Abadía.²¹ To resort for a moment to the 1960s vocabulary of sociologist Robert Merton, I would argue that Leroi-Gourhan’s claim should really be made sense of within archaeological ‘systematics’, rather than being confined to its mere ‘history’. After all, when duly repositioned with regards to ethnoarchaeology and ethnohistory, for example, and when further enriched with considerations emanating from post-colonial discourse, ontology, symmetry, interdisciplinarity, temporal scales and material culture studies as well, is not the proposition whereby ‘Archaeology is but Ethnology in the past tense’ particularly good to think with? Could it not, at the very least, serve as a fitting end-of-course essay assignment for undergraduate class ‘Archaeological Theory 101’?

There is indeed ample scope for further theorising with Leroi-Gourhan’s historically situated claim, and then expand upon this on the changing relationships between theory, ideology and practice in museums and university settings in the twentieth-century Francophone and Anglophone research traditions. Keeping however these strands of investigations for other occasions, I propose to round up our archival incursion by recalling that it was as a last-gasp quasi-claustrophobic space-filling addition of ‘a few lines of composition’ at the printers’ behest, that our protagonist’s bold claim came to be written. What we need therefore to appreciate in some detail are the intellectual technologies – or, if you will, the material constrains and opportunities – entailed and intertwined with Leroi-Gourhan’s processes of knowledge production.

4. …and the material processes of knowledge production

To start with, most obviously, the proof-sets in question have come to our attention thanks to a long-established and meticulously maintained archival chain. Kept at hand for several decades by their ‘producer’ Leroi-Gourhan (for reasons that could well be further delved into), these ‘papers’, to give them their generic name, were then physically collected, deposited, accessioned, inventoried, sorted and curated in archival holdings as per conservation norms, and then made available to be retrieved from their storage boxes and folders, maintained in their accession and shelf-mark sequences and also handled as cautiously as possible so as to keep their friable edges from crumbling any further than they inevitably would.

More to the point, the ‘papers’ in question highlight to the full the technicity of the printing process, a key feature of book-making since Gutenberg’s times and until fairly recently, when mechanization took command in the age of mechanical (and dematerialised) reproduction (to equally paraphrase S. Giedion and W. Benjamin). In this respect, Leroi-Gourhan’s proof-reading work reported above, and specifically the double sets of corrections he requested to be made on his p. 513, can help us foreground and appreciate various aspects of book production that have become nowadays rather occluded or difficult to grasp. In particular, we can now access something of the material and intellectual practices by which knowledge was couched on paper – before, that is, the absolute domination of the computer-aided text-editing and printing processes which we nowadays totally take for granted.

Casting another look at the typographic errors indicated by Leroi-Gourhan, they appear to evidence a manual process of typesetting and text composition. Indeed misaligned or inverted characters such as those found on p. 513 (e.g. ‘qu _i’, ‘documəntation’, ‘spéɔulation’) occur in principle only when the metal-cast letters by which the type-setter composes the text are each individually placed (and infrequently misplaced) one after the other on the line-composing stick. This observation, if further confirmed, leads us to intriguing questions of technological innovations and economic history. After all, why was this highly-skilled and labour intensive manual procedure used here – fastidious and (occasionally) error-prone as it is – when far more efficient and reliable processes of line- or galley- composition by automated ‘hot metal’ casting machines (Monotype or Linotype) have already been available for several decades? It will not do in his respect to attribute some amateurism or provincialism to the printing outfit. Since their establishment in 1849 and until their liquidation in 1981, the Mâcon-based Protat Frères were among the foremost family-owned printing firms in France. During the 1870s, partly for ideological (anti-Republican) reasons, they shifted their printing activities from administrative or semi-official material to the erudition sector, specialising in delicate and time-consuming bibliophile publications, notably using Hieroglyphic, Assyrian, Hebrew, Arabic and other complex scripts. These skilled services were much in demand throughout the first half of the twentieth century by archaeological, numismatics and fine arts learned societies, as well as the Bibliothèque nationale, the École nationale des chartes and indeed the faculty of letters of the University of Paris, which published the monograph series of the Institut d’ethnologie.

Given these higher-end printing skills, the manual type-setting of Leroi-Gourhan’s 1946 book (in Caslon serif font) becomes more understandable, even though this publication did not seem to have necessitated particular technical dispositions. Protat Frères might have thus been keen to uphold their standards and reputation for hand-crafted printing expertise, despite the fact that they had actually owned a Monotype composing machine since the 1920s, according to the information available in their archival fonds (at the departmental archives of Saône-et-Loire, under file 37J–838). The same archival source can moreover suggest some connections between the retention of manual typesetting and the war circumstances. Indeed it might have been important for the firm to maintain their highly specialised workforce in employment in these difficult times, all the more so that some of its ‘printing materials’ had apparently been requisitioned, for the period 1943–1946, by the Commissariat à la mobilisation des métaux non ferreux.²²

Whatever the specific techniques used in its production, Leroi-Gourhan’s book went on to be composed into galley blocks pages of cast-metal characters, inked and press-printed recto/verso on large sheets of papers. Once folded into the volume’s dimensions (a thrice folded in octavo 16 pages format), these individual quires were collated and placed in their sequence, stitched, glued, guillotine-trimmed and finally bound together under cardboard covers. It so happened that, in its printed version, the ultimate page of Leroi-Gourhan’s thesis fell on page number 513 – which was also the initial page of the thirty-third 16-pages quire of which the book was composed. Hence the indications in smaller font-size at the bottom of this page: ‘A. Leroi-Gourhan. – Archéologie du Pacifique Nord’, ‘33’. In line with the printers’ modus operandi, this labelling ensured the smooth assembly and binding process of the book, insofar as the name of the author, the title of the work, and the sequence number of the quire in hand were all displayed – though possibly not as unobtrusively as one might have wished (see again Figure 2).²³

We recall in effect that Leroi-Gourhan was somehow dissatisfied with the position of the last sentence of his text, placed at the very bottom of the page and furthermore abutting to these printers’ indications. The process of the proof-correcting not only tuned his eye to the formal, grammatical or typographic aspects of the page. It also led him, probably for the first time, to appreciate his own text, its appearance, its disposition, indeed its materiality, as a reader would. From this viewpoint, he aimed to spare his imagined reader the inevitable temptation, having reached the textually dense bottom of page 513, to turn the page in the (unfulfilled) expectation of further text on the verso. Hence the second set of ‘topographical’ corrections he requested from Protat Frères, precisely in order to fulfil this expectation of further text to be added. Without conjecturing any further on the matter, my suggestion is that Leroi-Gourhan ended up in these lines unwittingly inversing the ‘natural order’ of authorship: it was not some inner drive or creative impulse (‘having something to say’) that motivated his writing, as much as the need to have additional space occupied by ‘a few lines of composition’, including then the bold theoretical statement we have sufficiently commented on.

Overall, what clearly transpires from our archival foray is the materiality of the proofs of Leroi-Gourhan’s Archéologie du Pacifique-Nord. So much so that, in their physical properties and their tangible potentialities, these proofs play a part not simply in the fixation or the communication of scientific knowledge, but also in its very generation. Hovering in the background are infrastructural expectations on the reliability of the postal services (on the Paris-Mâcon axis) and also economic concerns over the expenditures incurred by each back-and-forth set of proofs, raising costs which had to be approved by the collection director, Paul Rivet (as indicated on top left in (Figure 3)). The same goes for the need to ‘drop the lines’ so as to ‘turn the page’. The thrust of the matter here is that Leroi-Gourhan’s textual additions on p. 513–514, however significant or otherwise they might be, came about for paratextual reasons. We know now how much attention he paid, as an author, to the external qualities of his text, to these ‘surrounding’ elements highlighted by literary criticism, including frontispieces, epigraphs, intertitles, notes, headings, and indeed page layouts with their so meaningful blank spaces.²⁴ Our brief illustrated excursus into these ‘intellectual technologies’ has thus converged with the investigations undertaken under the heading of ‘textual genetics’ (génétique textuelle), which sets to retrace the intellectual and material trajectory spanning from the hastily scribbled draft to the full-fledged publication.²⁵ Put otherwise, if I may draw together some particularly satisfying referential interconnections, the archival foray I have proposed here has effectively applied to Leroi-Gourhan’s own scholarly productions the chaîne opératoire approach for which he is so justly famous.