Déchelette’s contribution to Iron Age Studies : theory and practice
Je suggère dans cet article qu’on porte plus d’attention aux approches méthodologiques et théoriques de Joseph Déchelette de l’archéologie. Même s’il a pu être plus un homme de synthèse que d’innovation, il appartient à la période où se développait l’« histoire culturelle » appliquée à la préhistoire. Ses façons d’aborder les problèmes sont très typiques de cette période, par exemple dans ses tentatives de mettre au point un cadre chronologique précis où situer les changements culturels en usant du concept de périodes et de fossiles types, et également dans son usage de cartes de répartition pour localiser des groupes « raciaux » tels que les Celtes et les Ligures. Un problème majeur était d’établir un lien entre les données archéologiques et les classifications d’autres sujets, en particulier la linguistique, même s’il n’alla pas jusqu’aux positions extrêmes de son contemporain Gustav Kossinna dont les idées allaient former la base des développements racistes de la première moitié du xxe siècle.
Mots clés :histoire culturelle, cartes de répartition, racisme, techniques de fouilles, chronologie, archéologie, sociabilité savante, hommages et condoléances
PlanHaut de page
- 1 M.-S. Binétruy, De l'art roman à la préhistoire, des sociétés savantes à l'Institut, itinéraire de (...)
1The books and articles which have so far been published about Joseph Déchelette have mainly concerned themselves with his personal life and circumstances, and the development of his interest in archaeology in terms of his excavations, research projects and publications1. What has hardly been touched on is the origin of his theoretical ideas and methodology. Mainly this is because, like many of his contemporaries, he thought there was nothing to discuss, and it was not until the 1960s, or even later in France, that what were thought to be “commonsense” approaches to archaeology began to be questioned, with the development of new approaches and a new “paradigm”, the rejection of the so-called Culture-Historical paradigm in favour of what came to be called the “New Archaeology”. In fact Déchelette’s period of activity lies in some of the most formative of years of the Culture-Historical approach, while many of his contemporaries were still stuck in an Antiquarian approach with only an interest in the objects themselves rather than their physical context and interpretation, for instance in the pillaging of cemeteries in the Champagne with not even the name of the site recorded, let alone the grave associations.
- 2 J.R. Collis, The Celts: Origins, myths and inventions, Stroud, 2003 (second revised edition 2006).
2In my book on the historiography of the Celts2, I suggested that Déchelette was one of the key figures in the formulation of the archaeological definition of the Celts, one which was to dominate throughout the 20th century, bringing together ideas and approaches from a variety of sources: linguistic, historical, art historical as well as archaeological. He also drew on a range of international scholars, German, Czech, British, Irish, and Swiss, developing what was an entirely new synthesis. My analysis of his work was, and has remained, superficial, and I have not had time to pursue the full range of his writings and those of his contemporaries, let alone study his letters which may betray better where his ideas were coming from and how he developed them. In this article I merely wish of provide some signposts to avenues of research which may prove interesting to pursue, especially in his contributions of protohistoric archaeology.
3The first question one must ask is what Déchelette was trying to do with the archaeological data at his disposal. Clearly it was to write History, but what sort of History? As in other nation states, especially Britain, History in the 19th century was increasingly being written against an imperial and colonial background, as the major powers in Europe vied with each other to establish world empires. The emphasis was to document the rise of the nation state, from the primitive customs of the earliest documented ancestors, in France, the Gauls, with the gradual evolution to more complex and sophisticated forms of government, leading up to national unity, but accompanied by developments in art and literature.
- 3 A. Thierry, Histoire des Gaulois depuis les temps les plus reculés jusqu’à l’entière soumission de (...)
4In the period following the Napoleonic wars, the two most influential historians in France were the brothers Augustin and Amédée Thierry, and it is the latter’s Histoire des Gaulois depuis les temps les plus reculés jusqu’à l’entière soumission de la Gaule à la domination romaine which dominated the teaching of pre-Roman history in France, going through many editions after its first publication in 1828 until 18703. It was commonly used as a school book prize (e.g. the copy in the library at Mont Beuvray), and Camille Jullian recorded the impact of his prize copy in developing his interest in history. In it we see the rise of “nos ancêtres les Gaulois” and of Vercingetorix as the first French national hero, uniting the Gauls against the Roman invaders. Thierry considered the Gauls to be the first inhabitants of Gaul, arriving sometime early in the second millennium bc (he was still working on the short, biblical. chronology). The sources of Thierry’s reconstructed history was the written classical sources, and this largely remained the case for his academic successor, Camille Jullian, but by then the long chronology had come into use with the recognition of the length of prehistory documented by Archaeology and Geology, but the historical aims remained the same, e.g. the origins of the Gauls and their customs and development, and the Roman conquest; he also had the benefit of Napoleon III’s excavations and writings. In his introduction to Protohistoire, Déchelette likewise depends heavily, and discusses in detail, the classical sources, especially on the location of the Celts.
5Historical linguistics in the 19th century were dominated by the German school: Grimm, Bopp, Zeuss and others, with the development of the concept of the Indo-Germanic or Indo-European language group. At this level of abstraction, language was generally equated with race, and so the “Indo-Europeans” were equated with the “Aryans”, in contrast with Semitic languages represented by the Jews and the Arabs; to this was added the further factor of religion, with Christianity opposed to Judaism and Islam. Race was also used at a more refined level in contrasting the Teutonic or Germanic race with the Slavs and the Celts, classifications which were to be fundamental in European History up to the Second World War in the ideology of the Third Reich. At a yet further level of refinement, language became the marker of the Nation State–French, Spanish, English, etc.–often leading to the suppression of minority languages such as Galliego, Basque, Catalan, Provençal, Alsacienne, Breton, Welsh, Gaelic, Lappish, an attitude only relaxed in the 1960s and 1970s. The leading scholar in driving forward these ideas in the late 19th and early 20th century was Gustaf Kossinna, but I am unclear how far his thinking affected Déchelette, as his major papers were appearing at the end of Déchelette’s life, in the immediately pre-war period. As is clear from the way in which he volunteered for military service in the French army (despite his age) and in the report of his final words, he too was certainly swept along by the tide of nationalistic thinking which precipitated the outbreak of the war, despite the international nature of his academic work.
- 4 H. d’Arbois de Jubainville, Premiers Habitants de l’Europe d’après les Auteurs de l'Antiquité et le (...)
6In France the field of Celtic linguistics was dominated by Henri d’Arbois de Jubainville, who, after this retirement as archivist for the Department of the Yonne, became the first professor of Celtic Studies in the Sorbonne, and his book Premiers Habitants de l’Europe4 very much reflected the thinking at the time in correlating peoples with languages. This approach assumed that the populating of Europe, and their successive replacements, could be reconstructed from linguistics. Though Déchelette occasionally states his disagreement with his interpretations, these are largely over matters like dating rather than a rejection of the methodology. As I have argued elsewhere, Déchelette relied much more closely on d’Arbois de Jubainville than he did on the traditional histories typified by Jullian.
7In history and linguistics Déchelette was following well established disciplines, but this was less true for archaeology. Some of the models he was using were less than a decade old, and, while it is clear that he was following these new ideas, the way in which he used them was innovative.
- 5 G. de Mortillet, “Les Gaulois de Marzabotto dans l’Apennin”, Revue Archéologique 22 (1870-1), p. 28 (...)
8His first problem was how to translate historical “racial” groups like Celts and Germans and linguistic entities into archaeological terms. In the latter half of the 19thcentury there was in an increasing belief that material culture could be used to define linguistic and ethnic groups, for instance Kemble’s distinctions between Celtic Art and that of the Romans, the Anglo-Saxons and the Vikings, or Virchow’s identification of Burgwallkeramik with the Slavic expansion in the later 1st millennium ad. For the Iron Age, in 1871 Désor and Mortillet recognised the similarity between the ornaments and weapons found in the graves at Marzabotto, and those found in similar graves in the north of France as well as from the lake site at La Tène, and de Mortillet used them as evidence for the Gallic invasions of northern Italy described by Livy and Polybius, and specifically with the tribe of the Senones5. In Palaeolithic archaeology different artefact types were used to typify specific stages of development and named after type sites, such as the Abbevillian, the Mousterian and the Magdalenian. This had also been attempted for later periods, including the Iron Age, with Hildebrand’s stylistic distinction between Hallstatt and La Tène, though de Mortillet’s attempt at making finer divisions such as the Marnian and the Beuvraysian met with less acceptance.
- 6 G. Kossinna, “Zur Herkunft der Germanen. Zur Methode der Siedlungsarchäologie”, Mannus-Bibliothek 6 (...)
9Kossinna’s famous formulation of the concept of the “Culture Group”, and its correlation with a “people” did not appear until 19116, and probably had little or no impact on Déchelette’s thinking. Certainly by 1904 he was using the equivalent term in French “civilisation”, but he was more ambiguous in its meaning than Kossinna. La Civilisation d’Hallstatt he thought was shared by several different peoples, especially Celtic and Illyrian, while for La Tène he distinguished between Celtic, German and Insular versions. Thus it is not always clear how he understood the concept; perhaps as a heuristic device useful to archaeologists, but not necessarily correlating with entities in other disciplines (e.g. a language group) though he does sometimes use it to signify a “people”, in places equating La Tène with Celtic. I hesitate to use the term “ethnic” as this was a concept which had yet to be defined, and a “Culture” or “Civilisation” had more of a racial significance than ethnic. Certainly when comparing the material culture from the Late La Tène oppida of Mont Beuvray, Manching, Stradonice and Velem St. Vid, he was struck by the marked similarity which he assigned to their shared Celtic culture. On the other hand he did not use “Celtic Art” as a clear discriminant as it turned up in his Celtic and Insular cultural areas, but not in the Celtic areas of Spain.
- 7 H. d’Arbois de Jubainville, “Conquête par les Gaulois de la région située entre le Rhin et l’Atlant (...)
- 8 J. Déchelette, Manuel d’Archéologie Préhistorique, Celtique et Gallo-Romaine. II-1: Âge du Bronze, (...)
- 9 J. Filip, Keltové ve Strˇední Evrop, Prague, 1956; Keltská Civilizace a její Deˇdictv, (revised edi (...)
10His solution was to use burial rites as the distinguishing criterion. D’Arbois de Jubainville had suggested that the origin of the Celts lay east of the Rhine, especially around the Main valley on the basis that here natural features such as rivers had names which were of Celtic origin, unlike France where rivers such as the Seine/Sequana for him represented an older, pre-Celtic, layer of toponymy7. He had suggested on the evidence of historical evidence in the classical writers that the expansion of the Celts into France lay around the 7th-6th centuries, a period in Hallstatt I and II and early La Tène I when inhumation was the dominant burial rite in this area. At the same time, in the Germanic areas, cremation was the main burial rite, so Déchelette suggested that inhumation was Celtic and cremation was Germanic, with crouched inhumation representing the pre-Celtic Ligurians8. The shift in northern France from inhumation to cremation he linked with Caesar’s statement that the Belgae were originally Celtic, but had absorbed a large Germanic population coming from east of the Rhine. This approach was perhaps an innovation by Déchelette, though both on theoretical and factual grounds it is now unacceptable (e.g. the areas assigned by Caesar to the Celts show a similar shift to cremation in the Late La Tène, for instance at Mont Beuvray). But this is something which has survived a long time in the literature, for instance the use by Filip of the appearance of flat inhumation cemeteries in central Europe to mark the arrival of the Celts in the “Dux Horizon” in the 4th century bc9; or the continued dating of the arrival of the Belgae in northern France to the 4th century BC (Caesar actually gives no dates) by authors such as Brunaux to explain the establishment of sanctuaries such as Ribemont-sur-Ancre and Gournay-sur-Aronde. In fact the shift to cremation starts as early as the 5th century in some areas of northern France, and inhumation is not particularly characteristic of the areas assigned to the Celts in the latter half of the 1st millennium. None of the supposedly distinctive features assigned to the Belgae, such as the sanctuaries or the Fécamp ramparts with earthen dump ramparts are confined to Belgic territory. In terms of the Celts, Déchelette’s ideas have had an even greater longevity, with the belief that the Celts originated in southern Germany, but maps which supposedly show the “origin of the La Tène Culture” or of “Celtic Art” are rather the distribution of the flat inhumation cemeteries in the 6th-5th centuries bc.
- 10 S. Winchester, The Map that changed the World, London, 2002.
11One of the major achievements of prehistoric archaeologists at the beginning of the 20th century in the short period before the war was the construction of chronological systems which have remained in use a hundred years later, including those devised by Reinecke and Déchelette for the central and western European Iron Age. The methods used by both authors lie in 19th century developments in Geology, notably in Palaeontology, where certain “type fossils” of extinct animals were used to typify and correlate rock formations; this approach was quickly transferred to Palaeolithic Archaeology in the 1860s. In the case of Geology the succession of rock formations was recognised by people like William “Strata” Smith as early as the 1820s10, and for Archaeology it was the sequence of river terraces and especially the succession of layers in caves. But the schemes, using the type fossil approach, were then extended to situations where the stratigraphical relationships could not be demonstrated.
12In the late 19th century the approach was further extended to the dating and classification of later phases of prehistory, where there was at the time only limited stratigraphical evidence, and the main evidence for contemporaneity came from the association of objects in graves or hoards. In these methodological approaches the lead had been largely taken by Scandinavian archaeologists: Christian Thomsen for the “Three Age System”; Johannes Worsaae for the developments of type fossils and associations; Hans Hildebrand for the division of the Iron Age into the Hallstatt and La Tène periods; and Oskar Montelius for typology and cross dating (the association of traded or similar objects which could be dated in other parts of the world from historical records, for instance in Egypt, Greece or Rome). The type fossil remained the key element in both Reinecke’s and Déchelette’s chronologies and were used to construct “phases”, but they often had to deal with unstratified material from sites which were occupied for a relatively short duration, as in the case of the oppida. Cross-dating was especially important to Déchelette in looking for the date and area of origin of “Celtic Art”, which for the first time he was able to date to the 5th century bc in the area from northern France, southern Germany and Bohemia.
- 11 H.-J. Eggers, Einführung in die Vorgeschichte, Munich, 1959.
- 12 J. Déchelette, Manuel d’Archéologie Préhistorique, Celtique et Gallo-Romaine. II-1 : Âge du Bronze(...)
- 13 J. Déchelette, Manuel d’Archéologie Préhistorique, Celtique et Gallo-Romaine. II-2 : Deuxième Age d (...)
- 14 J. Déchelette, Manuel d’Archéologie Préhistorique, Celtique et Gallo-Romaine. II-3 : Second Age du (...)
13The shift to the Culture-Historical paradigm in the late 19th and early 20th centuries thus produced a need for a closer dating of archaeological finds, in order to date culture changes which were then interpreted in terms of migrations and the arrival of new peoples, most clearly demonstrated in the maps used by Kossinna and his followers as evidence for the gradual expansion of the Germans from their supposed Scandinavian homeland11. Already in 1910 Déchelette was using the distribution of finds of bronze sickles as an indicator of the areas settled by the Ligurians, and noting how they were concentrated in the valley of the Rhône. This he did, not with a distribution map, but with a table listing the number of sickles found in each departement12 and in 1913 he discusses the distribution of flat inhumation burial rites in the Hallstatt period13. In 1914 he uses a distribution map of La Tène cemeteries in France14.
- 15 A. Bertrand, Archéologie Celtique et Gauloise : mémoires et documents relatifs au premiers temps de (...)
- 16 A. Bertrand and S. Reinach, Les Celtes dans les Vallées du Pô et du Danube, Paris, 1894.
14He was not, however, the first in France to use maps to put forward ethnic interpretations of archaeological data. Alexandre Bertrand in 1889 had produced a map showing the distribution of archaeological remains which he interpreted as three successive phases of colonisation of France15: the earliest marked by megalithic tombs constructed by an unrecorded race of people; a second, invasions by Celts in the Bronze Age represented especially by hoards of bronze objects; and a third in the Iron Age with cemeteries which were evidence of the arrival of the Gauls. The difference in the density of finds he interpreted as indications of the differing longevity of these groups in different areas; the concentration of megalithic tombs in Brittany showed for him that this primary wave of settlers had survived for many years on the Atlantic coast, the latest being contemporary with the arrival of the Gauls in the Iron Age. In his later volume, co-authored with Salomon Reinach (1894), he listed the occurrence of Hallstatt swords and Gallic cemeteries16.
- 17 H.-J. Eggers, Einführung in die Vorgeschichte, Munich, 1959, abb. 29.
- 18 H.-J. Eggers, Einführung in die Vorgeschichte, abb. 26.
- 19 J.R. Collis, The Celts, 2003 (second revised edition 2006).
15It was not until the 1950s that a proper critique was made of Kossinna’s methodologies, especially the use he made of distribution maps. Hans Jürgen Eggers considered how the process of discovery could affect distribution maps, and also how the material of which an object was made had also to be taken into account17. He also looked at the way in which cultural factors could affect distributions, especially the nature of deposition. So, for instance, the distribution of Roman objects in the southeast parts of the Baltic coast led to mainly coins being deposited in one area where they occurred in graves, whereas in an adjacent area finds of imported vessels were dominant, deposited in hoards18. However, Eggers still considered that cultural reconstruction was the primary aim of prehistoric archaeology. It was not until the 1990s that a similar critical approach was applied to the archaeology of the Celts19.
- 20 J.G. Bulliot, F. Thiollier and N. Thiollier, Album, St. Étienne, 1899.
16Jacques Gabriel Bulliot’s excavation at Mont Beuvray, which Déchelette took over from his uncle in 1897, had, in certain respects, been innovative; Bulliot, for instance, was one of the first, perhaps the first, to recognise timber buildings in the form of post-holes. He also ensured that the walls and other features he discovered were plotted on to a master plan, itself innovative in being the first ancient site for which a contour plan was made. Déchelette himself introduced one other innovation, the photographing of features discovered and of the finds, the latter published in the Album of plates by Bulliot with Félix and Noël Thiollier20.
- 21 M. Bowden, Pitt Rivers: the life and archaeological work of Lieutenant Augustus Henry Lane Fox Pitt (...)
- 22 R.E.M. Wheeler, Archaeology from the Earth, Harmondsworth, 1954.
17But in other respects the excavations were typical of their time, with the digging of trenches by workmen in search of walls which would then be followed to provide a plan, very much like the contemporary excavations on the Roman town at Silchester. The recognition of chronological depth in the phasing of the houses was minimal. Although stratigraphy was recognised on Palaeolithic cave sites, it was not used on sites of a later date. There are exceptions, such as General Pitt Rivers who in 1875 recorded the ditch of the Iron Age hill-fort at Cissbury cutting through the Neolithic flint mines; he also recorded the sequence of infill of the ditch of the Neolithic Wor Barrow in 1893, but this had been dug in spits with the depth of finds recorded which could later be plotted on to a profile of the ditch21. But in no case were sites dug as a sequence of superimposed layers; that was not to appear in Britain until the 1920s and 1930s with excavators such as Sir Mortimer Wheeler22, and in France even later, in the 1970s.
18Déchelette was thus merely a typical excavator of his time rather than an innovator, and the complexity of the succession of superimposed buildings revealed by the recent excavations at Mont Beuvray was simply not recognised. The site, like Silchester, was virtually treated as a single period, but provided one of the most complete plans yet available of a first century BC oppidum. The interpretation of the plan was in terms of a Roman or medieval town, with public buildings labelled “forum”, etc., elite courtyard houses, and specialist artisan areas reminiscent of the quarters of medieval towns dominated by specific crafts or guilds.
19The picture that emerges from this discussion of Déchelette’s work is of someone who was not a major innovator like his Scandinavian and British predecessors and contemporaries, for instance, Christian Thomsen, Johannes Worsaae, Augustus Pitt Rivers or Oskar Montelius. Rather he was a synthesiser who brought together the best scientific approaches to produce major overviews of his time of the prehistoric, protohistoric and Roman periods. His writings on Celtic Art are very typical. The concept of this art he took over from British authors: John Kemble, Augustus Franks, Sir Arthur Evans and J. Romilly Allen, and he was the first continental author to write about it. At the time the construction of archaeological periods as a means of dating was in vogue with a number of authors such as Montelius and Reinecke, so he upset his own scheme for the Iron Age, and he was able to fix this in absolute terms by using cross-dating, especially for the La Tène I period with its classical Greek and Etruscan imports. These objects were associated with the earliest forms of Celtic Art, so he was able to date its appearance, and to plot where the earliest examples were to be found, in graves in northern France, southern Germany and Bohemia. Thus, though none of the concepts and methodologies were new, he was able to make a new and major contribution in understanding the origin of the art form.
20Dismissing Déchelette as someone who was not a major innovator, however, may not entirely do him justice. He was certainly a major player in establishing the methodologies of the Culture-Historical paradigm which was to dominate European Archaeology for much of the 20th century, as, for instance, his approach to the identification and study of the Celts. A more detailed inquiry into his work and its context than I have been able to offer here may substantially change our view of his status among his peers.
1 M.-S. Binétruy, De l'art roman à la préhistoire, des sociétés savantes à l'Institut, itinéraire de Joseph Déchelette, Lyon, 1994. F. Déchelette, Livre d’Or de Joseph Déchelette : centenaire 1862-1962, Roanne, 1962. E. and J. Gran-Aymerich, “Les grands archéologues : Joseph Déchelette”, Archéologia 185 (1983), p. 71-73.
2 J.R. Collis, The Celts: Origins, myths and inventions, Stroud, 2003 (second revised edition 2006).
3 A. Thierry, Histoire des Gaulois depuis les temps les plus reculés jusqu’à l’entière soumission de la Gaule à la domination romaine, Paris, 1828.
4 H. d’Arbois de Jubainville, Premiers Habitants de l’Europe d’après les Auteurs de l'Antiquité et les Recherches le plus Récentes de la Linguistique, Paris, 1877 (second edition 1889).
5 G. de Mortillet, “Les Gaulois de Marzabotto dans l’Apennin”, Revue Archéologique 22 (1870-1), p. 288-290, pl. 22.
6 G. Kossinna, “Zur Herkunft der Germanen. Zur Methode der Siedlungsarchäologie”, Mannus-Bibliothek 6 (1911).
7 H. d’Arbois de Jubainville, “Conquête par les Gaulois de la région située entre le Rhin et l’Atlantique”, Revue Celtique (1903), p. 162.
8 J. Déchelette, Manuel d’Archéologie Préhistorique, Celtique et Gallo-Romaine. II-1: Âge du Bronze, Paris, 1910, p. 13.
9 J. Filip, Keltové ve Strˇední Evrop, Prague, 1956; Keltská Civilizace a její Deˇdictv, (revised edition 1963, English translation Celtic Civilisation and its Heritage, 1962), Prague, 1960.
10 S. Winchester, The Map that changed the World, London, 2002.
11 H.-J. Eggers, Einführung in die Vorgeschichte, Munich, 1959.
12 J. Déchelette, Manuel d’Archéologie Préhistorique, Celtique et Gallo-Romaine. II-1 : Âge du Bronze, Paris, 1910, p. 14.
13 J. Déchelette, Manuel d’Archéologie Préhistorique, Celtique et Gallo-Romaine. II-2 : Deuxième Age du Fer ou Époque de Hallstatt, Paris, 1913.
14 J. Déchelette, Manuel d’Archéologie Préhistorique, Celtique et Gallo-Romaine. II-3 : Second Age du Fer ou Époque de La Tène, Paris, 1914.
15 A. Bertrand, Archéologie Celtique et Gauloise : mémoires et documents relatifs au premiers temps de notre histoire nationale, Paris, 1889.
16 A. Bertrand and S. Reinach, Les Celtes dans les Vallées du Pô et du Danube, Paris, 1894.
17 H.-J. Eggers, Einführung in die Vorgeschichte, Munich, 1959, abb. 29.
18 H.-J. Eggers, Einführung in die Vorgeschichte, abb. 26.
19 J.R. Collis, The Celts, 2003 (second revised edition 2006).
20 J.G. Bulliot, F. Thiollier and N. Thiollier, Album, St. Étienne, 1899.
21 M. Bowden, Pitt Rivers: the life and archaeological work of Lieutenant Augustus Henry Lane Fox Pitt Rivers, dcl, frs, fsa, Cambridge, 1991.
22 R.E.M. Wheeler, Archaeology from the Earth, Harmondsworth, 1954.Haut de page
Pour citer cet article
John Collis, « Déchelette’s contribution to Iron Age Studies : theory and practice », Anabases, 9 | 2009, 239-247.
John Collis, « Déchelette’s contribution to Iron Age Studies : theory and practice », Anabases [En ligne], 9 | 2009, mis en ligne le 01 mars 2012, consulté le 27 avril 2017. URL : http://anabases.revues.org/480 ; DOI : 10.4000/anabases.480Haut de page
© AnabasesHaut de page