Version classiqueVersion mobile

Batellerie gallo-romaine

 | 
Giulia Boetto
, 
Patrice Pomey
, 
André Tchernia

3 – Technological Transfer from the Mediterranean to the Northern Provinces

Transfert technique de la Méditerrannée aux Provinces du Nord

Ronald Bockius

Résumé

L’article concerne le problème de l’origine et des influences techniques des anciennes traditions de construction navale mises en évidence en pays celtique entre les Alpes et les Îles Britanniques. Après l’occupation romaine de ces régions, la plupart des bateaux et des navires ont été construits selon les techniques méditerranéennes alors que d’autres témoignent de caractéristiques techniques typiquement gallo-romaines considérées comme celtiques. Comme les premiers peuvent être identifiés à des navires militaires romains à rames et à des navires de commerce probablement destinés à l’approvisionnement de l’armée, il semble que les techniques méditerranéennes n’ont atteint les Provinces du Nord qu’à cette période seulement. De même pour les productions des charpentiers gallo-romains qui associent des techniques locales à des caractéristiques de type classique. En l’absence de témoignage de navires assemblés d’époque pré-romaine sur le territoire continental de la Celtique, on ne sait si cette région a été influencée, ou non, par la construction navale méditerranéenne avant l’occupation romaine.
À partir des données historiques et archéologiques qu’il a rassemblées, l’Auteur soutient l’hypothèse de l’existence de relations techniques pré-romaines entre les traditions de construction navale de la Méditerranée et de l’Europe du Centre et du Nord-Ouest. De plus, l’étude d’un groupe d’épaves gallo-romaines révèle des éléments et des détails de construction inhabituels aussi bien dans la construction navale méditerranéenne contemporaine que dans la construction navale des régions barbares.

Entrées d'index

Mots clés :

chaland, bateau cousu

Keywords :

barge, sewn ship

Texte intégral

  • 1 For constructional criteria and geographical grouping
    cf. Bockius 2004, p. 126-144, fig. 2-9. Anoth (...)

1The river Rhine especially at its lower section is known as an area rich of ship archaeological remains, not only of wrecks but also of secondary sources, as anchors, fittings and the like. Most of the material belongs to the Roman period (fig. 1). Different to pre-medieval ship finds from the hinterland and distant provinces with access to the Mediterranean as to groups in the Saône-Rhône-Region and from the Swiss Jura lakes with their predominant discoveries of barges, there is a multiplicity of types of inland craft closely distributed along the lower Rhine’s military zone (fig. 1). Roman period ship wrecks came up to light along the Dutch, Belgian and North-Western French coasts and at the opposite side of the English Channel in the river Thames, near the Severn inlet and once in Ireland. Not all of them are seagoing vessels. We are used to define most of the latter ship finds by their constructional characteristics as remnants of Romano-Celtic shipbuilding tradition (McGrail 1995; Bockius 2003, p. 192; 205-214, map 2); others show Mediterranean ship construction features (fig. 1). The same holds true not only for the Rhône-Saône-group but also for the Rhenish group of ancient inland boats and ships which reflects not only a diversity of vessel types and functions but apparently also strong technological influences from Mediterranean shipbuilding traditions1.

Fig. 1. Ship finds of the Roman Period found in Central and Northwest Europe.

Fig. 1. Ship finds of the Roman Period found in Central and Northwest Europe.

2The aim of this paper is either to provide a critical review mainly of the published material found north of the Alps and to discuss aspects of technological continuity, interaction and innovation against a background of cultural and political history. It concentrates on three major questions and problems respectively:

  1. for what reasons was one Roman vessel of central European origin built in Mediterranean fashion with mortise and tenon whilst the other was without?

  2. what do we know about the provenance of certain constructional features which look strange in the concert of Romano-Celtic ship finds?

    • 2 The topic has been discussed from the author in a paper published in German: Bockius 2007a.

    are there archaeological and historical clues which can shed a bit more light on the problem of continental Celtic plank-built boats and ships of the Pre-Roman Period2?

3Answers can be found especially by a look onto the northern periphery of the Roman world, i.e. the Roman frontier along the rivers Danube and Rhine, but also the Atlantic zone to the north-west.

4The examination may start with ship-archaeological relics belonging to a family of inland craft whose builders indisputably followed Mediterranean procedures of ship construction. In the year 1994 the Roemisch-Germanisches Zentralmuseum Mainz together with Bavarian state authorities excavated two wrecks at the Bavarian section of the Danube in Oberstimm, near Ingolstadt (Bockius 2002) (fig. 2). The site was located in front of a Roman fortification. As has been shown by the author previously, these vessels suited Roman military purposes. In a functional sense, they can be interpreted as the smallest class of ancient galleys, known to the ancient as moneres – in this individual case, approx. 16 m long light open boats (fig. 3). Driven by oar crews of 18 and 20 men respectively, remains proved the original presence of an auxiliary sailing rig. Rowers were seated at relatively short rooms whose distances between imply the usage of classical measuring systems. As both hulls were carvel built both from soft and hard wood on the basis of shell-first planking by mortise and tenon edge joining, the mostly oaken frames were fastened to the planking by tree-nails, their heads nicely made by machine, a turning lathe (fig. 4). At first hand, every feature observed point to Mediterranean shipbuilding, which according to the dendrochronology took place in the German Danube area at the beginning of the 2nd century AD.

Fig. 2. Oberstimm, Bavaria, wreck No. 2 during excavation (after Bockius 2002).

Fig. 2. Oberstimm, Bavaria, wreck No. 2 during excavation (after Bockius 2002).

Fig. 3. Oberstimm, Bavaria, wrecks Nos. 1 & 2. Excavation plan (after Bockius 2002).

Fig. 3. Oberstimm, Bavaria, wrecks Nos. 1 & 2. Excavation plan (after Bockius 2002).

Fig. 4. Oberstimm, Bavaria, wreck No. 2. Treenail used to fasten a rib to planking, turned on a lath (photo R. Bockius).

Fig. 4. Oberstimm, Bavaria, wreck No. 2. Treenail used to fasten a rib to planking, turned on a lath (photo R. Bockius).

5There is a parallel to the Oberstimm boats, a wreck excavated in the late 19th century near the Roman military camp of Bunnik-Vechten (ancient Fectio), a Dutch site in the vicinity of Utrecht (Bockius 2002, p. 105-118, fig. 23-25, pl. 7-14). The publication of the year 1895 may not present the modern standards one would prefer but the documentation is clear enough to feel sure that the Vechten wreck belongs to the same ship type and the same shipbuilding tradition: edge-joining, oar and sailing propulsion, small-spaced crew accommodation, even the ground-plan of the boat as far as to be retrieved from the published drawings and archive material seems to match the Oberstimm findings. Two aspects may confuse: the Vechten wreck was found much further to the West at the Rhine, and it is apparently much older than the Oberstimm boats, presumably Tiberian, if not of an earlier date.

  • 3 Discussed within the framework of Rhenish finds of Roman barges by Bockius 2004, p. 138-144, note 3 (...)

6Nevertheless, from the discussed Dutch and Bavarian discoveries can be learnt that, in the first two centuries AD Mediterranean shipbuilding tradition reached the continental military zone of the Roman Empire along the Northern boundary, and locally built vessels also in a functional sense were connected to the Roman army3. Does that mean that all the other ship finds in the northern provinces also belong into a military context as long as they share elements of Mediterranean shipbuilding?

7We can not solve this problem at this point. However there are arguments which tend to a positive answer of the above question. The majority of traces of Mediterranean shipbuilding to be identified along the Roman provincial boarders are found in places of a certain importance which are linked with Roman military and administration. The artefacts to be listed here are by no means all wrecks, but also ship equipment and in one case even a dislocated tenon. It may be enough to mention the sites and to give either a short comment:

  • (A) London with the County Hall ship built from oak wood but with lots of mortise and tenon joints (Marsden 1994, p. 109-129, fig. 105-106);

  • (B) the Roman camp Nigrum Pullum/Zwammerdam at the Rhine limes in the Dutch province of South Holland where planking remains and a steering oar have been uncovered both showing edge-joining, but also by barge No. 6 with its single mortise and tenon fastenings (De Weerd 1988, p. 155-170, fig. 83; 89-90; 94-97; 105-106);

  • (C) the remains of a late 1st century Roman barge has been discovered in Vleuten-De Meern near Utrecht. De Meern 4 sank together with cargo, building material, which indicates an official purchase, as the location could because the wreck was uncovered in front of the limes track which followed the river boundary. Planking was fastened with pegged mortise and tenon joints (de Groot, Morel 2007, p. 38; 41-46, fig. 5.7; 5.9-10) (fig. 5).

Fig. 5. De Meern, prov. Utrecht, The Netherlands, Roman barge No. 4. Sectional views (after de Groot, Morel 2007).

Fig. 5. De Meern, prov. Utrecht, The Netherlands, Roman barge No. 4. Sectional views (after de Groot, Morel 2007).

8Different to the Vechten and Oberstimm boats, the typological classification of the above-mentioned ship finds tells us nothing about the assumed military background. Other traces of presumable Mediterranean shipbuilding tradition found in the area prove anything but solid: a supposed ship timber with lead sheeting discovered at the Roman naval base Velsen (pers. comm. J. Morel, Amersfoort. Cf. Bockius 2003, p. 196, note 17), located at the mouth of the contemporary river Rhine, an outlet of Lake Flevo. And, finally, the find of a displaced tenon may be noted. The artefact (fig. 6) came up to light during pebble dredging close to the colonia Ulpia Traiana, the modern town of Xanten which had been an important place on the Roman lower Rhine with a legionary camp nearby.

Fig. 6. Xanten, Nordrhein-Westfalen, Germany. Wooden tenon.

Fig. 6. Xanten, Nordrhein-Westfalen, Germany. Wooden tenon.

9It would not be surprising if, by the Roman occupation, Mediterranean shipbuilders or at least their skills reached the military zones, as had happened with lots of other conveniences and knowledge, e.g. pottery production, Latin language etc. If so, we simply may continue to distinguish Romano-Celtic and non-Romano-Celtic relics uncritically with the risk to be misled. As long as we are dealing with Roman ship finds, the simple separation in classical Roman and provincial Gallo-Roman tradition seems to be satisfying. However, one could ask: what do we know about the origin of Romano-Celtic shipbuilding, the definition of which so nicely matches archaeological and historical sources as well from the latter descriptions of Celtic ships and building procedures given by Caesar and Strabo? What can be traced back for pre-Roman Celtic shipbuilding in the strict sense?

10It is Béat Arnold who has best separated the continental Bronze and Iron Age features which survived prehistory (Arnold 1992, p. 57-105). He pointed to aspects as the continuity of dugout tradition in the wider sense, not to forget details as moss caulking and correlated constructional procedures that can be paralleled in north-western Europe at least to the Early Bronze Age. To avoid misunderstanding: the author of this paper does not wish to contradict him, nevertheless, from our point of view it is only half of the truth what we can learn.

  • 4 Bockius 2006, p. 195-197. In a wider range: Bockius 2009. The phenomenon is also known from Roman b (...)

11Few scholars concede that the scope of what is called Romano-Celtic shipbuilding tradition could include technical influences from the Mediterranean, caused by the evolution of Roman culture in the provinces during the centuries of the Romans physical and spiritual presence. Perhaps the best examples are the Mainz boats and the Guernsey ship; they reflect Romano-Celtic tradition but they also prove Mediterranean mentality by the use of measuring units calibrated on the basis of late Attic-Roman foot or provincial pes Drusianus ranges4. One could argue that this was caused by late-Roman influence as could be assumed for the custom of concealing coins in the mast-step of a ship or at secret locations (fig. 7) as is revealed in the Romano-Celtic London-Blackfriars wreck and Mainz, wreck No. 1 where a Theodosius I coin had been deposited presumably as a votive between frame and planking. Such clues may be understood as outstanding accompanying phenomena of Romanisation (Bockius 2006, p. 11; 191-192 map 1, pl. 14.5). But what is about technical features of Romano-Celtic vessels which in principle point to Mediterranean origin although they had become obsolete and were out of fashion within contemporary Mediterranean shipbuilding since hundreds of years?

Fig. 7. Distribution of ship finds with coin ex-voto (after Bockius 2006).

Fig. 7. Distribution of ship finds with coin ex-voto (after Bockius 2006).
  • 5 Jézégou et al. 2009. The author is especially indebted to M.-P. Jézégou for giving him the opportun (...)

12The topic will be underlined with just a few issues. Starting with the most problematic comparison: the mast-step cut into a massive floor timber ought to be discussed. This characteristic element is typical for Romano-Celtic boats and ships, found with inland and coastal vessels, round-built merchantmen, sharp-ending oared boats, and barges as well. Two varieties exist: floor timbers with squared socket holes to insert the mast foot (fig. 8), and such with a construction made from iron hoops and a cross-bar to open the mast-step abaft in order to slip the mast in a leaning position during the rigging manoeuvre, e.g. when lowering the mast when a ship had to pass bridges. The archaeological evidences are dated from the end of the 1st century AD (Xanten-Wardt) to AD 390 (Mainz, wreck No. 5). Distributed within the area of Romano-Celtic shipbuilding on the Thames, the Channel region, along the Rhine to the Lac de Neuchâtel, the most southerly find comes from Lyon, Parc Saint-Georges, wreck No. 4 (Bockius 2007b, fig. p. 34) (fig. 9). In the maritime environment of Mediterranean shipbuilding practise mast-step frames are missing. Nevertheless, as can be traced back from a shipwreck found at Mèze (Conque des Salins, étang de Thau, dép. Hérault), the principle could have been known to shipwrights in the area. This wreck, by its edge-joined planking and bipartite transversal with a mast-step in the centre, meshes together features of Mediterranean ship construction and of Romano-Celtic tradition (Rouquette et al. 2004, p. 35-38, fig. 22-26; and M.-P. Jézégou in this volume). Unfortunately, the age of the wreck seems by no means clear. In a first attempt, a wood sample from the keel has been radiocarbon dated. It resulted in a range of 280 BC to AD 70 with a maximum probability within 210 to 85 BC5. Doubts caused by stratigraphic observations led to a second analysis with calibrated dating to AD 2 to AD 236 (Jézégou et al. 2009, p. 78). In the former case, this ship find would in a timely fashion anticipate the Gallo-Roman mast frame, whilst in the latter case it would indicate the inspiration of shipbuilders in the south of France, who were familiar with mortise and tenon technology, by those of the Romano-Celtic sphere where mast frames are typical.

Fig. 8. Bevaix, Suisse. Romano-Celtic mast frame “socket hole” type (after Arnold 1992).

Fig. 8. Bevaix, Suisse. Romano-Celtic mast frame “socket hole” type (after Arnold 1992).

Fig. 9. Distribution of Romano-Celtic mast steps. Dots: mast frames; triangles: longitudinally assembled log.

Fig. 9. Distribution of Romano-Celtic mast steps. Dots: mast frames; triangles: longitudinally assembled log.

13Should the Pre-Roman, Hellenistic dating of the Conque des Salins wreck hold true, the Romano-Celtic principle to use strong ribs as mast foundations would had been familiar to western Mediterranean shipbuilders before the oldest remnants of Romano-Celtic watercraft occurred in Central and North-Western Europe during the 1st century AD.

14Even before the time of Caesar’s conquest, the northern part of ancient Gaul was influenced by Greek culture, adopted from the French Mediterranean coast via its hinterland. Greek and Etruscan bronze vessels of the 6th and 5th centuries B.C. found around Bourges and at the lower river Loire may be noted here (Kimmig 1983, p. 3-78, fig. 21, 23, 25, 27-32. Cf. also Cunliffe 2001, p. 311-328, fig. 8.1-2, 5-6. – To Bourge: Gran-Aymerich 1997), but also Polybios’s and Strabo’s notices on the Emporion Korbilon should be kept in mind. This place was probably located near St. Nazaire, in the vicinity of the Celtic Veneti (whose ships according to Caesar’s description would match Romano-Celtic ship construction), and was still flourishing at the end of the 2nd century BC (Bockius 2003, p. 227-232, notes 151-152, map 6). One can imagine that Iron Age Celtic shipbuilding in Gaul as it has been portrayed by Caesar in the middle of the 1st century BC was influenced by Western Greek technological transfer. So, the occurrence of an early, perhaps even pre-Roman mast-step frame with a ship built according to Mediterranean traditions would not make surprising – of course, we must accept an amalgam of different boatbuilding traditions, Celtic and Greek, in Pre-Roman Gaul.

15Though initially nothing more than a hypothesis, the theory gains support from the analysis of certain ship finds. However, there are three observations further to the above-mentioned arguments, which suggest an archaic layer of shipbuilding skill not only in Romano-Celtic ship construction but also in more advanced-looking vessels built in the former Celtic area, and notably, they all belong to inland watercraft.

16As the first category, special types of keelsons may be discussed. Of lesser interest, the keelson type found with Oberstimm wreck 1 meets a comparable girder in one of the Ostia ship finds. However, the timber found with Oberstimm 2 (fig. 10) appears more instructive: the characteristic massive swelling at the section where the mast-step has been located is exhibited by parallels found with the Zwammerdam 2 and 6 barges and Woerden 7, all dated to the 2nd century AD (fig. 11). Similar shaped keelsons are known from Hellenistic wrecks, such as La Chrétienne A and Plane A, but the better parallels are found with Late Archaic and Classical Period ships of sewn construction, by name Bon-Porté I, Gela 1, and Ma’agan Mikhael; the Golo wreck apparently had a similar keelson. In contrast, comparably shaped timbers from the Mediterranean contemporary to the 2nd century AD wrecks from Germany and the Netherlands are still to be found. However, used as longitudinal re-enforcements and mast foundations of Romano-Celtic barges, the keelson type appears to be strange because from a ship architectural point of view in flat-bottomed barges stiffness is already guaranteed by their L-shaped bilge-strakes. No doubt keelsons found with Gallo-Roman barges only figured as an equivalent of the usual mast step frame typical for Romano-Celtic shipbuilding. So much seem to be clear, such elements of peculiar shape in the Mediterranean had been out of fashion for centuries, at least as components of seagoing ships.

Fig. 10. Oberstimm, Bavaria, wreck No. 2. Keelson with mast step and socket holes of thward stanchions. No scale (after Bockius 2002).

Fig. 10. Oberstimm, Bavaria, wreck No. 2. Keelson with mast step and socket holes of thward stanchions. No scale (after Bockius 2002).

Fig. 11. Zwammerdam, prov. South Holland, The Netherlands. Barge No. 2. Sectional view. No scale (after De Weerd 1988).

Fig. 11. Zwammerdam, prov. South Holland, The Netherlands. Barge No. 2. Sectional view. No scale (after De Weerd 1988).

17The second aspect leads us again to the family of ancient vessels of stitched construction: some of the Romano-Celtic barges (listed by Bockius 2000, p. 442; 445; 455, fig. 2; 8) including the De Meern 4 wreck of mortise and tenon construction (de Groot, Morel 2007, p. 37-41; 49, fig. 5.7-9A; 5.11, pl. fig. 5.11) show a unusual pattern of “limber holes”, which, on a regular basis, were cut into the lower side of the floor timbers exactly where plank seems and even scarfs were crossing underneath ribs (fig. 5 and 11). As a practical explanation of that phenomenon, Dutch colleagues have emphasised that the purpose of such recesses might have been to prevent ribs from rotting (J.-M. A. W. Morel in: Jansma, Morel 2007, p. 141-142). That explanation appears surprising because who has ever seen wooden ribs, either ancient or modern, that were rotted along the positions where seams crossed? Thus, it is hard to believe in the idea of always leaking plank seams, all the more it is well-known that timbers do not rot by being permanently wet, but rather by long term fluctuations of wet and dry conditions. As it seems floor-timbers of Roman barges without having recess systems correlated to plank seams show no traces of rotting. Were only those ships always watertight?

18However, the pattern of “limber holes” sequences recalls the typical scheme for frames of ancient ships of sewn planking and/or lashing, where recesses were needed for the luting material and stitching inside planking (“crenellated” frames: Kahanov 2004, p. 49-63; 66-77, Tables 9-11) (fig. 13.2). Thus, such a feature should be considered as a survival from a time when Celtic plank ships still were of sewn construction. That sewn shipbuilding existed in Iron Age Britain (synopsis by McGrail 1996, p. 24-38 with references; Clark 2004) only but not in ancient Central Europe is hard to believe. Rather it seems plausible to assume that the apparent non-existence of Celtic inland plank ships – with the exception of the barge from Lipe (Slovenia) (Bockius 2000, p. 465-468; 477, fig. 24, pl. IV 2; see also infra Boetto, Rousse) – reflects a much smaller number of such vessels built in pre-Roman times than were in use during the later period, whether in Gaul or in Britain.

Fig. 12. Mainz, Rheinland-Pfalz, Germany. Late Roman wreck No. 3. Excavation plan showing the framing system with segmented ribs. No scale (after Bockius 2006).

Fig. 12. Mainz, Rheinland-Pfalz, Germany. Late Roman wreck No. 3. Excavation plan showing the framing system with segmented ribs. No scale (after Bockius 2006).

Fig. 13. Segmented ribs connected by scarfs: 1. Mainz, Late Roman wreck No. 3, frame A3; 2. Marseille, wreck Jules-Verne No. 9, end 6th century BC (2 after Pomey 1997).

Fig. 13. Segmented ribs connected by scarfs: 1. Mainz, Late Roman wreck No. 3, frame A3; 2. Marseille, wreck Jules-Verne No. 9, end 6th century BC (2 after Pomey 1997).

19Perhaps the third observation is needed more to trust in the hypothesis of an obscure layer of Mediterranean influence to be found with Romano-Celtic plank ships: when Mainz wreck No. 3 was first examined, the author was much confused from its framing pattern as it seems to be unique in the concert of Romano-Celtic and Mediterranean-Roman ship construction. Ribs were made each from two and three segments of grown wood, which have been assembled to the planking alternately (fig. 12). Z-shaped scarfs were arranged either single near the plank keel or pair-wise at the bilges (fig. 13.1), fitted together by iron nails which also protruded inside the hull planking. Apart from joints sporadically found with frames on the London, County Hall wreck
(Marsden 1994, p. 117-118, fig. 105-106 [scarf joins frames no. 12, 16, 20, 28, 36, some probably being repairs]), such a system of two- and three-segmented scarfed frames looks strange within contemporary Roman ship building. Nevertheless, the same feature again is known to be a characteristic of late Archaic and Classical sewn shipbuilding (Kahanov 2004, p. 49-63, Table 9) (
fig. 13.2). Albeit some seven to eight centuries younger, the phenomenon shown by Mainz 3 may be interpreted as a late reflection of a much older tradition which survived into the Roman period by means of a construction skill as sewn shipbuilding likewise did in the Northern Adriatic Region (Bockius 2003, p. 192-193; 197-200, maps 2-3; Kahanov 2004, p. 66-77, fig. 75; Glusčević 2004) (fig. 14). Finally, it should not be forgotten that even the preference of Greek ship builders of the 6th to 4th centuries BC to use metal nails was shared by Gallo-Roman ship building.

Fig. 14. Distribution of Late Archaic and Classical Periods (triangles) to Roman Period (dots) ship finds in the Mediterranean of sewn construction, and wrecks showing related shipbuilding features (cross symbols).

Fig. 14. Distribution of Late Archaic and Classical Periods (triangles) to Roman Period (dots) ship finds in the Mediterranean of sewn construction, and wrecks showing related shipbuilding features (cross symbols).
  • 6 For Late Bronze and Iron Age disc wheels with un-pegged plug fasteners, similar to tenon joining cf(...)

20It appears reasonable to assume that Mediterranean technology as mortise and tenon construction had been transferred to the Rhine and Danube in course of the Roman occupation. The same could be true for the utilization of certain measuring systems (cf. above) for they were identified even with single ships of Romano-Celtic tradition (Mainz, St. Peter Port wrecks, and Rhenish barges). This may be understood as an effect of Romanization which was introduced to local Gallo-Roman shipyards e.g. by experts of the Roman army, who doubtless were most familiar with foot and cubit as units of every day life in the Classical World. Whether such a late, i.e. Roman influx holds true also for the selected phenomena described before seems less plausible, because one should bear in mind the long distance of time between Greek and Phoenician knowhow respectively and Gallo-Roman ship construction. From our point of view, the Celtic sphere had been influenced by western Mediterranean boat builders long before the employment of the Roman northern provinces. The Lough Lene wreck found in an Irish lake could be claimed to be an indication for that hypothesis, though its 14C-dating to 300 B.C. at the earliest and the 1st century AD at the latest is not that satisfying (Farrell 1989, p. 223-227, fig. 6 top; O hEailidhe 1992, p. 185-190, fig. 2; dating: Brindley, Lanting 1991). As an extended dugout that boat was made from an extremely flat hollowed bottom element which was lashed together from two pieces and mechanically secured by a dovetail joint. The boat’s sides originally were made each from a single plank fastened to lateral uprising components of the carved bottom segments by pegged mortise and tenon joints (fig. 15.1-2). According to its constructional features the Lough Lene extended logyboat combines different building mentalities (fig. 15.2): (1) lashing as a technique used in British shipbuilding since the Early Bronze Age, and dovetailing which even proves to be of Neolithic origin; (2) edge-joining by mortise and tenon as a fastening method supposed to be unknown to prehistoric Northern Europe. Thus, this ship find in principle has to be seen as a hybrid with mixed Northern European and Mediterranean construction features. On the other hand, the occurrence of mortise and tenon fasteners does not strictly point to a relation to shipbuilding in the Roman provinces or even to the Mediterranean world: the plug-shaped tenons of the Irish wreck look much different to fasteners found within vessels reflecting Mediterranean shipbuilding tradition, whether seagoing or inland craft. The only coincidence with contemporary material exists in their pegging whereas un-pegged mortise and tenon joinery in principle was known to Central European carpenters from the Late Bronze Age when bipartite disc wheels were stuck together6, and additionally fastened by dovetail joints (fig. 16 and 17). Nevertheless, the Irish boat reflects not only the philosophy of prehistoric British boat and wagon builders because it includes pegging as a procedure to fasten tenons transversally. It is hard to say at what period such a skill inspired Northern European handicraft and was then taken into practical boatbuilding, but it took place probably long before the Roman occupation of Britain when luxury goods moved from the area of oil trees and grapes into Celtica (e.g. Kimmig 1983; Cunliffe 1988, p. 13-58; 2001, p. 302-310; 331-336). The late pre-Roman Iron Age spoke wheel from Holm Pierrepont, dated to the 2nd century BC at the latest (Musty, MacCormick 1973, pl. LVa ; Stead 1979, p. 43, fig. 13) may give a clue as to when un-pegged mortise and tenon joints were introduced to British woodworking techniques before the Roman conquest. Its multi-piece felloe was fastened by thin tenons well comparable with joints known from ships.

Fig. 15. Lough Lene, Ireland, extended dugout with mortise and tenon joints.

Fig. 15. Lough Lene, Ireland, extended dugout with mortise and tenon joints.

1. Reconstructed cross-section;
2. Plan drawings (after O hEailidhe 1992).

Fig. 16. Distribution of multipartite wooden disc wheels showing dovetails and un-pegged mortise and tenon joinery.

Fig. 16. Distribution of multipartite wooden disc wheels showing dovetails and un-pegged mortise and tenon joinery.

Dots: Late Bronze Age finds; triangles: Pre-Roman Iron Age finds.

Fig. 17. Flag Fen, near Peterborough. Late Bronze Age wheel (after Taylor 2001).

Fig. 17. Flag Fen, near Peterborough. Late Bronze Age wheel (after Taylor 2001).

21However, it is a pity that the Lough Lene find can not provide a more accurate dating. To a better proof of the discussed observations, more Iron Age plank-built boats and ships had to be discovered not only on the British Isles but especially on the continent.

Bibliographie

Arnold 1992 : ARNOLD (B.) – Batellerie gallo-romaine sur le lac de Neuchâtel, tome 2, Saint-Blaise, Éditions du Ruau, 1992 (Archéologie Neuchâteloise, 13).

Bockius 2000 : BOCKIUS (R.) – Antike Prahme. Monumentale Zeugnisse keltisch-römischer Binnenschiffahrt aus der Zeit vom 2. Jh. v. Chr. bis ins 3. Jh. n. Chr. Jahrbuch Römisch-Germanisches Zentralmuseum, 47.2, 2000, p. 439-493.

Bockius 2002 : BOCKIUS (R.) – Die römerzeitlichen Schiffsfunde von Oberstimm in Bayern, Mainz, Verlag des Römisch-Germanischen Zentralmuseums-R. Habelt, 2002 (Monographien Römisch-Germanisches Zentralmuseum, 50).

Bockius 2003 : BOCKIUS (R.) – Abdichten, Beschichten, Kalfatern. Schiffsversiegelung und ihre Bedeutung als Indikator für Technologietransfers zwischen den antiken Schiffbautraditionen. Jahrbuch Römisch-Germanisches Zentralmuseum, 49, 2002, p. 189-234.

Bockius 2004 : BOCKIUS (R.) – Antike Prahme und ihre Rolle in der Binnenschiffahrt der gallisch-germanischen Provinzen. In : Brandt (K.), Kühn (H.J.) eds., Der Prahm aus dem Hafen von Haithabu (Wrack Haithabu IV). Beiträge zu antiken und mittelalterlichen Flachbodenschiffen, Neumünster, Wachholtz, 2004, p. 125-151 (Schriften des Archäologischen Landesmuseums, Ergänzungsreihe, 2).

Bockius 2006 : BOCKIUS (R.) – Die spätrömischen Schiffswracks aus Mainz. Schiffsarchäologisch-Technikgeschichtliche Untersuchung spätantiker Schiffsfunde vom nördlichen Oberrhein, Mainz, Verlag des Römisch-Germanischen Zentralmuseums-Schnell & Steiner, 2006 (Monographien Römisch-Germanisches Zentralmuseum, 67).

Bockius 2007a : BOCKIUS (R.) – Spuren griechisch-etruskischen Knowhows im keltischen Schiffbau? In : Kelzenberg (H.), Kießling (P.), Weber (S.) eds., Forschungen zur Vorgeschichte und Römerzeit im Rheinland, Festschrift H.-E. Joachim, Leiden, Zabern, 2007, p. 253-267 (Beiheft der Bonner Jahrbuecher des Landschaftsverbandes Rheinland, Rhenisches Landesmuseum Bonn und Verein von Altertumsfreunden im Rheinlande, 57).

Bockius 2007b : BOCKIUS (R.) – Forschungsschwerpunkt 4: Mediterrane Traditionen im Schiffbau in West-, Mittel- und Nordeuropa. In : Römisch-Germanisches Zentralmuseum. Forschungsinstitut für Vor- und Frühgeschichte, Jahresbericht 2006, Mainz, Verlag des Römisch-Germanischen Zentral Museums, 2007, p. 32-34.

Bockius 2009 : BOCKIUS (R.) – Markings and Pegs: Clues to Geometrical Procedures of Roman Naval Architecture? In : Nowacki (H.), Lefèvre (W.) eds., Creating Shapes in Civil and Naval Architecture. A Cross-Disciplinary Comparison, Leiden-Boston, Brill, 2009, p. 73-92 (History of Science and Medicine Library, 2).

Brindley, Lanting 1991 : BRINDLEY (A. L.), LANTING (J. N.) – A boat of the Mediterranean tradition in Ireland: preliminary note. International Journal of Nautical Archaeology, 20, 1991, p. 69-70.

Burmeister, Fansa 2004 : BURMEISTER (S.), FANSA (M.) eds., Rad und Wagen. Der Ursprung einer Innovation, Wagen im Vorderen Orient und Europa, Mainz, Zabern, 2004 (Beihefte Archäologische Mitteilungen Nordwestdeutschland, 40).

Clark 2004 : CLARK (P.) ed., The Dover Bronze Age Boat, Swindon, English Heritage, 2004.

Cunliffe 1988 : CUNLIFFE (B.) – Greeks Romans & Barbarians: spheres of interaction. London, Batsford, 1988.

Cunliffe 2001 : CUNLIFFE (B.) – Facing the Ocean. The Atlantic and its Peoples 8000 BC – AD 1500. Oxford, University Press, 2001.

Farrell 1989 : FARRELL (R.T.) – The Crannόg Archaeological Project (CAP), Republic of Ireland II: Lough Lene – offshore island survey. International Journal of Nautical Archaeology, 18.3, 1989, p. 221-228.

Glusčević 2004 : GLUSČEVIĆ (S.) – Hydroarchaeological excavations and the discovery of the third sewn Liburnian ship – seriliae, in the Roman harbour of Zaton near Zadar. Archaeologia Maritima Mediterranea, 1, 2004, p. 41-52.

Gran-Aymerich 1997 : GRAN-AYMERICH (J.) – Les premières importations méditerranéenes de Bourges. In : Brun (P.), Chaume (B.) dir., Vix et les éphémères principautés celtique. Les VIe-V e s. av. J.-C. en Europe centrale-occidentale. Actes du colloque, Châtillon-sur-Seine (1993), Paris, Éditions Errance, 1997, p. 201-211.

de Groot, Morel 2007 : GROOT DE (T.), MOREL (J.-M. A. W.) eds. - Het schip uit de Romeinse tijd De Meern 4 nabij boerderij de Balije, Leidsche Rijn, gemeente Utrecht: waardestellend onderzoek naar de kwaliteit van het schip en het conserverend vermogen van het bodemmilieu, Amersfoort, Rijksdienst voor archeologie, cultuurlandschap en monumenten, 2007 (Rapportage archeologische Monumentenzorg, 147)

Guyon, Rieth 2009 : GUYON (M.), RIETH (É.) – The Gallo-Roman wrecks from Lyon, Parc Saint Georges (France): new archaeological data on Ancient inland “bottom-based shipbuilding”. In : Bockius (R.) ed., Between the Seas. Transfer and Exchange in Nautical Technology, Proceedings of the 11th International Symposium on Boat and Ship Archaeology (ISBSA 11), Mainz (2006), Mainz, Verlag des Römisch-Germanishen Zentralmuseums, 2009, p. 157-165 (RGMZ-Tagungen, Band 3).

Jansma, Morel 2007 : JANSMA (E.), MOREL (J.-M. A. W.) eds. – Een Romeinse Rijnaak, gevonden in Utrecht-De Meern. Resultaten van het onderzoek naar de platbodem ‘De Meern 1’, Amersfoort, Rijksdienst voor archeologie, cultuurlandschap en monumenten, 2007 (Rapportage archeologische Monumentenzorg, 144).

Jézégou et al. 2009 : JÉZÉGOU (M.-P.), ROUQUETTE (D.), WICHA (S.) – The Roman wreck at Conque de Salins (Mèze, Étang de Thau, Département de l’Hérault). Skyllis. Zeitschrift für Unterwasserarchäologie, 9, 1, 2009, p. 78-84.

Kahanov 2004 : KAHANOV (Y.) – The sewing of the ship. In : Kahanov (Y.), Linder (E.), The Ma’agan Mikhael Ship. The Recovery of a 2400-Year-Old Merchantman: Final Report, vol. II. Haifa, Israel exploration Society-University of Haifa, 2004, p. 3-79.

Kimmig 1983 : KIMMIG (W.) – Die griechische Kolonisation im westlichen Mittelmeergebiet und ihre Wirkung auf die Landschaften des westlichen Mitteleuropa. Jahrbuch Römisch-Germanisches Zentralmuseum, 30, 1983, p. 3-78.

Kimmig 1992 : KIMMIG (W.) – Die „Wasserburg Buchau“ – eine spätbronzezeitliche Siedlung. Forschungsgeschichte-Kleinfunde. Stuttgart, Theiss, 1992 (Materialhefte zur Archäologie in Baden-Württemberg, 16).

Kunwald 1970 : KUNWALD (G.) – Der Moorfund im Rappendam auf Seeland. Prähistorische Zeitschrift, 45, 1970, p. 42-88.

Marsden 1994 : MARSDEN (P.) – Ships of the Port of London. First to eleventh centuries AD, London, English Heritage, 1994 (Archaeological Report, 3).

McGrail 1995 : McGRAIL (S.) – Romano-Celtic boats and ships: characteristic features. International Journal of Nautical Archaeology, 24, 1995, p. 139-145.

McGrail 1996 : McGRAIL (S.) – The Bronze Age in Northwest Europe. In : Gardiner (R.) ed., The earliest ships. The Evolution of Boats into Ships, London, Brassey’s, 1996, p. 24-38 (Conway’s History of the Ship).

Musty, Mac Cormick 1973 : MUSTY (J.), MAC CORMICK (A. G.) – An early Iron Age wheel from Holme Pierrepont, Notts. Antiquaries Journal, 53, 1973, p. 275-277.

O hEailidhe 1992 : O HEAILIDHE (P.) – ‘The Monk’s Boat’ – a Roman-period relic from Lough Lene, Co. Westmeath, Eire. International Journal of Nautical Archaeology, 21.3, 1992, p. 185-190.

Pomey 1997 : POMEY (P.) – Un exemple d’évolution des techniques de construction navale antique : de l’assemblage par ligatures à l’assemblage par tenons et mortaises. In : Garcia (D.), Meeks (D.) dir., Techniques et économie antiques et médiévales, “Le temps de l’innovation”, Colloque international, Aix-en-Provence (1996), Paris, Éditions Errance, 1997, p. 195-203.

Rouquette et al. 2004 : ROUQUETTE (D.), JÉZÉGOU (M.-P.), WICHA (S.) – Épave Conque des Salins. In : Bilan Scientifique du DRASSM 1999, Paris, Ministère de la Culture et de la Communication, 2004, p. 35-38.

Schovsbo 1987 : SCHOVSBO (P. O.) – Oldtidens vogne I Norden, Frederikshavn, Bangsbomuseet, 1987.

Stead 1979 : STEAD (I. M.) – The Arras Culture, York, Yorkshire Philosophical Society, 1979.

Taylor 2001 : TAYLOR (M.) – The Wood. In : Pryor (F.), The Flag Fen Basin. Archaeology and environment of a Fenland Landscape, London, English Heritage, 2001, p. 213-215.

van der Waals 1964 : WAALS VAN DER (J. D.) – Prehistoric disk wheels in The Netherlands, Groningen, Wolters, 1964.

de Weerd 1988 : WEERD DE (M.D.) – Schepen voor Zwammerdam. PhD thesis, University of Amsterdam, 1988.

Woytowitsch 1978 : WOYTOWITSCH (E.) – Die Wagen der Bronze- und frühen Eisenzeit in Italien, München, Beck, 1978 (Prähistorische Bronzefunde, Abt. XVII, vol. 1).

Notes

1 For constructional criteria and geographical grouping
cf. Bockius 2004, p. 126-144, fig. 2-9. Another approach to distinguish regional peculiarities of shipbuilding: Guyon, Rieth 2009 and the articles of É. Rieth and M. Guyon and É. Rieth in this volume.

2 The topic has been discussed from the author in a paper published in German: Bockius 2007a.

3 Discussed within the framework of Rhenish finds of Roman barges by Bockius 2004, p. 138-144, note 36.

4 Bockius 2006, p. 195-197. In a wider range: Bockius 2009. The phenomenon is also known from Roman barges: Bockius 2004, p. 127; p. 129, note 15; p. 132, fig. 10.

5 Jézégou et al. 2009. The author is especially indebted to M.-P. Jézégou for giving him the opportunity to discuss the finding and providing him with her manuscript in 2008. See also the article of M.-P. Jézégou in this volume.

6 For Late Bronze and Iron Age disc wheels with un-pegged plug fasteners, similar to tenon joining cf. van der Waals 1964:, p. 69-74, fig. 24-25, pl. VIII; Kunwald 1970, p. 53-54, fig. 9; Woytowitsch 1978, p. 25-28; Joachim (H.-E.) – Bonner Jahrbücher, 179, 1979, p. 448-450, fig. 5-6; Schovsbo 1987 (variety of type D3); Heussner (K.-U.) – Bodendenkmalpflege in Mecklenburg, Jahrbuch, 1985, p. 125-131; Pugin (Ch.), Corboud (P.), Castella (A.-C.) – Archäologie der Schweiz 11, 1988, p. 146-154; Kimmig 1992, p. 61-62, pl. 37,3; Taylor 2001, p. 213-215, fig. 7.48. For a general view of the prehistoric wheel cf. Burmeister, Fansa 2004.

Auteur

Leiter des Forschungsbereiches Antike Schiffahrt
Museum für Antike Schiffahrt
Römisch-Germanisches Zentralmuseum
Neutorsraße 2b, D – 55116 Mainz (Deutschland)
bockius@mufas.de

© Publications du Centre Camille Jullian, 2011

Conditions d’utilisation : http://www.openedition.org/6540

Lire

Open access

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search