Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 15 October 2019

Lake-Dwellings in the Alpine Region

Abstract and Keywords

This chapter discusses difficult extensive archaeological excavations; stratigraphy, dendrochronology, and the rhythm of the habitations; the first wetland dwellings south of the Alps; lake-shore villages north of the Alps; Neolithic village space in southwest Germany and Switzerland; and lake-shore villages and terramare villages of the Early/Middle Bronze Age. The wetland villages in the Circum–Alpine region constitute one of the richest documents of European prehistory, the reason for which is the extraordinary state of conservation of the remains in waterlogged conditions. It is however a historical document with gaps, for periods of abandonment are as long as those of occupation. This interrupted history cannot be understood without placing it in the wider context of the regional histories from the Neolithic to the end of the Bronze Age.

Keywords: wetland sites, lake-shore villages, wetland archaeology, Alps, Circum–Alpine region, terramare, habitation, Neolithic


Since the recognition of the very first prehistoric lake-shore village at Obermeilen (Lake Zurich) in January 1854, the number of sites identified the wetlands of the Circum-Alpine is close to 1,000 (Suter and Schlichtherle 2009; Schlichtherle 1997). This could well be a minimum, given the number of sites discovered at Clairvaux and Chalain (Jura, France) and the eastern shore of Lake Neuchâtel (Switzerland) (Hafner and Suter 2000), where systematic prospecting and core drillings have recently led to the doubling of the number of lake-shore dwelling sites known in 1990. Since it covers the entire Circum-Alpine region, the lake-dwellings phenomenon can no longer be considered an exception of a limited area, but a widespread occurrence (Coles and Lawson 1987). Within this relatively unified geographic unity, many marshes (but by no means all), lands subject to flooding, and lake shores were selected to create permanent hamlets or villages. This does not mean that the phenomenon occurred in a contemporaneous manner over such a wide area, or that the reason for the creation of these singular dwellings was the same everywhere. Other than the chronology and the cultural affiliations, where obvious differences are apparent, the only common denominator between these dwellings (rapidly named ‘pile dwellings’) was the choice of a wetland area, either subject to flooding or permanently under water, on which to build houses (Fig. 15.1), far from dry land, cereal fields, grazing lands, and hunting territories.

Difficult Extensive Archaeological Excavations

 Lake-Dwellings in the Alpine

figure 15.1 Aerial view of the Prise d’Eau archaeological zone at Chalain (Fontenu, Jura, France). Three dwelling sites appear as small wooded spurs, separated from the shore by low-lying marshes subject to flooding. (Photograph: P. Pétrequin.)

On these sites, which favoured the conservation of vegetable matter, the wood used in construction and manufacture of artefacts, the first archaeologists were mostly concerned with the discovery of collectable objects, often in a remarkable state of conservation: wooden (p. 254) receptacles, cloth, axes with wooden shafts, and antler tools. These extraordinary objects, extremely rare in the context of dry sites, rapidly led to the international reputation of these marsh and lake-shore villages (Munroe 1908). Their very number had a notably detrimental effect on the quality of the fieldwork: the race for exceptional collectable items to display in curiosity cabinets, and feed private and museum collections, often led to poorly organized excavations, in which the stratigraphic sequences were considered of little importance, until the beginning of the 20th century, and occasionally well beyond. From these first field explorations, in which the landowners went as far as granting ‘exploitation rights’, grew the notion of a ‘lake-shore civilization’, considered totally different from the farming communities settled on dry land elsewhere in Europe. The need therefore grew to question the reasons for this so-called cultural specificity, perhaps motivated by reasons of defence, or specific to a fish-based economy.

Furthermore, the abundance of extremely well-preserved remains, the depth of successive stratigraphies to be investigated, and the technical difficulties in reaching the anthropic levels below the water table rapidly limited the extent of archaeological surfaces investigated on wetlands. As a consequence, little was then known about the villages themselves and how these were organized.

Archaeological research was re-initiated in the marshes of the Federsee, in southwest Germany, where R. R. Schmidt was able to uncover the entire surface of several Neolithic hamlets (Schmidt 1930). These extensive excavations were made possible at that time because they were bog villages, buried at shallow depths, in a marsh in the process of being artificially drained and dried out; furthermore, the anthropic stratigraphies were neither thick nor complex. Rectangular houses with beautiful plank floors thus appeared, successively built to approximately the same plan; the archaeological remains were sparse, so that extensive areas (p. 255) could be rapidly uncovered without being literally submerged by tens of thousands of artefacts—which still today is the key factor limiting the extensive archaeological excavations of hamlets or small agglomerations. It is within this context, in which the stratigraphic observations have much improved, that archaeologists developed a special interest in the architecture and layout of the villages. Two schools then opposed each other: the one arguing for houses built at ground level when the water level and the water table sank sufficiently, the other arguing for raised-floor houses (the famous ‘pile-dwelling villages’). This squabble came to a head with the pamphlet published by Paret (published in German in 1948, but only translated into French in 1958), which purported to demonstrate that all wetland villages were built at ground level; there were therefore no differences between villages built on dry land and wetland ones, other than the fact that the latter were subsequently covered by the rising waters, and lake sediments or peat. At one fell swoop the special features of wetland sites were thus eradicated. Paret's observations were however limited to shoreline marshes, and could not be generalized. From this standpoint, the new theory obscured the different architectural adaptations to lake shores and marshes which we now know how to recognise (Fig. 15.2) (Pétrequin and Pétrequin 1984).

A mechanistic interpretive model was postulated to explain the dynamics of wetland habitations in the Circum-Alpine region: the farmers built their villages in the marshes and on the lake shores during dry climatic periods, specifically to supply water for their flocks and herds; when the climate became more humid and the water level rose, the villages had to be abandoned (Vogt 1954). This explanation is still widely accepted, especially given that variations in the lake level are excellent climatic indicators (Magny 2004).

Stratigraphy, Dendrochronology, and the Rhythm of the Habitations

 Lake-Dwellings in the Alpine

figure 15.2 One type of foundation adapted to mechanically unstable lake sediments: the Middle Bronze Age example from Fiavè (Trentino, Italy), with posts with horizontal tenons resting on a mesh of orthogonal perches. (Photograph: P. Pétrequin.)

The development of new hypotheses coincides with the reorganization of the archaeological services after the Second World War and the restarting of archaeological excavations. The observations then began to benefit from the stratigraphic approaches developed for the Palaeolithic, whilst new techniques now gave improved access to submerged sites and archaeological levels below the water table; these included the use of pumps, shoring up with planking, and earthwork retaining walls, as well as scuba diving in the 1960s. Starting in 1970, the (p. 256) number of wetland excavations began to increased rapidly, especially in the context of large-scale salvage operations during the construction of motorways and the excavation of deep foundations for high-rise buildings in Switzerland (Collective Papers 2004), the draining of wetlands and the construction of yachting harbours in southwest Germany (Schlichtherle 1997), the creation of tourist areas and beaches in France, and more recently, large-scale research excavations in Germany, France, and northern Italy (Suter and Schlichtherle 2009).

The extent of archaeological excavations now covers the main part, if not the entire extent, of certain villages, where the foundation posts are systematically charted and studied. The development of dendrochronological dating—which in the best cases allows determination of the year in which the tree was felled, especially with oak and ash—now appears as a real revolution in wetland archaeology, opening the possibility of determining (in solar years) the building chronology of houses and villages, as well as that of the stratigraphic sequences and the relative pace of the technical and cultural evolution (Billamboz 2006).

The rhythms of wetland dwellings, as well as variations in the water level, could then be determined with an unequalled level of precision, on average to within a decade; this permitted comparisons between villages and regions hundreds of kilometres apart. The most recent syntheses now confirm the relationship—previously postulated—between periods of low water and the intensive development of wetland sites (Magny 2004). This unequivocal result merely determines the dynamics in terms of presence/absence, which needs to be nuanced. There are in fact many sites where structural timbers cannot be dendrochronologically dated, either because the trees are too young, because they come from strands totally manipulated by people, or because they belong to undatable species (Pétrequin 1997). Moreover, the systematic surveying of the land some distance from the lake shore has brought to light lake-shore villages constructed during periods of high water, such as at Chalain (Jura, France) during the Iron Age, which were contructed some 100m from the shore. This means that the absence of dwellings, at certain periods, on the ancient lake-shore shelves could well be linked to the displacement of villages in accordance with the variations in the water level (Pétrequin et al. 2005).

We need to face the obvious: the proven correlation between low water level and the occupation of wetland villages masks a much more complex reality. Certain periods of low water offer no examples of lake-shore villages, such as the Bell–Beaker period, during the second half of the third millennium (Pétrequin et al. 2005). Divergent situations are also apparent during the Middle Bronze Age between north of the Alps, where wetland villages are virtually absent (Magny 2004), as opposed to the southern slopes of the Alps, where they reach their apogee precisely during this period (Perini 1984; 1988).

Moving to the regional level, we can observe that Neolithic villages are particularly numerous on the shores of Lake Neuchâtel ( Switzerland) during the 28th century bc, whilst 80km away, on the far side of the Jura, they are completely absent at Chalain and Clairvaux (Pétrequin 1997). Judging from a purely local level, the lack of symmetry of lake-shore dwellings at Chalain and Clairvaux (Jura, France)—only 12km apart as the crow flies—is striking: when the lake at Clairvaux is occupied, the lake shores at Chalain are for the most part deserted, and vice versa, as though there were a more or less regular shift in the population from one lake to the other (Pétrequin et al. 2005).

 Lake-Dwellings in the Alpine

figure 15.3 Differing rhythms of occupation from one period to another. Top: a rapid succession of short-lived villages at Chalain 3 (Fontenu, Jura, France), at the end of the fourth millennium. Bottom: continuity of occupation during a century or longer, at the beginning of the third millennium at Clairvaux/La Motte-aux-Magnins (Jura, France). (Photographs: P. Pétrequin.)

The rhythms of occupation can also differ between villages and between periods. Thus, the rapid succession of short-lived villages (10–20 years of occupation) during the fourth millennium in southern Germany, western Switzerland, and the French Jura (Fig. 15.3, top), contrast with the longer lasting villages in the third millennium (Fig. 15.3, bottom) (Wolf and Hurni 2002). (p. 257)

Finally, by using dendrochronology, it is sometimes possible to date the periods of occupation of the villages precisely, even though the time lag between the abandonment of the villages and the rise of the water level is generally not known. Any number of reasons can be envisaged other than the rising water: destruction by fire often marks the end of the occupation of a village, sometimes with such regularity that we can ask ourselves whether it is not a deliberate ritual at the time of abandonment. In the French Jura, we have been able to demonstrate that all the villages on one of the small lakes were abandoned several tens of years prior to the first indications of rising waters (Pétrequin 1997).

Without a shadow of doubt, we are faced with singular rhythms of occupation which not connected with the sole variations in water level. We have thus reached the limits of the climatic hypothesis; we therefore need to turn to the manner in which society functioned—the economic models, the demographic pressures, and the state of the environment exploited by agriculturalists—to apprehend these striking differences (Pétrequin et al. 1998). It would seem that the differences between one lake and another, between one site and another, have been underestimated in an attempt to reach a monocausal explanation: the climatic factor was postulated as the primary mover to explain the dynamics of the occupation of wetland sites, without paying any real attention to adaptive processes or social and historical factors (Schibler et al. 1997).

In the Sixth Millennium, The First Wetland Dwellings South of the Alps

The oldest wetland site villages correspond, as far as we know, to a southern European phenomenon which appears very early on, south of the Alps—in fact, from the second half of the sixth millennium bc. This corresponds to the second phase of the Early Neo (p. 258) lithic, after the initial agricultural colonization coming from the shores of the Mediterranean. Interestingly enough, it is not a strictly Alpine phenomenon, tied to the lakes and humid depressions of the maximal extension of the glaciers. In fact, the dwellings of La Marmotta, on Lake Bracciano (Lazio , Italy), also belong to this current of Neolithic colonization. Radiocarbon-dated to 5600–5200 cal bc (Fugazzola Delpino 1995), La Marmotta yielded characteristic cardium decorated, engraved and painted pottery. We know little, however, about the structure of these villages, other than their location in relation to the shoreline and their present submerged position. By the same token, the site of La Draga at Banyoles (Catalonia, Spain) belongs to the Cardial. This lacustrine village, whose duration of occupation is not clearly defined, was constructed on the lake shore around 5400 cal bc, and yielded some remarkable wooden artefacts, previously totally unknown from dry-land sites, such as sickles with obliquely set flint blades (Bosch et al. 2000).

These two sites, La Marmotta and La Draga, established within an unmistakably Mediterranean context, demonstrate the special attraction of lacustrine environments in which to build villages, within a few centuries of the initial Neolithic colonization, possibly when demographic expansion began to bite, and the Neolithic began to spread onto different types of soil. The same can be said about the basin of Lake Varese (Lombardy), just south of the Alps, where a number of wetland sites have been identified dating from the end of the sixth millennium bc: Pizzo di Bodio and Isolino Virginia. The oldest settlements, dated to approximately 5400 bc, belong to the Early Neolithic culture, and show cultural relationships with Vho, Gaban, and the Ligurian Impressa. In the case of Isolino Virginia (c.5060 cal bc), beds of timbers were laid flat in an orthogonal manner, suggesting rectangular shaped habitations, whose flooring was placed directly on the damp marsh soil (Banchieri 1997). The village was built on a small island and occupied for a long period of time, as evidenced by the 1.90m-thick archaeological strata and the pottery, which together supply evidence for almost the entire Neolithic up to the end of the third millennium bc (Banchieri 1990). Another example of a lake-shore settlement in the Mediterranean region, dating from before 5000 bc, also exists in northern Greece, at Dispilo (Lake Kastoria) (Touloumis and Hourmouziasi 2003).

In northern Italy, the occupation of wetland sites continues to develop during the entire fourth millennium bc, in the period of the Square Mouth Pottery (Vasi a Bocca Quadrata) culture (Aspes 1982). The best-known sites are Pizzo di Bodio, already mentioned, on Lake Varese, and Molino Casarotto on Lake Fimon (Veneto); there again, badly preserved flooring was excavated, suggesting rectangular structures (Bagolini et al. 1973).

North of the Alps, where the Neolithic appears from 5400 bc, wetland locations do not appear to have been favoured. Of course, around the Federsee (southwestern Germany) as well as the former Lake Wauwil (Lucerne, central Switzerland), Mesolithic camp sites (e.g. Schölz 7) evidence special interest in deer hunting. The only indication of an Early Neolithic lake-shore settlement has been identified at Chalain (Jura), dating from around 4400–4200 bc, possibly in association with Hoguette-type pottery, and Egolzwil (central Switzerland). The regional populations appear to favour instead the valley-floor lands best suited to the cultivation of cereals, in line with the Danubian expansion model on sedimentary or loess soils. (p. 259)

North of the Alps from the 43rd Century bc

The oldest lake-shore villages north of the Alps belong to the Egolzwil culture, whose pottery attests to a strong southern tradition, connected to the Cardial expansion in the Middle Rhone Valley. As far as village organization is concerned, only the eponymous site of Egolzwil 3 (Lucerne, Switzerland), on the edge of the former Lake Wauwill, has been the subject of extensive excavations. It is a large village dating from the 43rd century bc: more than 25 rectangular wooden buildings have been identified, made up of housing units with domestic hearths and possibly storage houses. The village space, bounded by a fence, extends parallel to the ancient shoreline, with approximately three rows of constructions 4m wide and 8–10m long, whose rooflines were perpendicular to the shore (Wyss 1994). In the past few decades, several villages belonging to the Egolzwil culture have been identified in central Switzerland; for instance on the shores of Lake Zug and Lake Zurich, indicating the permanence not only of the material culture but also of the organization of village space, over at least two centuries (Collective Papers 2004).

When the construction of lake-shore villages took off on the Swiss Plateau (with the Cortaillod culture) and in the French Jura (with the Néolithique Moyen Bourguignon) starting in the 39th century bc, the following general layout was used more often: one or two rows of rectangular buildings parallel to the shore, their roofs perpendicular to the shoreline, and a solid continuous fence or palisade at least on the landward side. This southern building tradition will continue in western Switzerland and the Jura until at least the 35th century bc. At Clairvaux (Jura, France) (Fig. 15.4), the layout is particularly clear, with the single row of houses, with their long sides close together, whilst the entrances face the lake with the work areas, two raised walkways, and a group of granaries built on a shallow in the lake.

This stable type of architectural layout can comprise tiny hamlets, such as the one at Burgäschi (Berne, Switzerland), or more important villages made up of some twenty structures hidden behind a strong palisade, such as at Hauterive-Champréveyres (Neuchâtel, Switzerland). In all the instances, an impression is given of isolation from dry land, a desire to bar and control the access points, and soft ground into which it was easy to sink the supporting timbers of the structures. The choice of these damp lands, which could be subject to flooding, and where it was not possible to circulate without sinking in, meant that the architecture needed to be adapted (flooring timbers laid on the soil, or raised on piles), and the virtually systematic use of vegetable matter to stabilise the soil.

 Lake-Dwellings in the Alpine

figure 15.4 Clairvaux-les-Lacs (Jura, France), station II. A Cortaillod-type hamlet, with a palisade on the landward side, a single row of houses, and a group of granaries in shallow water (35th century bc). (Documentation: P. Pétrequin; drawing: A. Viellet.)

Stratigraphic and dendrochronological studies have always shown that the life span of these Middle Neolithic villages was short, barely exceeding 10–20 years, before being abandoned for varying lengths of time, then occasionally being rebuilt on approximately the same ground plan. Such cycles of building, occupation, abandonment, and reconstruction are particularly evident at Clairvaux and Twann-Bahnhof (Berne, Switzerland); they were perhaps related to swidden agriculture, which presupposes large tracts of forest clearance, low population density, and the need to move villages on a more or less regular basis (Pétrequin et al. 1998). In all the instances, domesticated animals do not appear to have had access to the village centres, (p. 260) where the soil was too soft; furthermore, pollen analyses indicate that grass fields and pasture land are missing, and that large farm animals, sheep and pigs, were mostly fed in the forest.

Neolithic Village Space in Southwest Germany and Switzerland

It would appear that population density in the most fertile valleys had to attain a certain threshold before the need to colonize regions between river systems and on the plateaux made itself felt, involving the clearance for agriculture of lands substantially less appropriate for the introduction of slash-and-burn and fallow-forest swidden agriculture.

During the second half of the fifth millennium bc, the oldest villages are to be found in the marshes of the Federsee, north of Lake Constance (Germany). In a similarly manner to Egolzwil, the first hamlets group a number of rectangular houses with wooden floors made of perches laid in a criss-cross pattern, as in the cases of Aichbühl and Schussenried (Schmidt 1930). The structures are laid out in approximately one or two rows, perpendicular to the shoreline, as in the Swiss model at that time.

The layout of the villages tends to change from the 39th century bc onwards. In the case of Hornstaad-Hörnle I (Lake Constance), the rectangular buildings were still laid out in an (p. 261) irregular manner (Dieckmann et al. 2006), but later on the roof-lines would be mostly parallel to the shore. We are thus witnessing a reorganization of the village space which is in opposition to the Egolzwil–Cortaillod type of village we have previously seen. This new arrangement will become more common with the Pfyn culture, strongly influenced by oriental stimuli, which develops in northeastern Switzerland from the 38th century bc. We here have a totally new social concept of village space: several rows of tightly packed houses, lining parallel streets perpendicular to the shore; the circulation streets are now privileged, whilst solid fences or strong palisades protect these demographically larger villages on the landward side (Société Suisse de préhistoire et d’archéologie 1995).

Whilst the oriental cultural influence expands towards the Swiss Plateau and the French Jura northwest of the Alps, this new model of village will progressively gain acceptance in a southwesterly direction. The Zurich region is the first to be touched with the Pfyn culture, around 3800 bc, followed by the Horgen, starting in the 33rd century bc. With the arrival of Horgen pottery styles in western Switzerland, the model is transmitted as far as Lake Neuchâtel, and a little later, to the French Jura.

This organization of the axes of communication between dry land and the shore, such as at Arbon-Bleiche 3 (Thurgau, Switzerland), leading to reorientation of village space, could well be related to the increased sedentism of the population, with certain villages being uninterruptedly occupied for several decades towards the first half of the third millennium (Leuzinger 2000). There were new methods of managing forest land and new agricultural practices, in particular with the introduction of yoked oxen, the ard, and the first wheeled vehicles (Pétrequin et al. 2006), even if it is doubtful whether these innovations were available to all. It is also in the third millennium that we discern the first pastures for domestic animals, and find that game rapidly dwindled in favour of domesticates (Schibler et al. 1997; Pétrequin et al. 1998).

As can be seen, the model of village street, ensconced behind a strong palisade, is directly tied to the social and economic conditions induced by the increasing pressures in the Late Chalcolithic from the Balkans to the northwestern Alps on wetlands and dry land, whether in Bavaria, at Pesternacker and at Unfriedhausen, or in Switzerland, until the middle of the third millennium (Schönfeld 1990; Schlichtherle 1997).

These lake-shore villages, in which the houses were (although not all) tightly packed behind defensive enclosures, are not an indication that the lacustrine communities had developed specialized economies. All the archaeological cultural elements found on wetland sites are echoed on dry-land ones, which were also fortified at this time. One possible exception consist of the villages of the Lüscherz culture (found on Lakes Neuchâtel, Bienne, and Morat, western Switzerland, during the 29th and 28th centuries bc), which could have represented a special form of habitat tied to a specific cultural assemblage (Société Suisse de préhistoire et d’archéologie 1995).

Lake-shore Villages and Terramare of the Early/Middle Bronze Age

With the 24th century bc, the wetland habitats disappear completely from the Circum-Alpine region. Only a few pots can be signalled here and there, not even necessarily in relation to a village: one example comes from Sutz (Lake Neuchâtel), another from the Lake of Bourget (p. 262) (Savoie, France), a third, finally, from Lake Varese-Isolino (Italy). It is not much to show for a crucial period of history, when Europe-wide large-scale cultural phenomena affected regional cultural evolutions. Is this the reason for the abandonment of the lake-shore settlements? It is likely that time saw the beginning of a long period of low lake-water levels, which only came to an end in the 16th century bc (Magny 2004; Pétrequin et al. 2005).

North of the Alps, a new period of colonization of wetland sites slowly got under way during the Early Bronze Age. Starting around 2100 bc, the first manifestations appear with a hamlet at Clairvaux-La Motte-aux-Magnins (Jura), built on a small peninsula in the lower marsh; its defensive position cannot be doubted. It is, however, only during the 18th century that the phenomenon will gather pace. The large lakes are now affected, with larger villages, on occasion comprising tens of buildings, and systems of successive palisades which permit the strict control of access from dry land: Lake Constance with the sites of Ludwigshafen-Seehalde, Haltnau-Oberdorf and Arbon-Bleiche; Lake Zurich, with ZH-Mozartstrasse, where three rows of rectangular houses were built on chassis of crossed beams; Baldeggersee (Lucerne, Switzerland), with posts held in place by foundation tenons; Lake Neuchâtel, with Concise-Sous Colachoz, which incorporated impressive palisades; Lake Geneva, where Morges-Les Roseaux is a key site of the Rhône culture; and Lake Annecy, where a sturdy palisade, framing the entrance to a village at Sevrier-Les Mongets with its plank walkway, has recently been identified (Suter and Schlichtherle 2009).

Thus, north of the Alps, wetland habitation sites underwent a brief period of expansion between approximately 1800 and 1600 bc, at the same time as the increase in metallurgical output, and the occupation of fortified sites on dry land.

The dynamics of lake-shore dwellings then collapse rapidly, and around 1500 bc, the last known settlements are those of Forschner in the Federsee marsh (Upper Swabia, Germany), Bodman-Schachen 1 (Lake Constance), ZH-Mozartstrasse (Lake Zurich, Switzerland), and Arbon-Bleiche 2 (Lake Constance, Switzerland), where the final construction phase occurred at the end of the 16th century bc (Menotti 2001). The correlation between the generalized abandonment of lake-shore villages and the worsening weather conditions, which corresponds to higher water levels between 1400 and 1200 bc, cannot be questioned here (Magny 2004b). During this entire period, unfavourable for wetland villages (at least so far as north of the Alps is concerned), the dynamics of farming communities’ site expansion will be played out on dry land.

South of the Alps, in the region of maximum extension of the glaciers as well as on the floodplains of the River Po, the dynamics of villages adapted to wetland sites is quite different. The story starts in a broadly similar manner as in Switzerland, with a steady growth in the number of lake-shore settlements, after a long hiatus from the end of the third millennium bc. Rapidly, however, large villages are built, which demarcate the organizational network of cultivation on the heavy morainic soils. On the lake shores, amongst the tens of sites identified since the end of the 19th century, those of Lake Ledro, Fiavè (Trentino), and Lavagnone (Brescia) are the most remarkable (Perini 1984; 1988). Some can be considered veritable little agglomerations, such as Viverone (Piedmont), reached by log tracks and totally enclosed by strong palisades (Fozzati et al. 1998). Such lake-shore agglomerations, which counted several hundred souls, clearly bring to mind the so-called terramare villages, established at this time on the floodplains of the River Po. Those around Parma (Emilia Romagna) are amongst the best known, with their fortified entrances and many rows of closely packed houses served by parallel streets, which represent good examples of ‘proto-urbanism’; a the (p. 263) good example being the terramare of Santa Rosa di Poviglio (Villagio Grande) (Bernabó Brea et al. 1997).

In fact, the whole of northern Italy is affected by generalized occupation of these lands with heavy soil, which begins around 2100–2000 bc (Aspes 1982). We are faced here with an expansionist economic and social dynamism, in the context of a society in which craft specialists expand their skill at the same time as long-range exchanges. This period of economic progress and demographic expansion is long-lasting in northern Italy (De Marinis 1999). Contrary to what appears to be the case north of the Alps, the climatic deterioration of 1400–1200 bc progressively puts a damper on the terramare habitation system and on the lake-shore dwellings, which disappear soon after.

During the Early–Middle Bronze Age, we can thus discern two broadly similar dwelling systems developing north and south of the Alps. But the dynamics of their evolution were markedly different: a slow development followed by a rapid disappearance in Germany and Switzerland (coinciding with a period of climatic degradation); and a rapid expansion, a long apogee, and at the end a gradual decline of a much more resilient technical and social system in northern Italy.

‘Proto-Urbanism’ North of the Alps?

The period 1100–800 bc corresponds to a period of lower water levels. While the population grows at a pace not previously seen before 1200 bc, the marshes and shoreline shelves are once again colonized with permanent villages. Interestingly enough, this is no longer a unique phenomenon of the Circum-Alpine region, but it occurs in northern Europe as well (for instance the famous fortified village of Biskupin).

The utilization of houses with raised flooring is generalized between Lake Constance and Lake Bourget, since villages are built far from the shore line, with remarkable and complex palisades. Two main architectural techniques were employed: cross-beam chassis with tenons to block the upright posts, used in central Switzerland as well as in the Federsee marshes (see Zug-Sumpf (Zug, Switzerland) and Wasserburg Buchau (Ferdersee)); and dwellings on stilts, more widely used everywhere else (Vogt 1954; Société Suisse de préhistoire et d’archéologie 1995).

What is most remarkable is the organization of village space inside the palisades. In one of the best case studies, Cortaillod-East (Neuchâtel, Switzerland) (Fig. 15.5), narrow passages separate the rows of rectangular closely packed houses, which cover the entire site (Arnold 1986). It is most probably the organization also observed in the terramare of northern Italy. The internal architecture of the buildings is also identical. We are therefore entitled to ask ourselves whether this form of close-packed, symmetrically laid out proto-urbanism was not introduced from northern Italy, where wetland dwelling sites were in rapid decline. However, different layouts were also present in the northern part of the Alps (for instance Greifensee-Böschen, Lake Pfäffikon and Ürschhausen-Horn, Lake Nussbaum, Switzerland—Eberschweiler et al. 2007; Gollnisch-Moos 1999).

 Lake-Dwellings in the Alpine

figure 15.5 The village of Cortaillod Est (Neuchâtel, Switzerland) during the Late Bronze Age. Note the palisade on the landward side and the parallel pathways, marked out by the large number of posts. (Drawing: A.M. Pétrequin, according to the Neuchâtel Dendrochronological Laboratory.)

The extraordinary wealth of these villages north of the Alps should also be stressed, in terms both of the weight of bronze objects—most of which were of local manufacture—and of the imports from afar, whether small objects made of tin or personal adornments made of (p. 264) Baltic amber. We must therefore recognize that the ‘fine period of the lacustrine Bronze Age’, as it was called in the 19th century, demonstrated undoubted originality as regards both its architectural designs and the quantity of metal objects manufactured. No equivalent sites have been found on dry land, so far as eastern France and southwest Germany are concerned. We have here, without a doubt, the material expression of an original social system and of a successful regional economy, substantially equivalent to the large terramare from south of the Alps, during the Early/Middle Bronze Age.

Around 800 bc, possibly consequent on rapid deterioration of the climate and rising water level of the lakes, the lake shores were quickly abandoned (Magny 2004). We could be looking at a relocation of the villages in response to the changing climatic conditions. During the entire Iron Age, encompassing the Hallstatt and La Tène periods, there will no longer be any wetland villages, even when the climatic conditions improved around 600 bc; the changed social systems no longer favour them. (p. 265)


The wetland villages in the Circum-Alpine region constitute one of the richest documents of European prehistory (Coles and Lawson 1987; Menotti 2004). The reason for this is the extraordinary state of conservation of the remains in waterlogged conditions. It is however a historical document with gaps, for periods of abandonment are as long as those of occupation. This interrupted history cannot be understood without placing it in the wider context of the regional histories from the Neolithic to the end of the Bronze Age. To merely explain the correlation of the lake-dwellings phenomenon with the variations in the lake levels (which is not in doubt) appears simplistic, and of little heuristic value. If the settlement of these villages away from dry land in an unfavourable environment can most certainly be explained in terms of defence, we need to look at the manner in which society functioned and how social relations between groups took place, to understand what drove the lake-dwellers to take such actions.

(Translation: Michael Templer)


Arnold, B. (1986) Cortaillod-Est et les villages du lac de Neuchâtel au Bronze final. Structure de l’habitat et proto-urbanisme. Archéologie neuchâteloise 6. Saint-Blaise: Editions du Ruau.Find this resource:

Aspes, A. (1982) Barche di Solferino. In F. Gaggia (ed.) Palafitte: Mito e Realtá: 193–5. Verona: Museo Civico di Storia Naturale.Find this resource:

Bagolini, B., Barfield, L.H. and Broglio, A. (1973) Noticie preliminari delle ricerche sull’insediamento neolitico di Fimon-Molino Casarotto (Vincenza). Rivista di Scienze Preistoriche 28 (1): 161–215.Find this resource:

Banchieri, D. G. (1990) Il neolitico del Pizzo di Bodio nelle Prealpi varesine. In P. Biagi (ed.) The neolithisation of the Alpine Region: 191–6. Brescia: Museo Civico di Scienze NaturaliFind this resource:

——  (1997) L’abitato preistorico dell’Isolino Virginia-Biandronno. Varese: Musei Civici.Find this resource:

Bernabò Brea, M., Cardarelli, A. and Cremaschi, M., (eds) (1997) Le terramare: la più antica civiltà padana. Catalogue of the exhibition. Modena: Electa.Find this resource:

Billamboz, A. (2006) Dendroarchäologische Untersuchungen in den neolithischen Ufersiedlungen von Hornstaad-Hörnle. In Landesamt für Denkmalpflege (ed.), Siedlungsarchäologie im Alpenvorland IX: Forschungen und Berichte zur Vor- und Frühgeschichte in Baden-Württemberg 98: 297–414. Stuttgart: Theiss.Find this resource:

Bosch A., Chinchilla, J. and Tarrùs I Galter, J. (ed.) (2000) El poblat lacustre neolític de La Draga. Excavacions de 1990–1998. Girona: Monografies del CASC.Find this resource:

Coles, J. M. and Lawson, A. J. (eds), European Wetlands in Prehistory: 1–21. Oxford: Clarendon Press.Find this resource:

Collective Papers (2004) Sur les traces des palafittes. Archéologie Suisse 27.2: 2–89.Find this resource:

De Marinis, R. C. (1999) Towards a relative and absolute chronology of the Bronze Age in northern Italy. Notizie Archeologiche Bergomensi 7.2.Find this resource:

Dieckmann, B., Harwarth, A. and Hoffstadt, J. (2006) Hornstaad-Hörnle IA: die Befunde einer jungneolitischen Pfahlbausiedlung am westlichen Bodensee. In Landesamt für Denkmalpflege (ed.), Siedlungsarchäologie im Alpenvorland IX: 8–276. Stuttgart: Theiss. (p. 266) Find this resource:

Eberschweiler, B., Riethmann, P. and Ruoff, U. (2007) Das spätbronzezeitliche Dorf von Greifensee-Böschen: Dorfgeschichte, Hausstrukturen und Fundmaterial. Zurich: Kantonsarchäologie Zürich.Find this resource:

Fozzati, L., Martinelli, N. and Evans, S. P. (1998) Le strutture palafitticole dei siti sommersi del lago di Viverone (TO-VC). In Preistoria e protostoria del Piemonte (Atti della XXXII Riunione Scientifica IIPP, Alba, 29 settembre–1 ottobre 1995): 46–9. Alba: Atti della XXXII Riunione Scientifica IIPP.Find this resource:

Fugazzola Delpino, M. A. (1995) Un tuffo nel passato: 8000 anni fa nel lago di Bracciano. Rome: Museo nazionale preistorico etnografico Luigi Pigorini.Find this resource:

Gollnisch-Moos, H. (1999) Ürschhausen-Horn: Haus- und Siedlungsstrukturen der spätestbronzezeitlichen Siedlung. Frauenfeld: Departement für Erziehung und Kultur des Kantons Thurgau.Find this resource:

Hafner, A. and Suter, P. (2000) –3400 v. Chr.: Die Entwicklung der Bauerngesellschaften im 4 Jahrtausend v. Chr. am Bielersee. Bern: Lehrmittel- und Medienverlag.Find this resource:

Leuzinger, U. (2000) Die jungsteinzeitliche Seeufersiedlung Arbon Bleiche 3. Frauenfeld: Kanton Thurgau.Find this resource:

Magny, M. (2004) Holocene climatic variability as reflected by mid-European lake-level fluctuations, and its probable impact on prehistoric human settlements. Quaternary International 113: 65–79.Find this resource:

Menotti, F. (2001) ‘The Missing Period’: Middle Bronze Age Lake-Dwellings in the Alps. Oxford: Archaeopress.Find this resource:

—— (ed.) (2004) Living on the Lake in Prehistoric Europe: 150 Years of Lake-Dwelling Research. London: Routledge.Find this resource:

Munroe, R. (1908) Les stations lacustres d’Europe aux Âges de la Pierre et du Bronze. Paris: Schleicher.Find this resource:

Paret, O. (1958) Le mythe des cités lacustres. Paris: Dunod.Find this resource:

Perini, R. (1984) Scavi archeologici nella zona palafitticola di Fiavè-Carera, part 1: Situazione dei depositi a dei resti strutturali. Trento: Servizio Beni Culturali della Provincia Autonoma di Trento.Find this resource:

——  (1988) Gli scavi nel Lavagnone. In Annali Benacensi (ed.), Atti del simposio internazionale sui modelli insediativi dell’età del Bronzo: 109–54. Cavriana: Annali Benacens.Find this resource:

Pétrequin, A. M. and Pétrequin, P. (1984) Habitat lacustre du Bénin: une approche ethno-archéologique. Paris: ADPF.Find this resource:

——  (ed.) (1997a) Les Sites littoraux néolithiques de Clairvaux et Chalain (Jura), III: Chalain 3, 3200–2900 av. J.-C. Paris: Maison des Sciences de l’Homme.Find this resource:

—— Arbogast, R. M., Bourquin-Mignot, C., Lavier, C. and Viellet, A. (1998) Demographic growth, environmental changes and technical adaptations: responses of an agricultural community from the 32nd to the 30th centuries BC. World Archaeology 30.2: 181–92.Find this resource:

—— ——  Pétrequin, A. M., Van Willigen, S. and Bailly, M. (2006) Premiers chariots, premiers araires: la diffusion de la traction animale en Europe pendant les IVe et IIIe millénaires avant notre ère. Paris: CNRS.Find this resource:

—— Magny, M. and Bailly, M. (2005) Habitat lacustre, densité de population et climat: l’exemple du Jura français. In P. H. Della Casa and M. Trachsel (eds), WES ’04:Wetland Economies and Societies: 143–68. Zurich: Chronos.Find this resource:

Schibler, J., Hüster-Plogmann, H., Jacomet, S., Brombacher, C., Gross-Klee, E. and Rast-Eicher, A. (1997) Ökonomie und Ökologie neolithischer und bronzezeitlicher Ufersiedlungen am Zürichsee: Ergebnisse der Ausgrabungen Mozartstrasse, Kanalisationssanierung Seefeld, AKAD/Pressehaus und Mythenschloss in Zürich. Zürich: Kantonsarchäologie Zürich/Fotorotar. (p. 267) Find this resource:

Schlichtherle, H. (ed.) (1997) Pfahlbauten rund um die Alpen. Stuttgart: Theiss.Find this resource:

Schmidt, R. R. (1930) Jungsteinzeit-Siedlungen im Federseemoor. Augsburg: Filser.Find this resource:

Schönfeld, G. (1990) Die Ausgrabung in der jungneolithischen Talbodensiedlung von Pestenacker, Ldkr. Landsberg am Lech. Bericht der Römish-Germanischen Kommission 71: 355–80.Find this resource:

Société Suisse de préhistoire et d’archéologie (1995) La Suisse du Paléolithique à l’aube du Moyen Âge, II: Néolithique. Basel: Verlag Schweizerische Gesellschaft für Ur- und Frühgeschichte.Find this resource:

Suter, P. J. and Schlichtherle, H. (2009) Pile dwellings. UNESCO World Heritage Candidature. ‘Prehistoric Pile Dwellings around the Alps’. Bern: Palafittes.Find this resource:

Touloumis, K. and Hourmouziasi, A. (2003) The man and the lake: living in the lake-side settlement of Dispilo, Kastoria, Greece. Mediterranean Archaeology and Archaeometry 3.2.Find this resource:

Vogt, E. (1954) Pfahlbaustudien. In U. Guyan (ed.), Das Pfahlbauproblem: 119–221. Schaffhausen: Schweizerische Gesellschaft für Urgeschichte.Find this resource:

Wolf, C. and Hurni, J.P. (2002) L’Environnement forestier au Néolithique final à travers l’étude dendrochronologique des sites littoraux du lac de Neuchâtel. In H. Richard and A. Vignot (eds), Équilibres et ruptures dans les écosystèmes durant les 20 derniers millénaires: durabilité et mutation: 387–402. Actes du Colloque International de Besançon (2000). Paris: Les Belles Lettres.Find this resource:

Wyss, R. (1994) Steinzeitliche Bauern auf der Suche nach neuen Lebensformen: Egolzwil 3 und die Egolzwiler Kultur. Bern: Schweizerisches Landesmuseum Zürich.Find this resource: