Times fade away. The neolithization of the southern Netherlands
Verhart, L.B.M.
Citation
Verhart, L. B. M. (1999, June 2). Times fade away. The neolithization of the southern Netherlands.
Retrieved from https://hdl.handle.net/1887/13346
Version:
Corrected Publisher’s Version
License:
Licence agreement concerning inclusion of doctoral thesis in the Institutional
Repository of the University of Leiden
Downloaded from:
https://hdl.handle.net/1887/13346
3.1 Introduction
This chapter deals with the core region Grave-Cuyk (fig. 3.1). It is a study of a regional distribution pattern of sites in relation to the investigation of a single site. The analysis of the settlement pattern using a.o. GIS is outside our scope.
This analysis will be published elsewhere1.
3.2 The Middle Neolithic in the south of the Netherlands
In 1973 the term ‘Middle Neolithic of Limburg’ was first
introduced as a result of the Koningsbosch investigation2.
This denoted chronologically the period between the Early Neolithic Bandkeramik and the Late Neolithic beaker-cultures. Culturally this Middle Neolithic encompasses the Michelsberg(MK)-culture (Middle Neolithic A) and the
Wartberg/Stein/Vlaardingen(WSV)-complex3(Middle
Neolithic B). In this outline the emphasis will be on the MK-phase and less attention will be paid to the WSV-phase. The Middle Neolithic can in a material sense be recognized in the Meuse valley by the — rare — occurrence of mostly undecorated pottery with few typical features and large flint concentrations, both in number and in size. The sites usually palimpsests represent an accumulation of moments of use, not only from the Middle Neolithic, but often as early as the Mesolithic and as late as the Iron Age. The limited
typological differences make it hard to distinguish pottery and flint by period. Another remarkable phenomenon is the large number of sites dating from the Middle Neolithic. The origins of the MK-culture lie in the German Lower Rhine area. There a gradual typological development in the pottery can be discerned from Rössen to MK I and
subse-quently MK II4
. Lüning assumed that not until phase MK III
an expansion towards the north and west occurred5. This idea
has been superseded by the occurrence of MK II-pottery
outside this supposed centre, even as far as the Paris Basin6.
As a result of recent research in the Limburg Meuse valley itself, but particularly in the adjoining areas like the west of the Netherlands, West Germany and Belgium, the early phase of the Middle Neolithic, the period of the MK-culture,
can be described more accurately7.
In the south of the Netherlands Maastricht-Vogelzang8and
the recently discovered causewayed enclosure in Heerlen9
3
Linden, a Neolithic activity area in the core region Grave-Cuyk.
Macro region Core region
0 20 km
are the sites where this early pottery occurs. In rough outline the material culture of the northwestern group of the MK-culture is characterized by the use of almost undecorated pottery, of which pots with round bases and outward bending rims (tulip beakers) are most typical. In addition jars, bottles, spoons, storage jars and baking plates occur. The composition of the range of pottery is often tied to a particular region. Specific variations developed within the regions. A striking example is the Hazendonk-3 pottery decorated much more often and with a small range of shapes. In the flint industry macrolithic tools predominate, usually made of imported mined flint. Characteristic tools are various types of quite thick arrowheads with semi-surface retouch, pointed blades, large blade and flake scrapers and axes, both of flint and stone.
As yet, little is known of the economy of the Dutch MK10
. Information on the food supply is mainly provided by sites in the west of the Netherlands, like Brandwijk, Hazendonk, Rijswijk and Wateringen that date from this period, but may
not be considered as typically Michelsberg11. In this
particular wetland area, not highly suited for agriculture, it is not amazing that hunting and fishing were important activities. Besides livestock, agriculture too contributed to the diet, as grain is known to occur on these sites. It is however doubtful whether grain was cultivated at the site itself. This would imply that elsewhere — probably in the drier coversand area — agriculture contributed to a high
degree to the food supply12
. The MK-sites containing grain are mainly located in the löss area. It is striking that on those sites mainly wheat was cultivated, whereas on the sites in the west of the Netherlands emmer and einkorn wheat were
still used13. Apparently the use of emmer and einkorn wheat,
that already played such an important part in the food supply in the Bandkeramik, continued in the coversand area. It is remarkable that the production of tools was completely focused on good quality imported flint from the south, despite
the presence of suitable local material. The most important sources were the flint mines of Rijckholt, but flint was also imported from Spiennes, the Haspengouw in Belgium and even
the north of France, from the vicinity of Romigny-Lhéry14
. Little is known of the lay-out of the settlements. Until recently the clearest example of a dwelling in the sandy area was a small rectangular building from Kruishoutem in Belgium. However, this building turns out to have sprouted from the excavator’s imagination and should therefore be removed as a source of knowledge of MK-culture
settlements15
. Nevertheless, small rectangular buildings were in use during this period, as can be inferred from the recently excavated settlement Wateringen IV. This is a small rectangular structure with dimensions of 11 by 4 metres. From Thieusies in Belgium a structure is known that might
be a building16
. Within a causewayed enclosure a
configuration of poles has been observed that are assumed to be housing constructions, but might also be part of the defense system. In France at Mairy in the Meuse valley a large number of floor plans have been found with
dimensions as large as 60 by 10 metres17
. These floor plans are considered to be houses, but there is an alternative explanation that this is an equivalent of the ‘causewayed enclosures’, where various social and ritual acts were performed. On the Aldenhovener Platte, on site Koslar 10, a large number of pits and post holes have been excavated, allowing the inference of a rectangular structure of 6 by
12 metres18
. From some settlements in the eastern river district post traces are known, indicating the presence of
houses there. Examples are Kraaienberg near Linden19and
Pater Berthierstraat in Grave20
. In the latter settlement fragments of clay used for huts have been found as well. Besides permanent or semi-permanent settlements other types of settlements are known as well. A small temporary
encampment is e.g. the site excavated in 1981 in Gassel21. In
as ‘causewayed enclosures’. These are mainly known from the adjoining countries, Belgium and Germany, and dominate settlement research there. In the Netherlands the
first specimen was recently discovered in Heerlen22. It is not
impossible that there has been and partly still is one on the
Caberg near Maastricht23
.
a b c
d e f
Almost nothing is known of burial procedures. In the sandy area nothing has been found up to now, whereas in the west of the Netherlands two cemeteries are known, a smaller one
at Zoelen24and a larger one at Rijswijk-Ypenburg25. In
Belgium burials in caves are known26
The second phase of the Middle Neolithic is distinguished by the appearance of groups that are characterized archaeologically
as Wartberg/Stein/Vlaardingen27
. After the discovery of the
Stein vault28, it became clear that much of the undecorated,
quartz-grit tempered pottery belonged to this phase29.
Subsequently it became apparent that there were strong similarities in a material sense between the Vlaardingen-culture in the west of the Netherlands and the Wartberg group from Hessen in Germany, and that all three could be
considered a single cultural unit: the WSV-complex30.
Morphologically speaking, the range of pottery may be called poor. Simple cooking pots predominate, with occasional appearances of small trays, bowls and collared flasks. Few diagnostic features are apparent in the flint industry as well. Most noticeable are transverse arrowheads and tanged points shaped like pine trees. Although occasionally macrolites may occur, the importation of large semi-manufactured items like
flakes and blades appears to have come to an end31
.
Almost nothing is known of the economy of these groups on the coversands in the south. The only source of information on this period are the excavation data from the west and middle of the Netherlands. Although it is remains to be seen whether these are representative of the south of the
Netherlands, they do provide a picture of the possibilities. The food economy shows a relatively wide range with settlements where hunting and fishing predominate
(Vlaardingen, Hekelingen, Hazendonk)32, versus settlements
with the emphasis on livestock (Voorschoten, Leidschendam,
Ewijk)33. There are also indications for agriculture, but this
often appears not to have occurred at the site or in the
vicinity34. The picture is quite different in the south of the
Netherlands. From Maastricht-Randwijck a pollen diagram is known exhibiting the presence of a ‘landnam’ in the WSV-phase, which might indicate a considerable contribution of agriculture to the food economy in the south of the
Netherlands in this period35. Comparable observations have
been made from the adjoining area in West Germany36
and
the eastern river district near Nijmegen37.
Whereas there were hardly any indications for houses in the MK-phase, a small number are known for the WSV-period
from the west of the Netherlands38. In most cases these were
small, rectangular, two-naved houses, but incomprehensible post clusters occur as well which might each represent a large number of consecutive, non-rectangular huts. Despite the scarcity of excavation data, more is known of burial procedures in this period. The dead were individually
buried (Hekelingen)39
, with a wide range of burial rituals, or
interred in a collective grave, like Stein40. Cremation of the
dead occurs with remarkable frequency in this period.
3.3 The macroregion
In the macroregion there is a large number of sites from the Middle Neolithic, but it is hard to define the exact nature of these sites. Therefore no statements can be made about function and economy either. A relatively large number of guide artefacts can be distinguished for the Middle Neolithic (fig. 3.2). For the Middle Neolithic A (MK-culture) these are: three types of arrowhead, macrolithic flakes and blades, and pottery. Due to its fragility the last category is severely under-represented.
Finds occur all over the Meuse valley, from Maastricht to ‘s-Hertogenbosch, but with a distinct concentration in the middle of Limburg, southeast of Roermond (fig. 3.3a-f). Individual objects like arrowheads and axes may indicate the occurrence of occasional activities like hunting, gathering food or forestry work on that location or nearby. In order to ascertain the presence of a settlement, specific guide artefacts, or a combination of guide artefacts, have to occur on a site.
A settlement is in this study defined by: – the presence of pottery
– the presence of two out of five guide artefacts, excluding a combination of two points
In the settlement pattern, too, a concentration in the middle of Limburg can be discerned (fig. 3.3f), with a smaller number to the south and north. The differences in density may be caused by habitation intensity, but more likely reflect
the collection and reporting efforts of amateurs41
. A good instance are the MK-finds in the core region Venray. On the map of the macroregion no settlements are visible, whereas investigation of the core region unearthed a quite impressive
number of these in this region42. Despite these limitations the
pattern clearly shows that Middle Neolithic communities had settled over almost the entire area under investigation.
3.4 The core region Grave-Cuyk
The core region Grave and Cuyk is located in the eastern part of the river district, in the north of the province of Noord-Brabant. Its size is indicated in figure 3.4. To the north the core region is bounded by the Meuse. Within the area the major cities are Grave in the west and Cuyk in the
east. This core region has a surface of approx. 90 km2and is
one of the smallest investigated within the Meuse valley project.
3.4.1 GEOLOGY
For the description of the geology we used the study by Van
aimed at the area around the microregion Linden in particular, but thanks to its broad geographical set-up, it provides an up to date basis for the description of a large part of the core region as well (fig. 3.5). For the areas not investigated in the core region we used supplementary
literature44.
In more than one respect the geology of the core region Grave-Cuyk is unusual. Apart from the fact that the geological development is clearly visible, the so-called terrace intersection occurs here, the transition from the western river district mainly characterized by sedimentary processes to the southern river district characterized by almost continuous erosion.
In the penultimate ice age, the Saalien, the basis for the modern terrain was laid. The ice cap pushed far to the south and created the lateral moraine system at Nijmegen and
Groesbeek. As a consequence Meuse and Rhine sought their way west to the sea as a single meandering river. Due to infilling of the valley floor by coarse-grained sand and gravel a river plain came into being that sloped to the west. This river plain is known as the Rhine and Meuse lower
terrace45.
In the Allerød construction of the lower terrace was complete and its destruction began. The Meuse drained into a number of deeply cutting meandering main channels — the Rhine had shifted its course to the north in the middle of the Weichselien. The channels carrying less water now began to infill.
In the Younger Dryas a new terrace level came into being, due to infilling of Allerød-cuts and erosion. This level is
referred to as the X-terrace46
. In this period large sand drifts occurred as well, creating a large number of dunes of
Mook Linden Beers Cuijk Grave Gassel 0 2km
varying shapes and sizes. As a result of predominantly western winds and the relatively northern course of the Meuse in the middle and north of Limburg, these dunes occur mainly on the east bank. In the research area the Meuse ran a more westerly course and the river dune systems are more numerous and smaller in size. One of the largest now houses the villages of Klein- and Groot-Linden. Small river dunes usually remained undeveloped. Instances of this are Kraaienberg, Nielt, Op den Drul and Hoge Renning. In the south of the research area coversand layers formed, under the influence of aeolian processes. At the same time a layer of loam was deposited from the active channels on the valley plain, referred to as ‘variegated
loam’47and covering almost the entire valley plain, with the
exception of the tops of the highest river dunes. The Praeboreal was characterized by a deep cut of the Meuse into the X-terrace; at the same time erosion lowered
this terrace by approx. 1 metre. The Meuse cut was concentrated on two channels, the northernmost identical to the modern bed of the Katwijkse Meuse, the southern one dividing into two beds at Den Drul and merging again at De Geest. In the middle Hoge Renning is located, as an island. This course was to determine the geological development of the area. At De Geest the channel was approx. 300 m wide and cut into the X-terrace to a depth of up to 4,5 m. In the next phase, the Boreal (fig. 3.6), the course of the Meuse centred on the northern bed, while the southern bed began to infill. From a C14-dating from the base of the peat layer in this channel, at Sluisgraaf, this infill appears to start
early in the Boreal48. Outside these channels developments
are momentarily at rest. There are no more sand drifts, no cuts of new channels and inside this area there is no
sedimentation at all until the 16th century. Occasional floods deposit some clay into the channels only.
Mook Linden Beers Cuijk Grave Gassel 0 2km
A radical change is caused by the realization of the Beerse Overlaat, as a result of which there is a forced sedimentation from the 16th century onwards, resulting in a clay layer of approx. 1,5 m thick covering the lower areas like a blanket. Only the tops of the sand dunes project over it.
3.4.2 TERRAIN
The area between Grave and Cuyk is gently sloping, with a clear division in land use between the higher, dry lands and lower, wet lands. The higher lands encompass the coversand area, the sandy fluvial deposits and the river dunes. Besides serving as settlement locations, these are mainly used as arable land. Occasional forests occur. The lower lands are dominated by livestock, although improved drainage and EC subsidies have recently stimulated the growth of agriculture, in the form of maize cultivation.
From southwest to northeast three geographical units may be
distinguished in the research area (fig. 3.6). In the southwest of the research area is the northeastern end of the extensive continuous Brabant/Limburg coversands. This coversand area was originally partly covered by peat, but has come to the surface as a result of the reclamation of the Peel. The next unit is a wide, relatively flat strip of land with fluvial deposits consisting mainly of sand. This area was originally cut by channels running southeast to northwest in general. This area contains river dunes as well.
The third unit is a strip of land with more relief, consisting of mainly clayey fluvial deposits with smaller and larger river dunes. The differences in height are partly related to the greater height of the river dunes here, but the relative difference in height is also caused by the low-lying fluvial deposits.
in drainage, the natural course of brooks and rivers is hard to distinguish. Originally the drainage will mostly have occurred by the old channel system. In addition some water courses can be recognized to have been natural originally. Instances of this are Graafsche Raam with its tributaries like Hooge Raam, Halse Beek, Laarakkerse Waterleiding and Sluisgraaf. Graafsche Raam drained surplus water from the Peel area, whereas Sluisgraaf handled the river district between Cuyk and Grave.
3.4.3 HUMAN INFLUENCE
Initially human influence in the area and its effects on the terrain was limited, due to the presence of the Beerse
Overlaat49
. From Carolingian times man has been actively fighting the flood waters of the Meuse, first by constructing a system of quays, designed to stop the water from the river and pools. In the final stage of this development the quays and dikes were built along the river itself. From 1300 onwards a single continuous Meuse dike exists, stretching
raised and it became possible to close the Beerse Overlaat (fig. 3.8). In 1941 the work had been completed and the Beerse Overlaat no longer existed.
The results of the activities of the Beerse Overlaat are clearly visible in the terrain. Besides the numerous places where the dike was punctured (pools) the Overlaat can also be inferred from the habitation pattern, concentrated on the highest points in the terrain: the river dunes. No farms existed in the lower areas before 1940. The persistent threat of flooding caused people not to settle in the area of the spillway and stopped villages from developing there. The result is an extremely pristine agrarian terrain.
The pristine nature of the terrain is by now its greatest threat. The absence of habitation turns out to offer excellent economic opportunities for sand and gravel extraction. After all, agricultural land is cheap and relatively easy to
expropriate, in contrast to built-on land that is usually spared
under those circumstances. In order to meet the growing demand for sand and gravel, exploitation of these resources started in this area in 1971. Over the past 25 years a large lake district has been created here as a result. Current plans for the period to 1998 will result in the exploitation of another 155 ha.
Apart from the fact that a beautiful (in our opinion) and original river district is fast disappearing, the consequences for archaeology will be disastrous as well. Most seriously, archaeological sites will disappear completely. Ironically, due to the sand and gravel extraction many sites will be discovered as well. Although hardly any archaeological finds were known from this area prior to 1971, this has changed dramatically, thanks to the unremitting activities of a small group of amateur archaeologists.
Among human influence (albeit unintentional) on the terrain, drifting sands should be reckoned. To the southwest and east
Mook Linden Beers Cuijk Grave Gassel 0 2km
of Grave relatively large areas of drifting sands have developed in the Middle Ages.
More active human influence did not occur to a large extent in the recent past. In the southwest of the research area the Peel lies, reclaimed by the cutting of peat, and scattered over the entire research area a small number of plaggensoil. In the north some instances of plaggensoil occur, mainly in the higher parts of the river dunes, larger numbers are located in the southwest of the research area.
The area has at present a strongly agrarian nature, with mostly mixed farming. The emphasis in the southern part of the area is on pig farming, whereas diary farming is more prevalent in the north. In agriculture mainly maize and beetroots are grown.
3.4.4 ARCHAEOLOGY OF THE CORE REGION
Until the seventies not many finds were known. For a long time many Roman finds had been retrieved in Cuyk and as a result, research often concentrated on this. In the countryside between Grave and Cuyk several Neolithic axes had been found and in 1928 an investigation had been carried out at Dommelsvoort, south of Linden into the presence of urns (fig. 3.9).
Fig. 3.9 Newspaper cutting from the ‘Echo van ‘t Land van Cuyk’, dated 9 June 1928 concerning finds in Linden.
Collector fields mining operations
Arts 0 1 Van Hassel 0 1 De Jong 0 2 Keereweer 2 1 Koeling 1 2 Koolen 1 3 Nagengast 0 1 De Wit 1 3 Verscharen 0 1
no activities (0), occasionally (1), regularly (2), intensively (3)
Table 3.1 Summary of major collectors in relation to the relative intensity of their labour and search locations in the period 1970-1990.
Site nr. indet. Early -Middle- Late- Neo.
45F-109 x 45F-113 x x 45F-116 x 45F-125 x x 46A-108 x x 46A-113 x x 46A-118 x x 46A-119 x x 46A-122 x x 46A-127 x x 46A-130 x x 46A-131 x x x 46A-133 x x 46A-136 x x 46A-138 x x x x 46A-142 x x x 46A-144 x 46A-146 x x x 46A-147 x x 46A-161 x x x 46A-176 x Total 9 7 3 8
Table 3.2 Dates of Mesolithic sites. Last column indicates the presence of Neolithic material on the site.
The amateur activities also led to the realization of a large number of excavations in the area; settlements from e.g. Stone, Bronze, and Iron Ages were investigated, as well as
from the Roman and medieval periods50.
3.4.4.1 Amateur activities
Amateur archaeologists only became interested in this area at a late stage. An important incentive was the formation of
the AWN-branch Nijmegen and Vicinity in 1968. This branch concentrated mainly on Nijmegen itself and the eastern part of the river district. As some members from south of the Meuse, the Land of Cuyk and Grave, joined, interest in this area increased. Two pioneers deserve mention: Mr. J. Koeling from Cuyk and Mr. M. Koolen from Grave. They were walking encyclopedias for other amateurs in the vicinity. As a result, ever more people became interested in archaeology and some form of cooperation arose. A contributing factor was that chance finds by laymen in the field of archaeology became generally known sooner, as the amateurs were regularly contacted by those regional finders.
As large parts of the research area have been covered by relatively recent sediments, the amateurs employ a research strategy different from that used in the south of the
Netherlands. There the finds are all over the place and can be collected by surveying the fields. This search strategy is only useful in the region Cuyk-Grave in those parts that have not been covered with sediments. As only a relatively small part of the area meets this requirement, it is not surprising that such a strategy is barely used in this region. The determining factor in the collection strategy of the amateur archaeologists were the numerous mining operations in the area. From the early seventies onwards the large mining operations on behalf of gravel exploitation began to the north and northwest of Linden. The levels exposed after the removal of the overlying sediments were regularly checked, as well as the profiles of the excavation walls. And successfully, as quite soon the first finds were recovered. Over the years attention was concentrated on these areas, not only because of the good opportunities for finding anything,
Mook Linden Beers Cuijk Grave Gassel 176 163 168 183 180 179 184 99 174 171 162 161 159 153 58 155 136134 122 125 124 126 139 149 148 145 132 120 133 127 130 142 144 147146 150 138 121 112 116 110 108 114 113 112 113117 115 119 114 118 33 92 90 100 107 106 109 131 0 2km
unknown exact location known exact location newly discovered sites
155 112
62 156
80
but also because of the large number of artefacts and the large amount of work that collecting entails.
Thanks to the protective covering relatively more pottery than flint is found in this area, unlike the vast majority of the coversand area in the south of the Netherlands, where persistent weathering ensures an under-representation of pottery.
Amateur activities were therefore concentrated on several areas opened up by digging activities (table 3.1). Therefore the archaeological data collected in this way present a biased picture of the overall find spread. We therefore do not have an even distribution pattern of the entire region. The pattern displays marked clusters in those areas that were searched often, the areas where mining operations occurred. The areas with less search activity are apparent as white spots.
3.4.4.2 Inventory and composition of distribution maps
The majority of the Stone Age finds from this core region date from the Neolithic. Occasionally older finds occur, like a Middle Palaeolithic hand-axe and some artefacts of
possibly the same age51
. In addition a small number of Late Palaeolithic and Mesolithic finds are known. In this chapter emphasis will be on a summary of the available archaeological data from the Neolithic.
As far as we could ascertain, all amateur collections, finds by non-amateur archaeologists and collections in museums have been inventoried. In early 1992 the inventory was closed. The items have been described according to a standard code and entered into a computer file. The number of sites in the inventory is 118. The numbers of finds per site range from 1 to 36.367 artefacts.
0 2km
The distribution map is a perfect illustration of the collection strategies of local amateur archaeologists (fig. 3.10, 3.11). The map is characterized by large areas with no or hardly any finds, alternating with small areas displaying a strong clustering. The distribution map is to a large extent therefore the result of amateur activities and barely reflects Mesolithic and Neolithic habitation intensities. Recent human disturbances like plaggensoil, drifting sands and peat cutting did not play a significant part in the realization of the pattern. Actually the mining activities at Linden and the construction of a new housing estate to the south of Cuyk dominate the pattern.
Mesolithic
The number of Mesolithic sites is 21 (fig. 3.12). The numbers of finds per site are too low to allow any statements
to be made on the function of the sites. But it is possible to date a number of sites more accurately than simply Mesolithic, on the basis of the typochronology of points (table 3.2).
On almost all sites Neolithic material has been recovered as well, with the exception of two Early-Mesolithic sites and possibly one site that can not be dated due to a lack of characteristic points. In all areas where Late Mesolithic activities can be identified, Neolithic activities occur as well. Although the number of sites is not large, the differences in distribution pattern by phase are striking. The Early Mesolithic sites are mainly located in the southern part of the research area, on the transition from coversand to river plain (fig. 3.13). In this period the river dunes do not appear to have been attractive locations for settlement. Apparently
0 2km
0 2km
0 2km
this changed in the Middle to Late Mesolithic (fig. 3.14-3.15). Sites from this period, scarce though they are, are more often located in the river dune area and less in the transition area between coversand and river plain. This choice of location was apparently also preferred in the Neolithic, since Neolithic material was found on all of these Late Mesolithic sites as well
Neolithic
The majority of the finds from the core region consists of Neolithic artefacts. Most of these have been chance finds -for instance axes (fig. 3.16)- and/or single finds collected in field reconnaissance. In addition some excavations were conducted. The most important sources are three Middle-Neolithic settlement areas that were excavated, to wit:
Linden-Kraaienberg52
Northeast of Linden, on the former hill Kraaienberg an area with strongly disturbed edges was investigated in 1977, with traces of habitation from Middle and Late Neolithic and the Iron Age. Scattered over the area a large number of soil traces have been found that we have been unable to arrange in any sort of order. The numerous traces of posts and their size indicate these were part of some sort of dwelling that can not be defined more precisely. Exceptionally a pit was found containing at least seven well-preserved pots of the Stein-group. This pit is taken to be a ritual deposition.
The majority of the pottery can be attributed to the Middle Neolithic and, on the basis of typological and technological features, to the Hazendonk-2 phase53. Most striking features are
the tempering and coiled construction of the pottery, the rims bending outward, the presence of ‘Tupfenleisten’, the occurrence of carination and the small number of decorated sherds, whose
0 2km
most typical adornment has been applied with a stick, the so-called ‘pinpricks’.
Another indication for this age is provided by the flint. The use of imported Rijckholt-type flint for the manufacture of tools is typical. All tools recovered on Kraaienberg are made of this material. In addition, some large, unused flakes and a trio of blades fitting together have been found.
Besides pottery and flint, a number of fragments of millstones and grindstones have been excavated in the area. Although these can not be dated typologically, they are attributed to the Middle Neolithic habitation.
On the basis of the data provided above, the site Kraaienberg is considered to have been a settlement where a large number of non-specialist activities occurred. The size of the terrain and the presence of soil traces are indicative of (semi)permanent settlement, presuming an agrarian way of life.
No statements can be made on the nature of the activities dating from the WSV-phase.
Gassel54
Some years after the discovery of Kraaienberg, another Middle Neolithic site came to light north of Gassel.
In 1983 a small, sharply demarcated concentration of pottery and flint with a diameter of approx. 7 metres could be excavated there. Soil traces were not uncovered. If these had been present in first instance, they will have been destroyed completely by subsequent soil processes.
The large numbers of decorated pottery show great similarities to pottery recovered in Hazendonk phase 3. Gassel most likely dates from this phase. The pottery displays a wide morphological range. Besides small bowls and pots that may be considered tableware, some large pots can be distinguished that probably served for storage. The flint displays the same features as Kraaienberg. All tools have been made of imported Rijckholt-type flint. Some items have been made of flint originating in the Haspengouw, Belgium. Among stone tools mill- and grindstones occur once again.
0 2km
In its overall composition the find material displays marked similarities to that of Kraaienberg. The distribution of the finds and differentiation in distribution patterns of the various artefact classes point to a relatively short period of use of the settlement, however. This brief use of the settlement does not imply this would have been a ‘special activity site’. The presence of all find categories is inconsistent with this. It would therefore appear that the traditional basic functions known from agrarian settlements are represented here as well. This site may be considered a ‘base camp’, with the herding of livestock for a season in the immediate vicinity a likely but hard to prove option.
Grave-Pater Berthierstraat55
During a short investigation in the spring of 1988 a fragment of an originally larger settlement area at Zittert, west of Grave, could be investigated. The site is on the east bank of the river Raam which ran there originally.
Some post traces were recovered here with sherds and pieces of loam with wattle impressions, probably from the wall of a house. Around these traces sherds and flint were scattered. The pottery is similar to both Kraaienberg and Gassel in features. Both Hazendonk 2 and 3 are represented, so evidently the area was used over a longer period or on two separate points in time. The function of this site is considered to have been identical to the settlement at Kraaienberg. In Grave, too, we would be
dealing with a (semi)permanent settlement used over a long period of time, with a primarily agrarian function. The large number of arrowheads, like Kraaienberg, are indicative of a still large contribution by hunting to the food economy.
For the distribution of settlements and the use of the terrain in the Neolithic, the distribution pattern of the axes provides a first indication (fig. 3.16). The axes are the type of artefact most insensitive to amateurs, as even laymen can recognize and collect these tools. These occur in all types of terrain, but there is nevertheless a clustering in areas traditionally visited by amateurs.
A further investigation into the economy and use of the terrain in the Neolithic is hampered by the small amount of finds on most sites, making it difficult to make statements on settlement functions. In order to get an impression of the use of the area in the various periods, a series of distribution maps has been drawn up. The basic assumption is that pottery or a combination of two or more types of flint artefacts are indicative of a settlement. For a further distinction the pottery has been used as the main
distinguishing criterion. Sites that did not yield any pottery are more difficult to date as flint — in particular the arrowheads — is a less accurate cultural indicator. These
Site nr. indet Haz 1 Haz 2 Haz 3 Mi-Neo WSV beaker
46A-156 x 46A-159 x 46A-161 x 46A-162 x 46A-163 x 46A-168 x 46A-171 x 46A-174 x 46A-179 x x 46A-180 x 46A-183 x x 46A-184 x x 46A-210 x 46A-213 x 46A-214 x x 46A-216 x 46A-217 x 46A-218 x 46A-219 x 46A-221 x 46A-222 x 46A-223 x Total 24 5 11 19 8 21 21
sites without pottery can be attributed to the Middle Neolithic, on the basis of the presence of numbers and combinations of macrolites and various types of MK-arrowheads. The WSV-phase is characterized by the occurrence of transverse and pine-like points and the beaker-phase by the occurrence of types of points with a complete surface retouch. Sites with a single flint guide artefact have therefore not been included in the distribution maps.
Hazendonk 1-phase
A small number of sites (5) have yielded some sherds highly similar to material uncovered in the oldest layer of
Hazendonk, borough Molenaarsgraaf and in Brandwijk. The most important features of this pottery are the notched rims and impressions, ornamental impressions in squares and the
tempering with organic material56
.
All instances of Hazendonk-1 material from the region Grave and Cuyk are undecorated and may therefore only be
distinguished by the use of organic temper57. A single sherd
is equipped with an engobe; one sherd is apparently a fragment of a round base (fig. 3.17:3).
Finds from this phase occur exclusively on the small river dunes in the northern part of the research area and the large river dune of the Zittert, west of Grave (fig. 3.18).
Hazendonk 2-phase
For the identification of this material we may draw on the one hand on the results of research at the Hazendonk and Brandwijk, and on the other hand on the results of excavations at Kraaienberg near Linden and Pater
Berthierstraat in Grave58
. The number of sites in the core region has doubled in relation to the previous phase. The pottery is mainly undecorated, quite thin and fired hard, constructed in coils and characterized by a tempering of quartz, chamotte and sand. Morphologically plates and pots with outward-bending rims can be identified. Both flat and rounded bases occur. Most striking features are folded-over moulded rims, carinated shoulders, ‘Tupfenleisten’ as rim finish and decorative patterns in the form of ‘pinpricks’ and impressions on the carinated parts of the sides (fig. 3.17). In addition there are other features that, however, also occur in the next phase, Hazendonk 3, like surface coarsening and perforated lugs. The chronological continuity of many of these features makes distinction between these phases therefore quite difficult. The distribution of the sites is primarily the result of amateur search activities (fig. 3.19). The three clusters: in the west Zittert, in the centre the sand and gravel extraction between Gassel and Linden and in the southeast the new housing estate of Cuyk, all indicate that in the Middle Neolithic both 1 2 3 4 5 6 7 8 9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Fig. 3.17 (continued) 10-26 Hazendonk 2/3 (46A-138: 10, 12-17, 20-26; 46A-139: 19; 46A-148: 11, 18). Scale 1:2.
small and large river dunes were used, as well as the Pleistocene terraced lands in the southeast.
Hazendonk 3-phase
For this phase as well the research at Hazendonk and the settlements at Gassel and Grave-Pater Berthierstraat are
representative59. The number of sites in the core region is 22.
0 2km
0 2km
1 2 3 15 14 13 12 11 10 9 8 7 6 5 4 15 14 13 12 11 10 9 8 27 16 17 18 19 20 21 22 23 24 25 26
and fingertip impressions, drawn lines, applied lumps of clay and bands. Lugs occur as well, both perforated and
unperforated (fig. 3.17; 3.20).
Technologically the pottery appears to have declined, it is generally thicker compared to the preceding phase and the coiled construction is as often not visible. In the range of shapes plates do not occur frequently in this phase. There is a predominance of quite weak profiles and the pottery usually has a flat base, without a protruding foot. The distribution of the sites corresponds to a high degree with that of the previous phase, the only difference being an increase in the number of sites on the small river dunes in the north (fig. 3.21).
Middle Neolithic phase
Due to the absence of pottery -or the occurrence of pottery without specific features- a number of sites can only be
dated with the aid of flint guide artefacts. For the Middle Neolithic these are leaflike, teardrop-like and triangular arrowheads with semi-surface retouch and large tools, flakes and blades of imported flint, mainly of the Rijckholt-type. The distribution of the sites is markedly different from the preceding three phases (fig. 3.22). This time the sites are located in the area to the south and east of Grave. However, this is not due to a changing choice of location, but rather to different conservation conditions and amateur search activities. The sites south of Grave are almost exclusively surface sites, where pottery is badly preserved. By the preserving action of overlying sediments in the northern part of the research area and the manner in which these sites are discovered, the pottery has been better preserved and the chances of being found by amateurs have increased. Evidently the pattern of sites with pottery and that of sites without pottery are both distorted, but taken together, they do present a
0 2km
complementary pattern. The entire river valley was apparently in use during the Middle Neolithic, with the exception of the coversand area, where so far no sites have been discovered.
Wartberg/Stein/Vlaardingen (WSV) phase
This phase has been defined by Louwe Kooijmans60and is
characterized by the use of mostly undecorated pottery. Technologically it is distinctive by its quartz, sand and chamotte temper, the lack of coiled construction, the presence of base feet and S-shaped wall profiles. Another feature, in particular for the Vlaardingen-component is the use of rim perforations (fig. 3.23).
There is a great deal of uncertainty about the use of flint in this phase in the east and south of the Netherlands. On the basis of finds from the west of the Netherlands and some instances from the east and south, transverse and pine-like arrowheads are apparently characteristic. Macrolites do not
appear to have played a significant part in this phase, as these are absent from the delta.
The distribution of sites matches those of the Middle Neolithic and the phases of Hazendonk 1-3 (fig. 3.24). The same area is in use and the number of sites on the large river dunes appears to increase relatively. As far as can be determined, all sites may be considered settlements, with the exception of an infilled pit discovered during the excavation of Kraaienberg, which was probably a ritual deposit. The pit also demonstrates how hard it is to define this phase chronologically, as a rim sherd with herringbone decoration was retrieved from this pit as well. The Vlaardingen-material
in the delta continues into the AOO-phase61
. This also appears to be the case for SOM, which means we have to take into account that part of the distribution pattern surveyed might also belong to the next chronological phase,
0 2km
the beaker period, characterized by the occurrence of beaker material.
Beaker cultures phase
The Late Neolithic in this area is distinguished by the occurrence of protruting food beaker and bell beaker pottery and flint arrowheads with a complete surface retouch. In the preceding phase there were occasional signs that the finds did not solely reflect settlement activities. These signs become more distinct in this phase. Three sites may be interpreted as graves. In the vicinity of Kraaienberg a
bottom/body fragment of a bell beaker has been found62.
According to the finder it was located among cremation remains. Due to the presence of Iron Age urn interments in the same terrain and the rarity of beaker phase cremation remains, we doubted whether this could be interpreted as a burial find. As will become apparent later on, this find is not unique to the region. As more graves with cremation
remains are known from this period63, it is likely this is a
burial find after all, unlike the ideas expressed by one of us before.
The other two burial finds are an almost intact bell beaker found to the northwest of Gassel and a second bell beaker found in the immediate vicinity with a large number of
34 35 33 31 32 29 28 27 30
associated finds. To start with the latter: in the period of heightened mining operations Mr. J. de Wit from Grave found some sherds just before nightfall in a recently cut area on the edge of the dredge lake. Further investigation of the surface by the light of his car’s headlamps revealed even more. The final tally was the almost complete bottom of a Veluvian bell beaker, a grinding and a hammer stone, two large unprocessed flakes, an amber pendant decorated with spiral motifs and two thin gold clip-on earrings (fig. 3.25). Although de Wit had found no indications for a burial site such as a platform at that location and traces of ditches and post holes have not been observed, the find composition is in favour of an interpretation as burial inventory.
Some days afterwards de Wit surveyed the area again and approx. 80 metres to the north he found a second
concentration of bell beaker sherds. These proved to come
from an almost complete Veluvian bell beaker, but no other finds were recovered. Here, too, the almost intact nature of the find and the lack of attendant settlement material are in favour of an interpretation as a grave.
The remaining material from the region may be classified as settlement material (fig. 3.26). Apart from the distinctive decorated beaker sherds, sherds from normal ‘domestic’ pottery were found in these areas, mostly tempered with chamotte, sand and some quartz. These are often relatively thick sherds, decorated with nail, fingertip and spatula impressions.
The settlement pattern demonstrates a distribution and number of sites almost similar to the preceding phase. Remarkably, the number of sites in the area with small river dunes has decreased sharply. There appears to have been a certain preference for the larger river dunes and the remains
0 2km
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
of terraces south and southwest of Cuyk (fig. 3.27). Over time, the settlement pattern has been subject to quite some changes. In the Mesolithic the entire area was in use, initially only the south, but in the Late Mesolithic the northern river dune area as well. This latter zone was also the preferred settlement area in the Middle Neolithic. By the end of the Neolithic the settlement preferences changed: the river district falls out of favour and the coversand area appears to be more in use.
Although actual food remnants are absent, the settlement changes may be related to changes in the food economy. When hunting, fishing and food gathering are the main means of subsistence, the settlements are located mainly in the geographically heterogeneous terrains, also distinguished by their large geographical diversity and attendant wide range of flora and fauna. In the Late Neolithic less
heterogeneous geographical terrain is preferred, indicating a
more agrarian lifestyle. This pattern can also be
distin-guished in the core region Venray64.
3.5 The microregion Linden
3.5.1 INTRODUCTION
To gain a more detailed understanding of the settlement history and use of this river district area, we selected a small microregion for further investigation. The region lies to the south of Linden and measures approx. 2 by 1.5 kilometres (fig. 3.28).
This choice was partially prompted by the severe threat to the area. The core region in its entirety was to be the subject of the next stage of mining operations. As we write, the research area has been converted to concrete, cement and paving material and the area may be visited by boat. The main purpose of our investigation was to document which activities were performed here by prehistoric man
0 2km
and, if possible, find and investigate an undisturbed settlement from the Middle Neolithic before it was
destroyed. The various stages of the investigation, including
excavation, covered the years 1989-199265.
3.5.2 GEOLOGY
For a general description of geology and terrain we refer to paragraphs 3.4.1/3.4.2. The research area is dominated by a river dune. Compared to the surrounding dunes this may be considered a small one. Its shape is indicative of an original parabolic dune. To the west the dune is low, increasing in altitude towards the east, with a top at 11.3 metres above NAP. The west and south slopes are gentle, whereas the north and east slopes are much steeper. The north slope appears to have been eroded, due to a residual Meuse channel. The east slope was not affected. The river dune is an island surrounded on all sides by a complex of residual Meuse channels.
3.5.2.1 Historical development
Fig. 3.29 The microregion Linden in (a) 1803-1807, (b) 1859, (c) 1916, (d) 1928, (e) 1967.
a b
c d
complex Groot Linden, with mainly fields and to the west pastures.
The Topographical and Military Map of the Kingdom of the Netherlands, of 1859 (fig. 3.29b), plainly demonstrates that this area has hardly been subjected to change. The number of buildings has barely increased and the land use has basically remained the same as well. A striking difference is that the land division has become much more intricate, but this may be the result of the quality of the field survey. Early in this century several civil engineering projects have been carried out concerning the Beerse Overlaat (fig. 3.29c).
On the map of 1928 (fig. 3.29d) it is clear that in the core region the drainage of the Sluisgraaf has been improved and its course has been straightened. The previously waste land Groot Linden Broek has come under cultivation and is used for pasture. The number of buildings in the area is still low, due to regular flooding by the Meuse. Only in Groot Linden can some growth be seen.
The situation outlined above continues for quite some time. After the abolition of the Beerse Overlaat the areas suffering most from floods are partitioned and used for pasture. The remaining area barely changes, either topographically or in
1 2 3 4
land use, as does the pattern of roads. Only from the sixties onwards do some changes occur, but initially only within the traditional framework outlined above (fig. 3.29e). The roads are improved and if necessary widened to some degree, outside the old residential nuclei some farms are built, part of the pastures is transformed into fields and the village of Groot Linden gradually expands. The major influence on the area are the mining operations starting to the north of the village of Groot Linden in the seventies. Initially these are small-scale along the Meuse, but gradually the mine lakes increase. Hooge Voort, Kraaienberg, part of Maaskampen, Ganzenorgel, Over de Voort and the western part of the original Groot Lindens Broek successively disappear into the dredging machines. As we write, the area around Hoge Renning has disappeared as well and only the highest part of Hoge Renning is still clear of the water, like an atoll in the Pacific. Plans exist to extend the mining operations over the next few years to the east in the direction of Op den Drul and to the south between Grote Geest and Dommelsvoort as far as the secondary road Grave-Cuyk.
3.5.2.2 Palaeogeography
It is far from simple to reconstruct the Neolithic terrain. Although the physical terrain and its genesis have become known to a reasonable degree, there is not enough detailed pollen research to allow a reconstruction of the vegetation. In point of fact, we have to transpose data from elsewhere to this area to allow a rough reconstruction. The basis for the geographical reconstruction was provided by the paleogeographical map of the Boreal. As there were hardly any geographical changes after the Boreal, the situation is almost identical for the Atlantic. The map shows a river district north of a coversand area, with heights, the river dunes, as separate elements or continuous dune complexes. In between are older relicts like infilled channels and backlands of the Meuse. The river itself lies to the north, in the bed of the Katwijkse Meuse, where it still is today. A C14-sample of a peat layer at the base of a residual channel at the Sluisgraaf has yielded an age of 8830 ± 55 BP (GrN 17778). The area therefore appears to have been at rest since the Boreal, and only initially did peat formation occur in the deeper parts of the residual channels. We do not possess a pollen diagram or a C14-sample of the top of the peat, as a result of which no statements can be made about the end of peat formation. This has barely developed laterally between the residual channels, since no peat is found between the channels and the heights. If the peat has not been subject to erosion here, it would seem these areas were not extremely wet.
For a regional reconstruction of the vegetation we have to turn to a number of pollen diagrams from the immediate
vicinity. The sample that is closest and which covers the Allerød up to the Subatlantic without any gaps, is located at Dommelsvoort, in the southeastern corner of the research
area66. Two samples providing additional data are located in
the valley of the Raam, the first in Esterbroek, south of
Escharen, the second at Zittert, southwest of Grave67
. Other locations are further away, outside the core region.
From these data a rough picture can be inferred that Neolithic vegetation was characterized by relatively dense forest on the higher grounds: coversands and large river dunes (fig. 3.30). This forest consisted for the most part of oak, elm and lime, with hazel and ash in relatively open spots. These trees will have been predominant on the smaller river dunes as well, whereas the vegetation in the backlands, on clayey soils, will have consisted mostly of open mixed deciduous forest (riverine forest) with ash, elm, oak and hazel. The presence of hazel suggests there was a highly developed undergrowth as well. The high number of Alnus pollen are an indication that in the wetter parts of the area a alder carr will have existed, composed of alder and willow, with an undergrowth of fern.
Not much is known about the degree of wetness. Peat formation in the residual channels in the area suggests the possibility of a more open, but highly local, terrain distinguished by a marshy vegetation. The alder carr mentioned above might fit this in part, but a slightly more open vegetation dominated by reed and sedge is possible as well. This would however have been of limited size. For the reconstruction we have therefore assumed that the old channel system was mainly covered with alder, in the form of marsh forest, interspersed with small open zones with reed and sedge.
The terrain is therefore highly diverse, with relatively open and dense zones and dry and wet areas. This is the type of terrain that will have accommodated a wide range of animals. Species native to this kind of terrain and known
from excavations68include aurochs, brown bear, red deer,
been dated around 4910 ± 35 BP69. Plantago lanceolata first
appears in the eastern river district in this period as well and is indicative of human influence in the landscape around 3900 BC (cal), culturally defined as the Hazendonk 3-phase. Clearer signs of human influence are apparent in the pollen
diagram of Zittert70. This demonstrates a decline in the
forest, associated with Cerealia and Plantago pollen and with fluctuating percentages of Corylus and Betula, in a non-C14-dated pollen diagram, attributed to the Atlantic on the basis of pollen analysis.
It is questionable whether the influences that may be inferred from the pollen diagrams, may be attributed to agriculture. The percentage of cereal pollen in these diagrams is small. This may be explained by the fact that fields were not in the vicinity of these sample points. Cereal pollen is almost exclusively distinguishable in pollen
spectra close by settlements and arable fields71
. However, no samples of the settlement locations are present. Yet it should not be ruled out that livestock was a major factor in
the food economy. The data from Ewijk72-albeit
younger-0 500m
> 120 cm
0-120 cm gully
0
37 38 39 34 21 35 20 40 41 24 42 23 22 2 18 43 44 45 46 19 47 48 55 56 54 52 53 51 50 29 30 60 59 58 10 28 26 27 25 12 32 33 9 11 16 7 1 17 8 15 13 14 3 4 5 6 57 31 arable fields meadows stray finds exact location a b
clearly demonstrate this. A similar activity can be distinguished by the occurrence and increase of Plantago lanceolata.
3.6 The pilot study
3.6.1 INTRODUCTION
The local geology greatly influenced the research
opportunities. In particular the presence and thickness of the recent clay deposits of the Beerse Overlaat determined to a high degree the research strategy in this microregion. There is only a small area where this clay does not occur at the surface or in the plough soil (fig. 3.31). This zone may therefore be explored by surveying the fields. As mentioned before, this is a method of exploration employed only rarely by amateur archaeologists. Before the start of our
investigation the number of finds from this area was therefore small.
In the autumn of 1989 a single attempt was made to find out what the result of surveying the fields would be. At this stage it was only known that archaeological finds had been made on the highest part of Hoge Renning. Opinions were divided about the conclusions that should be drawn from these, ranging from Iron Age and/or Roman habitation to the presence of a medieval monastery.
In 5 metre-wide strips all fields surrounding Hoge Renning were surveyed (fig. 3.32). Results and research opportunities of this exploration have been reproduced in figure 3.32b. Thirty sites have been pinpointed, in almost all cases these were single finds of a single individual object. Indications for settlements were not present. The majority of the finds
0 500m
a 0 100 500m 0 25m 0 1m 0 1 2 3 4 5 0 1 2 3 4 5 6 7 8 b c
a b 1 7 4 10 19 18 20 12 13 11 24 2 23 22 25 16 26 5 3 17 9 8 6 14 15 27 21a 1 7 4 10 19 18 20 12 13 11 24 2 23 22 25 16 26 5 3 17 9 8 6 14 15 27 21a large drill
small drill Fig. 3.35 Distribution of all
comes from an area around the top of the river dune. The top itself, the exact location of surface finds in the past, could not be investigated, as it was used for pasture. The results of this part of the investigation have been included in the results of another part: the pilot drilling study.
3.6.2 Research method
Surveying the fields is clearly not a suitable method for finding sites in this area, hence an alternative strategy was developed. The top of the prehistoric surface lies below the recent clay deposits of the Beerse Overlaat. Two methods of
1 7 4 10 19 18 20 12 13 11 24 2 23 22 25 16 5 3 17 9 8 6 14 15 27 21a 26 a 1 7 4 10 19 18 20 12 13 11 24 2 23 22 25 16 5 3 17 9 8 6 14 15 27 21a 26 1 7 4 10 19 18 20 12 13 11 24 2 23 22 25 16 5 3 17 9 8 6 14 15 27 21a 26 1 7 4 10 19 18 20 12 13 11 24 2 23 22 25 16 5 3 17 9 8 6 14 15 27 21a 26 1 7 4 10 19 18 20 12 13 11 24 2 23 22 25 16 5 3 17 9 8 6 14 15 27 21a 26 b c d e
exploration may uncover sites in this kind of situation: digging test pits and drilling.
Mechanical digging of test pits is a research method often employed in order to gain a quick insight into the presence
of archaeological sites73
. For several reasons, this method is undesirable. First of all, it is not known where the sites are and how large they are. There is a ready chance of missing a large number of sites in digging the test pits. Another problem is the severe disturbance created in excavating the test pits. The finds may be collected at the dump heap, without having been measured in three dimensions. Only the profile and the finds it contains may be studied. In order to reduce these drawbacks as much as possible, a proportionate amount of extra work is required. Mechanical digging is less desirable then and more should be done by hand. This leads us to the third and often most important reason: expenses. Mechanical digging by itself is not inexpensive, but becomes considerably more expensive when this should occur on farmland still under cultivation at the time of the
investigation. The crop damage is extensive, as is the compensation that has to be paid. This also increases the chance that farmers are unwilling to cooperate.
In order to meet all these drawbacks, a new research strategy has been developed. The essence is to map as many sites as possible with a minimum of disturbance, as little expenditure as possible and without inconveniencing the farmers. In this research method four stages may be distinguished.
The first stage consists of the complete drilling of the entire research area, provided the prehistoric surface level is no more than 1 metre below the recent clay cover of the Beerse
Overlaat74. With the aid of an Edelman drill (ƒ 10 cm) soil
samples of approx. 1 litre were collected from the top of the prehistoric surface, in a grid of 25 by 25 m. Overall, 1858 holes were drilled (fig. 3.33). The samples were sieved on a sieve with 1 mm mesh. The residu was searched for the presence of small pieces of flint, pottery, bricks, charcoal and bone.
The results of this stage of the investigation may be translated into distribution maps with relative densities or presence/absence. Materials strongly susceptible to
pulverization, like charcoal, brick and bone, are indicated by presence or absence, flint and pottery in relative numbers. As the basis for the definition of a site the number of artefacts per drill residu was taken. This should amount to at least three pieces of flint and/or pottery. If drilling in the immediate vicinity yielded material as well, this was included in the site, even if there were less than 3 items. An additional criterion was the occurrence of surface finds at the site. The results of site 9 may be taken as example (fig. 3.34).
Overall, 27 sites could be defined (fig. 3.35a). However, this drilling method does not allow any statements on the size and date of the site. For this purpose a second stage of investigation was devised.
In the second stage the demarcated sites are drilled with a larger Edelman drill (diameter 15/20 cm) in a grid with finer meshes (12.5 by 12.5 m). This allows a larger sample of soil to be collected from the top of the sand. The different diameters were selected because during the investigation it became apparent that it was almost impossible to have a drill with a 20 cm-diameter penetrate the heavy clay for more than 80 cm manually. In cases where the sand was at a greater depth than 80 cm, we therefore employed a smaller drill. In this way 19 sites have been investigated, performing 265 drillings (fig. 3.35b).
The soil samples have also been sieved on a 1 mm-sieve. This usually allows a multiple of archaeological material to be collected. In this way relatively detailed distribution maps of the find spread can be made (fig. 3.36). In a number of cases this method also turned out to provide additional data, allowing a determination of the age of a site.
site indet. Stone Age Iron Age Roman Middle Ages
01 x x 02 x x x 03 x 04 x x x 05 x x 06 x 07 x 08 x 09 x x 10 x x x 11 x x x x 12 x x 13 x x x 14 x 15 x 16 x 17 x 18 x x 19 x 20 x x x x 21(A) x 22 x 23 x 24 x x x 25 x 26 x x 27 (21B) x 27 4 20 11 4 10
In the third stage of the investigation a test pit is dug in the part demonstrating the highest density of a site. The aim is to gain insight into the soil constitution, determine the degree of disturbance, find out whether any soil traces are present and try and collect enough material to specify the date. In this way 6 sites have been tested. The covering clay was removed mechan-ically and the pit was shovelled by hand to a depth of approx. 15 cm below the level of the last recorded find. The finds were collected in squares of 25 by 25 cm. The entire find layer of one square metre was excavated in squares of 25 by 25 cm and sieved on a 3 mm-sieve. Any soil traces, if present, at that level were sampled, cut and recorded in drawings and photographs. Once all these stages were completed and sites appeared suitable for further investigation, the fourth stage occurred: the actual excavation of a site. This occurred in one instance, which will be described in the next paragraph.
drill drill test pit excavation 10 cm, 15/20 cm grid 25m grid 12.5m stage 1 find sites x stage 2 demarcate sites x stage 3 determine quality sites x stage 4 determine nature size activities x Table 3.4 3.6.3 RESEARCH RESULTS
The results of the drill study, digging of test pits and
surveying the fields have been summarized in table 3.575.
Twenty-seven locations are concerned, representing one or more sites. By site we mean an activity that can be chronologically specified using the modest means at our disposal at this level of the investigation. The total number of sites is 50; these can be subdivided chronologically as summarized in table 3.5 and the following maps (fig. 3.36a-e). The contents of a sieve residu can easily be determined, with the exception of samples from the gravel-rich fluvial deposits. In these samples it is often difficult to distinguish between ‘flake’ material caused by natural processes and flint processed by man. As a distinguishing feature we have chosen the presence or absence of patina on the flint. Flint with patina was taken to be natural, whereas flint without patina was considered to be anthropogenic. Another problem
present in samples from gravel-rich fluvial deposits is that during collection of the sample pieces of natural flint are struck by the metal Edelman drill. This causes the flint to crack and the fragments to display a striking similarity to flint waste created anthropogenically. We overcame this problem by recording these occurrences during drilling in the field on the one hand, and on the other hand by considering samples with large amounts of identical flint and a very fresh appearance to have been damaged by drilling.
3.6.4 INTERPRETATION
The emphasis in interpretation will be on the Stone Age sites. A problem in the investigation of the drilling results is the frequent absence of enough dateable material.
Concerning the sites where no test pits were dug, we may only state that these can be dated Stone Age. Two of the 6 sites where a test pit was dug date from the Middle Neolithic, to wit MK-culture.
The distribution of the sites displays a rather uniform pattern and a strong correlation with the higher parts of the river dune. Some sites have been found in the lower area. It is difficult to determine whether these are the result of human activity, we cannot exclude the possibility that these are the reflection of unidentified ‘flake’ material caused by natural processes or unrecorded damage to flint during drilling. The number of finds per site is limited and the size of the sites is modest. The largest site lies on the highest part of the dune, but size is mainly determined by the spread of the Iron Age pottery. On the basis of the drill results, no statements can be made on the function of the sites at this stage.
3.6.5 CONCLUSIONS
The drill method is a good way to trace sites. Additional investigation by means of drilling with a larger diameter and in a denser grid provides data on the size of the site, but information on the age and function of the site can hardly or not at all be derived. The digging of test pits provides an opportunity to collect more data on soil constitution and age, but in that case the function of the site often remains shrouded in mystery. Therefore an excavation will have to provide a decisive answer to the meaning of these flint spreads, hence we selected one site for further investigation. The results of this investigation may be correlated to the results of the drilling study. The correlation might lead to an interpretation of the flint spreads for the entire area.
3.7 Excavation Linden-De Geest
3.7.1 INTRODUCTION
In the period 11 March to 3 May 1991 one of the Middle
Neolithic settlement areas, number 9, has been excavated76
0 10m Linden- De Geest 1990/1991
Fig. 3.37 Site 9 with 10 m-sample grid and relative density of flint finds. Maximum number of flints is 8.
100 105 110 120 125 130 135 140 145 150 155 160 165 testpit 1991 excavated area sample pits
The site was on the south flank of the small Late Glacial river dune (fig. 3.34). In the discussion of this excavation we shall also refer to a small pilot study conducted there in 1989. We assume that the traces found here, at a distance of approx. 30 metres from the 1991 excavation, are part of the overall settlement area. The exact position in relation to the 1991 excavation is hard to ascertain, as in the autumn all reference points had disappeared, as a result of digging up and dredging off the western part of the river dune. On the basis of the topographical data left we performed a site location as best we could.
The decision to excavate this particular site may come as a surprise, since there was a site with far more finds on the highest part of the river dune. Except for the fact that this terrain could not be investigated at that time due to an uncooperative tenant, we assumed that the degree of disturbance of the area first used in the Stone Age caused by subsequent Iron Age, Roman and medieval habitation activities would be so extensive that an investigation did not appear very worthwhile. This disturbance would be far smaller on site 9. The small size of the site appeared to be a favourable circumstance as well. Large sites are often the result of a series of habitation and use activities which are often quite hard to unravel in analysing the excavation data. These considerations led us to select site 9.
Preparatory to the excavation we drilled the entire area in a 10 m-grid, in order to determine its size. Overall, an area of
approx. 3600 m2was investigated. The greatest find density
occurred in the high western part of the area. The relative flint density (fig. 3.37) appeared to indicate a relatively small site with dimensions of approx. 20 by 20 metres. In the central part of the area under investigation some flint was found as well, but this was located in an elongated strip, more or less on the edge of the dune. During drilling pottery was only rarely encountered.
3.7.2 EXCAVATION METHODOLOGY
After the topsoil had been mechanically removed in two stages, a 96 metres long and 4 metres wide pit was dug over the presumptive core of habitation activity, in order to determine the eastern boundary of the site. The western boundary could no longer be determined, as this part had already been quarried for sand. Subsequently the investigation was extended to the north and south.
The procedure was to excavate a small, 25 by 25 cm square of each square metre, to a depth of 30 cm. The excavated soil was sieved on a 3 mm-sieve. In this way an area of
approx. 1150 m2
could be investigated. Subsequently in the areas with the greatest find densities the ‘culture layer’ was excavated manually, with a shovel. The aim was to collect testpit 1990 48 46 47 44 43 42 11 12 33 41 1 2 6 10 50 40 7 8 24 49 36 16 37 23 25 3 14 15 22 21 4 32 27 28 29 30 45 39 38 51 34 35 19 20 18 26 9 31 8 6 5 7 10 1 2 3 4 recent prehistoric indefinite 0 5m
34 14 51 1 12 1 (t) 6 (t) 0 50cm 34 51 1 (t) 6 6 (t) 0 26 31 33 17 25 16 38 21 45 35 32 36 2 (t) 3 (t) 11 0 50cm
