The MA4 locations in the Belgian Kempen area

(Words, marked in red are linked at the glossary of this website, opening in a new frame)  
Large quartzite plano-convex core from the MA4 location, belonging to the
oldest European cultures  

The  MA- locations
In 2007  a focused prospection started  for possible human (activity) traces from the pre- Saalian periods ( < 200.000 BP) in the Maas (=  Meuse) region south of Maastricht (Netherlands) ( location in Google maps) , to see if traces about these period could be found on this almost unknown period.
Surface finds, without any other stratification than a description like  'found at the upper terrace  of..."will give information, but maybe it was possible to find more...

Mainly Middle -Palaeolithic
In the cretaceous area around the Dutch city of Maastricht, 60 locations were visited. This lead, as almost expected,  to several possible (Early -) Middle Palaeolithic finds / sites at various locations, especially in  the region south of Maastricht, between Maastricht and the Belgian city of Liège.( see : Miscellaneous artifacts at this weblog)   Such,  often isolated surface finds from middle and high terraces of the Maas/ Meuse - river, only indicate a possible human presence in the earlier phases of the Middle Palaeolithic, but do not tell us more than that.
The information in this article is preliminary information, interesting enough to be published and discussed, but  the finds from the Belgian Kempen for this period are only discussed and validated by some archaeologists from France
The Belgian Kempen area
The Belgian Kempen area (in yellow)The investigated Pleistocene riverbank of the Maas is over 20 km long and in this project divided into 7 sections from MA1 to MA7.

The western Maas/Meuse- river area north of Maastricht (NL) from Lanaken  (B) in the south till Maaseik (B) in the north, is a large sandy region, covered with eolian fine grained sands (mostly originating to the late Saalian and the Weichselian periods), where Early Palaeolithic finds are scarce (1),  but  the (Middle) Palaeolithic, the Upper Palaeolithic, the Mesolithic and the Neolithic periods on the other hand are abundant.( An overview of the Palaeolithic of the region  at this weblog:Regional Limburg(NL) The Palaeolithic ) In the image left, brown lines indicating the boundaries of the plateau, the red square representing the investigated area. Several articles are handling  mainly  the MA4 section .
The highest points of the region can be found at southwest of Zutendaal (+ 103 m a.s.l.) and to the north the altitude is declining at only + 40 m. a.s.l.. The coordinates of the center of this investigated region is centralised at 51°00′N 5°40′E.

The former, Middle- Pleistocene Maas/Meuse channel edges ( = main river banks) (2) were build up  by thick layers of gravels during the B- Cromerian Complex ( stadial), ( Paulissen, 1974)) between the Belgian communities of Lanaken ( Pietersheim) and Maasmechelen- Opoeteren, which layers are covered by thick layers of sands (= cover-sands) during later stadial periods ( especially the Saalian glacial).
North of Maasmechelen, much  river terraces are gone, destroyed by incisions and other former dynamic river processes ( flush, erosion, Saalian- landscape equalizing).  
Eroded surfaces containing  possible prehistoric traces outside the plowed, reworked arable areas, almost only can be found after e.g. forest clearing activities, where big 'wounds in the landscape' are  leaving exposed surfaces at the terrace edges. Such temporary exposed situations demand quick response during investigations, so preparations for the study objects are essential.
The sandy Belgian Kempen area is very interesting in geology, because of the presence of (former) sand and gravel pits, of which some are given back to nature, demonstrating geologic walls; large forests, small villages combined with good  facilities ( walking tracks, cycle paths, restaurants, etc) makes this area a popular place to stay. Image: Valley of the Kikbeek, an old sand pit near Opgrimbie. 

The discovery

In august 2011, after 4 years of searching, a main discovery has been made in the sandy area of the Kempen area in Belgium. The site location , named MA4 ( e.g. MA= Maas, 4 = map-region 4) is sub -divided into 6 different site locations, based on similar finds at eroded surfaces at exactly the same altitude.

The artifacts, found at this site complex, are completely different in type or shape compared to  the previous finds of surface scatters at the Maas high terraces. 
Not only are large numbers of artifacts forming logical lithic assemblages, found at single find-spots, possibly clustered with others, eroded out of stratigraphical known layers, i.c. formed by depositional gravels from the Maas- river, but also the assemblages do have a direct link with gravel horizons in the local stratigraphy.  
The Maas- river sequence is one of the worlds best studied and completed 'time-lines' in the world, so correlation with the geological framework ( and thus the chronostratigraphy) is basically possible (  Rosenbaum, Turner 2003, Westaway et al., 2002).
In the past months, over 4000 artifacts are collected, samples were taken and pictures and drawings were made of site circumstances.

The so called 'pebble tools'  usually are not regarded as the most nice artifacts to collect and often they look like randomly broken stones. In reality such  stones are representing very old artifacts, belonging to the same tool technology stages like  old  tools found at  the Olduvai Gorge site in Kenya ( image here)  excavated by  Mr. L.S.B. Leakey, 1951, the Oldowan- tool technology.


Finding artifacts in an area of roughly 200 km 2,  where up to 15 meters of gravel has been deposited in a wide fan gravel "estuarine" of the Maas/ Meuse river is not easy.
99,9 % of the surface is covered by woodland, roads, cities and villages, very large amounts of scattered buildings..., industries, sand pits,railways,  private domains, heath-lands,  arable lands...
So the places to search for artifacts are very limited. Since it's nearly unlikely to find bones from early hominids in the region, due to bad geological conditions ( especially in the sandy northern part), the only remains to prove their existence are the artifacts, left somewhere in the landscape.
Topographical maps have been  compared with the fields, forests, etc. , elevations lines and possible eroded surfaces are studied. In the field, often good looking places seems to be inaccessible either by prohibited access, infrastructural works, or simply by the absence of any ( quality, unmixed) exposed eroding surface, of which you depend.  Few "disturbed" area's were found, where gravels from other locations were imported, transported and dumped, always very well visible by the deviant gravel types  compared with the local gravel depositions of the river.
So within this 1 % of possible usable locations,where artifacts could be found, there are only very few locations where artifacts really could be established. So, it is difficult, even to locate  few locations that are "good enough" to expose the right 'artifact bearing' levels.
Pebble tools mainly are traced by fieldwalking of terrace landform surfaces i.c. in a forested area like the Kempen area:. like after forest clearing activities, infrastructural activities ( draining activities of brooklands, clearing heath-lands, digging, etc.). Besides of useless contexts, as  pebbles could be naturally broken, this alone is not the way  locating a real Paleolithic location / site. For locating a real site, it is necessary to find established man- made tools, fitting in the local stratigraphical position ( analogous gravel types, analogous alteration, etc.) and with a remarkable deviation in presence ( concentrations).  So in reality we cannot always speak in terms like "surface- finds", as lots of artifacts were taken from an exposed part of any  eroding gravel layer, where artifacts still were found clamped in situ in i.e. the dark yellow loam layer of a late Elsterian or even a Cromerian deposition of the Maas/ Meuse river. Both this stratigraphical position in which artifacts were found, as well as the locations on the upper terrace of the river, brought  'reasonable' outcomes for the periods when the artifacts were produced  and used.

Large scale clearing of heath land, like here at Zonhoven- Molenheide brings back original cover-sand layers that might show some erosion after time. Such parts of the forest lands could bring new information about prehistoric archaeology.

Why are such large pebble tool collections not found before?

In the first place the Kempen - pebble tools do not come to us as very nice, easy recognizable (Acheulian) artifacts, that are easy to recognize. Moreover, tools are  processed at minimum effort and at first sight, modifications are not noticed easily. Such tools often are abjected by amateur- archaeologists. To recognize such tools and distinct them from geofacts, one should see lots of such artifacts and search at exactly the right locations - it took more than 3 years to find them: the limited number of sites makes it very hard to detect them in a forested area....
Why search at the surface and in possibly disturbed contexts?
The answer is simple: this is the only way to detect more about this period. Usually excavations do not take place in such forested areas, and gravel depositions are not part of usual excavations.

In situ

Several artifacts at MA4  were found  in situ. The eroded surfaces however could show a ( local) secondary context caused by geological processes and recent changes in soil structures by human activities. On the other hand, tools were found in such big concentrations,( estimated >500 /m2 for a local 4 m2= nearly 2000) and do not show any abrasion or other signs of fluvial or post- depositional transport, so it's unlikely they have been moved over a long distance. 
In this way, the find spots would even indicate the true locations where the artifacts were produced, what is a natural situation in case of pebble tools: produced,  used and discarded, and contemporaneous integrated into fluvial depositions, preventing weathering influences for ( the majority of) the artifacts.
Another indication pointing at local use in one " limited" period ( but probably during different phases within that period) is the fact, all locations within the site- complex are discovered at  about the same level, over an estimated area of ca. 200 x 500 m (= 1 km2). This could be the indication for the existence of a large site, which would be logical,  regarding the large amounts of artifacts at the different find-spots within the site cluster.The large amounts of artifacts would imply a habitation zone, as it is unlikely to make so much artifacts without staying nearby.

Local stratigraphy at the Upper Kempen Plateau Terrace, based on visible stratigraphy.  
Complex stratigraphy

The dominant factors for the in situ stratigraphy of several observed artifacts are the possible correlations with embedded gravel layers combined with the fact the entire site- cluster is positioned at the same altitude of the Maas/Meuse upper- terrace.
The top gravel horizons however, are heavily reworked during two different known periods: first after the fan gravel- depositions ( final -? Elsterian) and during another time during the Saalian II-III.(3)
In this way, artifacts originating from these parts of the gravel deposits, could have been moved upwards in the sandy cover-sands. Indeed, in the stratigraphy we notice a certain mixture of eolian ( very fine grained) sands, mixed with relative big granulates, i.c. pebbles, as shown in the picture above at level C2.  A "transitional horizon" would occur between the typical Cromerian - dated Zutendaal fluvial gravel depositions (MIS 13, early MIS12(?)) of the upper horizons and the Saalian dated eolian depositions, that originally were deposited above these gravels in younger periods (MIS 7/8).
The artifacts are likely to have been used after the first ( inverse) erosion of the Kempen fan gravels, ( towards the east), so after the Elsterian period. Indeed, there has been observed a relation with ( large?) high- energy deposits of the Pleistocene Maas /Meuse river, as these coarse sands are different from the eolian sands, by the bigger sand granulates dimensions.
So these more coarse aggrading clastic gravels, were deposed during a glacial period (i.c. most probably the Elsterian) and  when the Maas /Meuse river  had begun to down- cut itself into the landscape, during a post- Elsterian date, roughly after 450.000 BP, early hominids would have used stones from these gravels(4). This original aggrading process on the Maas /Meuse  channel might be the reason for a swaying, capricious channel edge just before site use, leaving a wide 'beach -like' channel edge, where gravels were exposed at the surface.
The same ( present) altitude for the whole site -cluster ( over a distance of 400 meters) is suggesting a contemporaneous use of the sites, in which case the gravels over this 400 -500  meters must have been simultaneously accessible for tool-production, which is commonly reported for sites with large good quality clasts depositions (5)
At locus MA4-2, a rather deep incision ( about 5-8 m deep) has been formed after site use: the currently divided MA4 -2-N and MA4-2- Z parts, show similar tool types, see the other article on this weblog Tool types and raw material of Acheulian -like tools in the Belgian Kempen area , indicating the site was not located at e.g. two sides of a former brook, forming an undivided 'activity zone'(6)
In the original stratigraphy the artifacts possibly were in a pre - Saalian I  ( before or at first stages of the  formation of the middle terraces) stratigraphical position between C2 and D in the image above: this would place them in a rather wide range between 450.000 BP - 300.000 BP,  (MIS 7a - MIS 12). (7)
Further investigations on the in situ artifacts is necessary to specify the stratigraphical position in more detail, if possible. (8)

Position of MA4 at the former Maas riverbank, during a post Elsterian  and pre- Saalian period. The Cromerian fan gravels were deposed by the Maas river and were eroded during glacial conditions of the Elsterian. Hominids used the site during/ after the forming of the As Paleosol.
Kempen fan gravels as mentioned in the text above, visible as large different horizons in a pit near As. The Manganese dioxide horizon is visible as a black horizontal layer. In situ finds from site MA6 are found on the edge of the overlaying red soils and the more light colored horizon.
Simplified cross- section of the situation in the sections MA2, MA3, MA4 and MA5. At the left  the Maas/ Meuse upper- terrace, with an eroding terrace edge;in the  middle we find the 'middle -terrace' ,the Eisden- Lanklaar terrace, remains of the alluvial plain during the Saalian, and right the current alluvial plain. At the MA1 section more terraces can be distinguished.

Geofacts- artefacts
Surface finds without any exact stratigraphy always are suspect. At first, they must be distinguished from geofacts and moreover they could originate from every possible period. This is why many archaeologists do not (like to) value them: the case is much more difficult and complex than finds from well dated horizons and the outcome is more limited, so less interesting for the archaeological record of an area. On the other hand, the lack of well preserved early Palaeolithic sites in the region brings, such 'minor 'cases, like assemblages from 'surface finds' , is all we have. Moreover, before these discoveries ( by the author ) no- one ever had thought such large scale European Oldowan cultures do exist in the wider region.
Man-made artifacts are different from geofacts by special features, that deviate from each other. Unifacial flaking patterns are very rare in trampling experiments ( as mentioned by Mortensen, 2008), and trampling and other accidental events usually are causing "random flaking". Besides of the regular flaking patterns, patinas, concentrations of pebble tool artifacts - so the features of the tools itself - there is stratigraphical evidence, derived from nearby local stratigraphies, especially from the numerous sand- pits in the region.
Flint - stones and pebbles, especially from the Late Elsterian gravels, sometimes have a man-made appearance. They must not be confused with prehistoric artifacts, though the shape of these geofacts often reminds more the later tools, made by men.
Especially flint from the Elsterian deposition has suffered a lot from frost . Notice  the abrasion and the lack of percussion rings. When such ripples occur, they are concentric, but irregular, fleeting of a visible navel point 
It is emphasized that the discussion about these artifacts is still open, till more experiments are made and more clearness is received about other possible semi- natural circumstances ( like trampling by large animals, see e.g. Neal H. Lopinot and Jack H. Ray Trampling Experiments in the Search for the Earliest Americans
Artifacts were published ( with images) at the prehistoric forum of Lorraine ( where such tools do occur) and recognized by others. Moreover, the artifacts were discussed and recognized as man- made tools in controlled flaking techniques,  and artifacts were reproduced  in some limited experiments to establish
  1. The assumed man-made artifacts show features that occur during experiments showing logic basic concepts for tool processing
  2. These features are suitable for appropriate tools with small cutting edges, but could very well have served as 'intermediary' tools for the production of wooden tools. weaponry (like spears, harpoons)
  1. Comparative tool forms are reported from the French Moselle region ( especially flakes) and from other North-West European sites, in Germany and England
  2. Further experimental production of quartzite and quartz tools will possibly give new perspectives 
  3. This, while such pebble tools only can be found in context of disturbing surface activities ( digging, forest clearing), so real undisturbed contexts are not possible to find for this group of artifacts. To rule out trampling, pebble damaging, crushing by machines, etc. further  experiments  are necessary. 
Special finds in the Benelux (Belgium/ Netherlands)(?)

The tools from the site complex can be regarded as very special, even when - for the moment- the latrge assemblages must be regarded as a very large surface- scatter. Large scale quartzite tools in a pebble tool- tradition are not common in the Lower countries. An exception is the hilly region of Luxembourg  where pebble tools are found, see http://www.prehistory.lu/vo_txt2.htm  See also another article at this weblog, a brief overview about pebble tools  found in The Netherlands and Belgium : pebble tool artefacts  in The Netherlands and Belgium (Arbannig). Notwithstanding the absence of a clear chronostratigraphy for the artifacts,  such large assemblies, if well documented, obtain much information.
Besides, the chronostratigraphy for these artifacts is estimated between 550.000 BP and 350.000 /330.000 BP, based on visible stratigraphy , etc. ( see below).
Stone tool-kits of the earliest Europeans solely consist of pebble tools and crude flake tools (9)

Very early occupations ( Pre- Neanderthal) of Belgium and The Netherlands (?)

These tools would be the markers of the arrival of one of very early migrating groups of tool- makers, and indicate the presence of very early hominids in the Lower countries (NL/B). Though pebble tools were used in different other, later periods, the combinations of find- spots ( also in situ) i.c. in the high river-terraces, the total absence of any imported tools, the heterogeneity of  the assemblage, the lack of artifacts from identifiable commonly known periods ( such as the Mousterian tool technology, the Upper Paleolithic, the Neolithic , etc.) and the tremendous similarities with the Mode 1 / Developed Oldowan tool assemblages, give a full impression of dealing with tools of Homo erectus / Homo heidelbergensis. Added to this there is the taphonomy of the artifacts, which would indicate a very  high age

In the month November 2011 over 4000 artifacts from one site complex and two different other sites were collected, showing big homogeneity in raw material choice, controlled flaking techniques, tool end products. Even with a known stratigraphy and in situ finds the period of these early occupations remains unclear, but will certainly have a pre- Saalian date, and even more likely   a Holsteinian date. But to be sure about the artifact's age, datable materials like carbon or fossilized animal remains must occur  in the stratigraphy . So far these materials are fully absent, but the search is still going on.

The sites in the palaeo - environments of Lower Palaeolithic/ Middle Pleistocene 

Finds of archaeological evidence at former ( main ) river channel edges of a Pleistocene river will have a direct link with the benefits of such palaeo environment, providing humans with large animal accumulations, especially nearby ( former ) attributes and lakes/  ponds. However, remains of ( large) Pleistocene fauna from the Kempen Plateau seems to be totally absent in the archaeozoological record.
This could  partially be caused by the bad circumstances in the acid  sands ( especially in the northern regions of the Kempen-plateau), where bones dissolve quickly, on the other hand the hydrology of the environment had changed over long periods, causing changing water-levels. Present drainage paths were formed during the  (late) Weichselian period and were especially influenced by changing vegetation covers, by land -use during late Holocene periods,especially the Middle Ages.
The Maas /Meuse river must have been a wide, longitudinal meandering river during post- Elsterian interglacial conditions ( i.c. MIS 11) , with low energy transport. Evidence for this slowly meandering  river type could be derived from the altitude of comparable finds ( small tool production) between the MA1 and the MA7 section, with an over all decline of ca. 12 m/ 20 km (0.85 /km; tertiary: 1,5 m/km; present: 1m/km)
This river type would be rather small ( up to several km's?) with changing gravel banks, depending on river dynamics directed by climatological conditions. Like in the present, small beaches occur in low height gradients and the cut- off channels were turned into larger swampy areas.  This type of environment shows a large diversity of biotopes with a large diversity of life, providing a wide spectrum of food supplies for humans, both vegetarian and animal: herbs, berries,plant-roots, fruits,seeds and nuts as well as meat from larger animals, such as horses, deer, badger and  boar and smaller animals like serpents, frogs,turtles,freshwater - mussels, fish, etc. This fits into the climate type of the long Holsteinian period ( over 40.000 years), with a variable warm climate with average temperatures more than 1- 2 degrees more than present; in the vegetation for the relating region we find Buxus sempervirens and some archaeological/ palaeolontological records from northern parts of The Netherlands speak of  the Hippopotamus amphibius bathing  in the North sea and at the Maasvlakte.
Such climatic conditions could explain the choice of early hominids for East- to South - East orientated sites, though MA4 could also be part of a real South orientated site.
During glacial conditions, the very large, up to over 20 km wide braided river system, cutting itself into the landscape,  causes many channels,  transporting large amounts of gravels and even very large ice raft blocks from the mountainous regions  Vosges and Ardennes. Gravel deposits could form inexhaustible sources of raw material for the production of ( pebble)  tools ( pebble clasts). Large animals crossing the river valleys, and at suitable crossing places (fords) people would get a chance for large scale hunting. 
The variety of blind river arms, river gravel-banks, small left over ponds, etc. with changing vegetation patterns must have presented a very rich fauna and flora in this region, satisfying the needs for early hominid migrating groups. 
The sites of MA4 / MA6 are located at current elevated, high points in the landscape and possibly they were similar  in the past, to keep dry feet and to have an overview on the landscape. This does not mean they coexisted in the same period, though this is most likely, based on comparable assemblages at the sites.
MA4 site cluster; size of triangle marks importance of site; black lines : current elevation lines of  + 85 m a.s.l. ; brown  line is  a possible Pleistocene riverbank during site use; mint green spots are current dry valleys ( erosional valleys)  formed during the Saalian 

Settlement: food and tools 

The Middle Pleistocene west bank of the Maas/Meuse river was an ideal place to stay during the post -Elsterian period.
Two important issues were decisive for the settlement in the Belgian Kempen area: the occurrence of large wildlife groups ( food), dwelling the alluvial plains  and the presence of  "good enough quality-" gravels along the Maas/ Meuse river, deposited during the Elsterian.
At first incisions into the present valley, the Maas/ Meuse river still had a very large alluvial plain, at Dilsen-Stokkem this plain would have spanned  about 20 - 25  km to the east. 
Old river channels were cut of, and were overgrown with lush vegetation. The large braided river system changed under climatological circumstances into a slow meandering river; the main river channel or one of the river channel(s) flowed somewhere in the western part of the large alluvial plain starting to cut itself into the large gravel- layers from Cromerian depositions.  

The first hominids arriving must have faced a rather cold climate. At the highest parts of the Kempen plateau, only relative large cutting / cleaving  tools were found, implicating the use of these tools for butchering large animals. It is very well possible, these hominids settled on small raised areas, located directly nearby small peri- glacial lakes, high on the plateau  which lakes would be the base for the later sources of e.g. the Bosbeek, the Asbeek and the Ziepbeek. Still today, we see the higher parts of these valleys are dry, supposing they started at higher altitude during pre - Saalian periods as melt water channel, forming tributaries for the main channel(s).
During the cold (-er)  climate type, so during  the late Elsterian  (Noordbergum MIS 13?) till early Holsteinian ( transitional phase MIS 12- MIS 11) the Maas /Meuse river beach reached even over + 90 m, so was directly laying over the Tertiary horizons at highest levels (see: Paulissen 1974, and Quartaire Geologische Kaart Belgie, Rekem nr, 26).(1)
 Large tools from these locations were made by hard hammer technique, and tools show large cutting edges.
From the raised places (orientation unknown, but not specific oriented to the east, to the riverside) , they must have had a splendid wide view over the region, they looked into what is now Germany and  movements of the game could be followed over big distances. Tributaries from the east plateau of the river caused wider alluvial plains - in combination with large swampy areas , ( former ponds), of which the Dutch city of Meerssen is a good example  (Meerssen = House in the marsh, the marsh of the Geul- brook).
The deserted river channels on top of the plateau filled up slowly with sprayed sands, and were mixed with the late Elsterian gravels, today still visible at many places on  top of the Kempen plateau. Melting waters must have caused erosion at the plateau. leading to new sources, forming the fossil floors of the present  brooks.
Afterwards, the climate changed from a glacial period into the longest interglacial ever known. The vegetation must have changed drastically. Coniferous forests disappeared and were replaced by mixed deciduous forests.

Changing climate and associated  changing vegetation - response causes the necessity for new foraging strategies.
Early hominids in the Kempen region were faced with a large alluvial plain, with  river beaches where gravels were exposed.
The prove for the fact, tool production must have taken place on / nearby such beaches is simple: such small pebbles (<3 cm cross cut) for the fabrication of micro -choppers and micro- chopping tools, only can take place when the raw material is available and visible. At the beaches such small pebbles were exposed, uncovered, so visible. Another fact is, the present Maas/ Meuse valley did not exist yet, so these early hominids had to search their tools at the places where gravels  were exposed: on what now is the upper terrace of the Maas/ Meuse river and on riverbanks of the alluvial plain, but this plain has been destroyed completely, so the only remains of gravel-banks are the plateau's edges.

Inhabited Pleistocene channel edge
Two different site complexes with possible dimensions of 400 / 400 m (MA4)  and   300 / 400 m (MA1), makes it believable there were organized camps at the Pleistocene Maas /Meuse  channel edge.
Where did these tool -makers come from? Did they use the Belgian Kempen area as a seasonable camps? Are the camps, oriented east -wards to the river,  forming a statement for a warm climate? Are the small tools a response to environmental conditions with small pebble clasts, or is it a reflection of a wider cultural determinated tradition? The numbers of the artifacts (> 3000) could point at repeated habitation, or indicate a long - term habitation.

The study of this site cluster will take many months / several years. The good news is, it might be possible that parts of the site cluster MA4 would still be covered by the Late Pleistocene eolian sands, so parts of the sites, especially those parts  located  more on top of the plateau, could be " intact".
In November 2011 a first publication ( in French) is  published on the internet. This full report in French could be downloaded   by clicking  "Galets aménagés de la région Campine, la Belgique."  ( page = argeocol. hpage. com, A.S.A.C. geological -pages)

(Above) example of an eroding surface at MA4- 4 where artifacts are eroding out of a part of the Middle Pleistocene  gravel-bed at the upper terrace of the present Maas/ Meuse river
An example of a quartzite pebble tool ( unifacial chopper) found in the Kempen.
An example of a quartz tool; quartz is usually difficult to recognize by  the lack of  final characteristics 
Image above: tool types in the assemblages from different sites show  very little variation, the general aspects are about the same: chopping tools dominate, choppers occur and in the assemblage we find picks, side scrapers, "sliced pebbles", "split pebbles ", flakes, core tools and a small amount of flint tools.

Image: impact point is visible as a light spot on the edge. Also visible are the percussion lines( direction left), which are still visible  in the grain - patterns, bending away from the impact point.
This image shows three transverse strikes, showing a developed mode 2: tool makers were very efficient in making the tools.
This image shows the well placed flake removals to get a chopping tool

Large quartzite side scrapers are normal at MA6 and MA4

core tool


Erosion processes of the East Kempen plateau

After the large fan gravel deposits during the Cromerian, during the pre - Elsterian phase (< MIS14), an inverse process took place. The Elsterian phase is characterized by the first incisions of the Maas river by large aggrading gravel deposits and ice drafts. The post- Elsterian- interglacial period was rather warm and caused heavy oxidation of soil layers, still visible as intense red soil layers, integrated in the pre- depositional layers.
The Saalian period caused large, deep incisions, bringing large amounts of the Cromerian gravel depositions towards the east edge of the Kempen area. These incisions were partially filled afterwards by eolian sand depositions ( cover sands) and partially eroded on the slopes, depending on the particular area ( the plateau edges in some parts of the Kempen area are on the other hand either formed by large dunes, or heavily eroded in longitudinal valleys, caused by incisions that suffered from heavy erosion during glacial conditions). 
 The image above is a compilation made of images taken in the Sibelco S.B.S. pit, Maasmechelen. According to this uppermost stratigraphy ( left), the pebble tools from MA4 would originally be positioned in the gravels of horizons between 30 cm and 1 m. For MA6 the situation looks more like the right picture, showing a heavily red oxidation horizon directly beneath the reworked and remixed Saalian lower horizon stratigraphy. Local stratigraphies can vary tremendously.

Generalized model of the stratigraphy established in the MA1 and MA4 sections ( east = right). Cover-sands have eroded  and often are found  at the surface of lower terraces, ( contrary to the sections MA2, MA3, MA5 and MA7). The coarse sand horizon has been established in every section, sometimes up to one km from the current plateau edge, forming a possibly late- Elsterian aggrading deposition.Eolian remix with gravels took place in Saalian and Elsterian gravels, but would be caused during Saalian conditions. Integrated As paleosol  is changing a lot in local stratigraphy. Former terrace [ channel ] edges are more invisible in the lower terraces due to erosion and eolian cover sands, worst visible at places where falling dunes were formed ( MA3 section, MA5).The Quaternary / Tertiary border is depending on the section, at MA1 ca. 10 m.

In the images below, an impression of what happened at site MA4, during the periods from MIS 11 till the Holocene (MIS 1). (Click to enlarge)
The start is, at the end of the Elsterian period ( ca. 420.000 BP) the river made first incision in the present valley. At small elevations, like at MA4-5 humans/ hominids settled to make their tools, from the riverbed nearby, at an estimated level between + 80 m and + 85 m. This period would coincide with the Holsteinian, a warm interglacial with average annual  temperatures of +1- + 2 degrees higher than present. The Maas/ Meuse river was meandering and must have followed a direction north- northwest, the first change after the Elsterian.

Between the Holsteinian  ( 420.000 - 380.000 BP)  and the beginning of the Saalian Complex (300.000 BP) the river cut itself farther into the large gravel depositions, building new terraces.During this phase we distinguish short interglacial periods like Oostermeer and Belvedere; The Eisden -Lanklaar terrace is formed, leaving a steep slope at the plateau ' s edge, best visible at Schootsveld near Opoeteren (Saalian bend in the river)

This period, roughly between 200.000 and  100.000 BP, changing climates occur; during polar desert storms eolian sands were blown from coastal regions, "cover sands", forming large covers of sand over the fan gravels; by erosion processes gravels down-sloping cut in wide valleys, transporting melting waters during interstadial conditions- leaving dry valleys up to 20 m deep

During the Weichselian period the landscape was more formed by melting waters, forming new gully 's . Linear sand- dunes were formed falling over the plateau' s edge at the east side of the plateau.

During the Holocene, plateau edges were eroding and the Maas/ Meuse river flows over the current alluvial plain. Eroding surfaces occur because of i.e. deforestation, a wrong use of the landscape and by changing water tables.

Tool technology
Naturally broken pebbles are rare in the Kempen area. Where we find them, they only occur sparsely, i.c.  showing  random fractures, but mostly on the edges ( where the pebbles are more thin. Naturally split pebbles show steep angle fractures, capricious edges,  and do not show controlled flaking ( e.g. bulb complex fracture or bulb setup ). Outside the sites where thousands of artifacts were found, there were hardly found any naturally broken pebbles, only sometimes on unpaved sandy forest paths.

The found objects were determined as man-made artifacts by following arguments:
1. repeating flaking patterns, showing flat bulbs and impact points, sometimes even with indications for the direction of the blow ( grain wave patterns)
2. Blows positioned at exactly the same positions on e.g. rounded pebbles, providing logical working edges combined with logical handles;
3. Identical blows on same raw type of  materials independent of size of pebble
4.low angle blows
5. large assemblages, found in concentrations/ outside the concentrations at the sites could  not be found  a single fractioned pebble
6. Unifacial flaking indicates controlled flaking by leaving useful tools such as choppers, chopping tools, core tools, spheroids
7. Only quartz was transformed into the  'pick' tool type
8. Concentrations of nice "spall burin - flakes", often partially secondary retouched
9. Denticulate edges
10. Total tool groups  form logical assemblages
11. Assemblages found at same altitude, strongly suggesting simultaneously made artifacts
12. Comparative artifacts from other similar locations ( river upper terrace) like "Haute Garonne"(France) and "Atapuerca" Spain (part  Sima de los Huesos)
13. Local gravel type, belonging to one period ( depending on loci)
14. Finds of several 'hammerstones', like with battered textures either distal or around
15. For an overview of tool types and technology see the article from this website
Tool types and raw material of Acheulian like tools in the Belgian Kempen area

The period of production, use and discard for these artifacts has been determined by the following arguments:
1. Position of the artifacts in the upper river terrace of the Maas/ Meuse river, =  Belgian Kempen Plateau, where the artifacts are part of the local Zutendaal gravels.
2. Artifacts are found 'in situ' in horizontal layers, showing a direct relation with the Zutendaal gravels; positioned  below the reworked ( mixed)  eolian cover sands and positioned  at the top of the red oxidation - layers of the As paleosol;
3. This is confirmed by  the comparisons of this stratigraphical position with visible stratigraphies nearby 
( pits)
4. In all assemblages at all sites not a single artifact has been found originating from 
possible other periods, nor any indication has been found of different tool production
within the assemblages- indicating one homogeneous standard of tool technology and
virtual tool concept, fitting in the Acheulean / i.c. European Oldowan
5. The taphonomy of the artifacts is in accordance with the estimated age of a Middle 
Pleistocene - early Middle Palaeolithic period, showing iron stains, sanded surfaces and
shiny mineralization of the surface "the sparkling effect"
6. Comparative assemblages from the same type of locations ( upper terraces) in 
different countries - France, Hungary, Germany, Spain , England- show similar tool
assemblages that all have been placed in the Acheulean period
7. Adapted flint artifacts show a similar bifacial tool technology concept as for the 
pebble tools, suggesting one concept for tool production, in a stadium before the 
Mousterian  core preparation techniques;
8. Both  within the different sites as well as between the sites, some small differences in 
tool production technology can be noticed. Such a thing is indicating a similar use
of the sites during a period, where pebble tools were the standard ( contrary to flint, 
found ca. 25 km to the south, which has become more or less the standard  during
the Mousterian 
9. It is very unlikely large amounts of exposed pebbles from the river coast 
were transported from the middle terrace to the upper terrace during a 
Saalian period or later ( so the period of the Neanderthals and later), 
for the production of tools. Moreover, it is not likely too, this would have happened 
at all places along the Maas river, as the tool production sites all are located at the 
upper Maas terraces 

Homo heidelbergensis (?).

The tools are very comparable to those tools attributed to H. heidelbergensis
( Thames river terrace, Haute Garonne river terrace.)
The "Kempen Stone Face"
At site MA4 a side scraper has been found, with an appearance ( en profile) of an human face, complete with heavy eyebrow, common to H. eretcus / heidelbergensis species. Ken Johnston, specialist on portable  palaeorock -art  writes about it at his website Portable Rock Art. Direct link to the articles about the Kempen Stone Face

Assumed situation for the Maas/ Meuse river during the Holstein interglacial (MIS 11). Dark blue = Holstein Maas and tributary the Rur, that changed its coarse later into northern direction (Vlootbeekdal, and later Roermond) . The wide alluvial plain ( brown lines ) has been incised  by late- Elsterian  heavy energy transport of the main channel, so this is  referring  the most wide  Holstein river channel edges , with an inclination < 10  m.. Small lakes, remains of  former river channels on loamy gravel -loam horizons  on the plateau formed small tributaries with the Maas/ Meuse river. This palaeo-geographical situation is based on correlated archaeological finds from the upper terraces, explaining possible early settlements or activity zones in their original geography.
Overview of  sites and artifacts ( including the MA4)
Section           Site                   Preforms       Large tools         Small tools            Flint
PHH 1 -3





PHH 10 /11

PHH 12 /13


PHH 15


PHH 17/18

PHH 25


PHH 28


PHH 30


PHH 32

MA4 -2

MA4- 4

MA4 -5






Some remarkable objects
In the continuous search for early hominid presence in the region, several remarkable objects were detected, of which it is not yet clear what the meaning and relation is with the tools found at the various sites. A number of these objects are presented here.
This object could be interpreted as a biface - form, made on a  flake ( split cobble).  In the assemblages  of MA4, where this object was found, random flaked objects were found. The triangle  on this object might have a meaning. 

 -----------------------------------------------------------------------------------------------------------------Notes in the text 
(1) see for an overview of archaeologic sites in Vlaanderen : Onderzoeksbalans Onroerend Erfgoed Vlaanderen  or only look the map for the locations of communities where early Palaeolithic sites were found  image here 
(2) The former Pleistocene Maas riverbanks at the Kempen plateau have been heavily eroded during the Saalian, At Opoeteren we can stil see a meander bend  positioned on the Eisden-Lanklaar terrace near Bergerven. The Saalian period causaed heavy erosion, making large incisions ( up to 25 m , visible like  present dry valleys) in the steep east slopes of the Kempen plateau.  So the exact distance between a findspot  and the Pleistocene  Maas river is unknown. All erosion of the surfaces of   discovered sites is a recent happening, due to human activities.
(3) See for this Saalian erosion : Gullentops et al. 1993: Saalian nivation activity in the Bosbeek-valley -PDF
(4) Gullentops  and Paullise, reporting ice rafted blocks in the Zutendaal Gravels, (Quarternary Lithostratigraphicunits ( Belgium) (2001) , 3.1.5., deposits from "Cromerian"probably  glacial B (= MIS 20) ; indeed I found large ice rafted blocks in the main original context on the terrace; artifacts were found in large amounts of ( disturbed) coarse sands on the slope
(5) It is  not very likely that after covering of the gravels during Saalian / Weichselian period, so after the glacial periods, people have dug out gravel stones for tool -production exactly at the same level, over a distance of 500 m. Gravels were very easy accessible without digging in the Maas riverbed during the pleniglacial and  (post-) glacial periods. The reason to produce large numbers of artifacts at one place is not always because this is a habited site ( compare i.c. The Ryckholt Flint Mines Complex (NL), where raw materials were transformed into tools ). Since no traces of fire are ( yet) discovered, there is  no real evidence of habitation. The large scale production of such archaic tools is suggesting  people produced and used these tools, especially because large cleavers were found.
(6) This is in accordance to the conclusions of Gullentops/ Paulissen ( 1972): the dry valleys in the eastern part of the Kempen plateau were formed by intense erosion during the Saalian period. This underlines the pre- Saalian date of the site MA4-2.( Gullentops, F. & Paulissen, E. 1972: Origine et âge des vallons du rebord oriental du plateau de la Campine In: Congrès et Colloques. Symposium de Géomorphology, 137-151.)
(7)This period would correspond with  the human occupation compared with  e.g. The Somme, Seine and Yonne- region in France , as a 'nearby' region. Human presence is established in interglacial conditions, mainly in MIS 11- MIS 13 stages ( 450- 500 kya) the  MIS 10-11 transitional stage  ( 400 kya)  ( dated to the northern European chronostratigraphy (Antoine et. al 2007),. The site cluster MA4 could reflect activities from the interglacial Holsteinian- phase ( ca 400 kya)  and the following undefined glacial period  (386 kya - 338 kya, this corresponding with the As Palaeosol and the overlaying gravels under the eolian sand deposits. For chronostratigraphy for this article see: chronostratigraphy table Version 1.2 ASAC at this website
(8) All these investigations are only made by vision, not by excavations or digging
(9) See : Sarunas Milisauskas, in : European prehistory (p. 30) : a survey; Springer 2002. Milisauskas warns in the book for naturally broken stones, causing problems with the identification of the earliest arrival of humans in Europe.

Ref. + internet
Armand. J.1979 The Middle Pleistocene Pebble Tool Site of Durkadi in Central Inida PDF, (Persée fr.) ,vol 5, nr 5 105-144
Bridgland, D.R. 2000 River terrace systems in north-west Europe: an archive of environmental change, uplift and early human occupation Quaternary Science reviews  Volume 19, Iss.13, September , Pages 1293-1303
Kolfschoten, T. van & Y. Vervoort-Kerkhoff,  1985 Nijlpaarden van Nederlandse bodem en uit de Noordzee. - Cranium 2(2); 36-43.
Lauerent, M. et al. 2007 ESR chronology of the Somme River Terrace system and first human settlements in Northern France Quaternary Geochronology, issue 1-4, volume 2
Mortensen, Peder 2008 Lower to Middle Palaeolithic artefacts from Loutró on the south coast of Crete Antiquity Vol 82 Issue 317 
Raynal  J.P.,  Magoga  L.,  Sbihi-Alaoui F.Z.,  Geraads  D.,  1995  -  The  Earliest Occupation of Atlantic Morocco : The Casablanca Evidence. in  The earliest occupation of Europe, W. Roebroeks & T. van Kolfschoten Ed., University of Leiden, 1996, 255-262.

North- West European rivers in the Elsterian/ Anglian stage 

AHOB publications list England ( Early dispersal of humans in Britain)

TopoEurope/ Pleistocene maps

Identifying and protecting archaeological remains; English Heritage 1998

Map of Kempen area, Wikipedia, 2010 by Willy Leenders ( free license)


Special thanks to Guillaume Willems from  Maastricht for his continuous support in many ways, as for  the transportation and for providing of facilities.

Guillaume Willems, during one of the many  field trips in the Belgian Kempen.