Paradise Lost? Insights Into the Early Prehistory of the Netherlands from Development-Led Archaeology

Nederlandse Archeologische Rapporten 062

Paradise lost? Insights into the early prehistory of the Netherlands from development-led archaeology

J.H.M. Peeters, D.C.M. Raemaekers, I.I.J.A.L.M. Devriendt, P.W. Hoebe, M.J.L.Th. Niekus, G.R. Nobles & M. Schepers Paradise lost?

Insights into the early prehistory of the Netherlands from development-led archaeology

J.H.M. Peeters, D.C.M. Raemaekers, I.I.J.A.L.M. Devriendt, P.W. Hoebe, M.J.L.Th. Niekus, G.R. Nobles & M. Schepers Colophon Nederlandse Archeologische Rapporten 62

Paradise lost? Insights into the early prehistory of the Netherlands from development-led archaeology

Authors: J.H.M. Peeters, D.C.M. Raemaekers, I.I.J.A.M.L. Devriendt, P.W. Hoebe, M.J.L.Th. Niekus, G.R. Nobles & M. Schepers Illustrations: P.W. Hoebe and J.H.M. Peeters unless otherwise indicated Cover image: excavations at Epse Olthof (Archeologie Deventer) Copy editor: Suzanne Needs-Howarth Authorisation manuscript: T. Terberger (Universität Götingen / Niedersächsisches Landesamt für Bodendenkmalpﬂege) Layout and production: Xerox/OBT, The Hague

ISBN/EAN: 9789057992926

Cultural Heritage Agency of the Netherlands P.O. Box 1600 3800 BP Amersfoort www.cultureelerfgoed.nl 3 Contents —

Preface 5 5 People and environment 165 5.1 Introduction 165 Summary 6 5.2 Exploitation and use of organic resources 165 5.2.1 Subsistence and food processing 166 Samenvating 12 5.2.2 Mesolithic pit hearths: wood for fuel and tar 178 5.3 Environmental dynamics and behavioural 1 Research framework 19 diversity 185 1.1 Introduction 19 5.3.1 Continuity and discontinuity of occupation 1.2 Analytical framework 20 and behaviour 186 1.2.1 People and materials 21 5.3.2 Human activity and environment 195 1.2.2 People and space 22 1.2.3 People and environment 22 6 Research agenda, ﬁeldwork and reports 201 1.2.4 Spatio-temporal dimensions 23 6.1 Introduction 201 1.2.5 Knowledge domains 23 6.2 The research agenda: knowledge domains and 1.3 Report selection 24 research perspectives 201 1.4 Practical approach 24 6.2.1 People and Material 201 6.2.2 People and Space 205 2 Knowledge system and report assessment 25 6.2.3 People and Environment 207 2.1 Introduction 25 6.3 Connecting to development-led research 2.2 Knowledge system 25 in practice 208 2.3 Report assessment 26 6.3.1 Predictive models 209 2.3.1 Report selection and preliminary assessment 26 6.3.2 Prospection: augering depth and sampling grid 210 2.3.2 Geographical and chronological coverage 30 6.3.3 Excavation: phased strategies 212 2.4 Thematic data mining 45 6.3.4 Analysis 215 6.3.5 Reporting 218 3 People and material 51 6.4 Conclusions 218 3.1 Introduction 51 3.2 Craf activities 52 7 The national research agenda and future 3.2.1 Production, use and discard of lithic tools 52 perspectives 219 3.2.2 Organic resources for artefacts 76 7.1 Signifcance assessment and scientifc 3.3 Socio-cultural meaning 82 knowledge gain 219 3.3.1 Materiality: the meaning of objects 83 7.2 The National Archaeological Research 3.3.2 Death and ritual 86 Agenda and research questions 220 3.4 Frames of reference 89 7.3 What’s next? 224 3.4.1 Typo-chronology of the Final Palaeolithic and the 7.4 Conclusion 227 Mesolithic 89 3.4.2 Chronological and regional trends in potery Bibliography 229 production and use 104 Appendices 247 4 People and space 113 4.1 Introduction 113 4.2 Lithic resources in the landscape and mobility 113 4.2.1 Late Palaeolithic long-distance transport 115 4.2.2 The Wommersom distance-decay distribution 117 4.2.3 Newly ‘identifed’ and ‘ghost’ raw material sources 119 4.3 Place and space 125 4.3.1 In search of spatial diferentiation 126 4.3.2 Understanding fnds scaters: activity areas? 132 4.3.3 Tents, huts, houses and other built structures 146

5 Preface —

This book presents the results of a synthetic P.W. Hoebe (BA) acted as database manager, analysis of archaeological reports dealing with assisted the other group members in extracting aspects of the early prehistory of the relevant data from the reports and analysed Netherlands. The reports were produced in the various aspects for the assessment of the context of development-led commercial information potential of the reports. archaeology over the past 10 to 15 years. Drs. M.J.L.Th. Niekus studied and analysed The Cultural Heritage Agency of the the reports in the context of the themes people Netherlands (Rijksdienst voor het Cultureel and materials and people and environment, but Erfgoed; RCE) commissioned a project, particularly focussed on lithic typochronology undertaken by the Groningen Institute of and Mesolithic pit hearths. Archaeology (GIA), which aims at a synthetic Dr G.R Nobles analysed the reports in the analysis of (mostly) excavation results context of people and space, with an emphasis concerning early prehistory (here defned as the on spatial analysis at the intra-site scale and Palaeolithic to the Middle Neolithic). The project various methodological issues regarding is part of the RCE’s Valleta harvest (Oogst voor feldwork strategies. Malta) project, which resides under the Dr M. Schepers analysed the reports in the ‘archaeological knowledge kit’ programme context of the themes people and materials and (Kenniskaart Archeologie). The aim of the people and environment, with an emphasis on programme is to develop tools or utilities to the use of organic materials, and aspects of support heritage management decision making environmental dynamics. from a national perspective. The book has greatly profted from The book is the result of a joint efort by the comments on various drafs by Dr Luc Amkreutz authors. Although we each made our own (National Museum of Antiquities, Leiden) and specifc contribution, none of the chapters is Dr Bjørn Smit and Dr Eelco Rensink (Cultural authored by a single contributor. In order to Heritage Agency, Amersfoort). Prof. Thomas account for the work done by each one of us, we Terberger (Universität Götingen; wish to provide a broad outline of who did what. Niedersächsisches Landesamt für Denkmalpfege, Dr J.H.M. Peeters acted as project leader Hannover) peer-reviewed and accepted the and wrote the running text on the basis of input manuscript in terms of its scientifc contents. (rough drafs, lists, tables, graphs and maps) Dr Suzanne Needs-Howarth (Toronto) thoroughly provided by the other project members and copy edited the text and improved it through her expanded on many topics throughout the book. critical comments and queries for clarifcation. Prof. D.C.M. Raemaekers acted as co- Furthermore, several individuals and project leader; he also studied and analysed the archaeological frms took the trouble to provide reports in the context of the theme people and us with high-resolution images from their reports: materials, with an emphasis on ceramics. Günther Noens (Genth), Archol BV (Leiden), ADC Dr I.I.J.A.L.M. Devriendt also studied and BV (Amersfoort), BAAC BV (‘s-Hertogenbosch), analysed the reports in the context of the theme RAAP BV (Amsterdam), and BOOR (Municipality people and materials, but particularly focussed of Roterdam). The authors want to thank all for on lithic technology (production, raw materials their work and their willingness to cooperate. Any and tool use); jewellery; and burial. mistakes and omissions remain our responsibility. 6 — Summary

Aim 2014. The results of our study will feed in to an Archaeological heritage is important. This notion evaluation of the national research agenda and has led the Dutch government to create will eventually help inform a proposal for legislation in order to protect archaeological changes to the research questions for NOaA 2.0, values. Since 2007, the Dutch heritage legislation which has recently been launched. is part of the so-called Malta legislation, which represents the implementation of the Valleta Approach Convention (1992). The 2007 law states that the We have approached the analysis in a way that ‘disturber’ has the obligation to investigate the permits to simultaneously broaden and deepen (potential) presence of archaeological remains insights under three overarching diachronic prior to the planned disturbance, and to themes, each of which, in turn, can be subsequently take measures to safeguard this approached from four analytical spatio- heritage, either in situ or ex situ. Archaeological temporal dimensions: (a) local and short-term, heritage is primarily the responsibility of (b) local and long-term, (c) regional and short- municipal or other local authorities. Hence, term, and (d) regional and long-term. The there is no nation-wide policy on how to deal overarching themes connect to three broad with the legislative obligations. In the context of arrays of the ‘human realm of existence’: people signifcance assessment and subsequent and materials, people and space, and people selection of which sites to excavate, version 1.0 and environment. All three themes have a direct of the national archaeological research agenda bearing on the study of human behaviour in the (Nationale Onderzoeksagenda Archeologie; past on the basis of archaeologically recorded NOaA; 2006) was intended to serve as a national data and subsequent interpretation of the scientifc reference framework for decision paterning found within. making and a starting point for research design. In order to enable structured searching of A recently conducted evaluation of the reports, it was necessary to build a fexible efectiveness of the heritage management cycle database structure which allowed for the under the Malta legislation has revealed that selection of reports concerning sites that reports resulting from development-led potentially provide information on particular research lag behind with respect to the topics. In order to connect our research interpretation and synthesis of excavation questions with the reports, we compiled a list of results, relative to the research themes and 50 ‘aspects’ that we also deemed to be questions in the NOaA 1.0. This raises the prerequisites for formulating answers to the question: What can we learn from all these research questions. These aspects range from investigations that have found their way into physical archaeological evidence to feld reports? For this purpose, the Cultural Heritage methods and applied analyses. The screening of Agency of the Netherlands (Rijksdienst voor het reports on the basis of these aspects indicates Cultureel Erfgoed; RCE) commissioned a project, which questions can potentially be answered undertaken by the Groningen Institute of based on the information provided in each Archaeology (GIA), which aims at a synthetic report. Furthermore, the connection between analysis of results concerning early prehistory each research question and report was stored in (Palaeolithic to the Middle Neolithic). The project relationship to NOaA 1.0 research themes and is part of the Oogst voor Malta (Valleta harvest) the overarching themes, as defned under the project of the RCE, which resides under the umbrella of our analytical framework. ‘archaeological knowledge kit’ programme (Kenniskaart Archeologie). The aim of the Assessment programme is to develop tools or utilities to The selected reports provided unbalanced support heritage management decision making coverage in terms of the representation of from a national perspective. In the particular geographical regions, archaeological landscapes, case of our study, the aim is to establish to what and chronological and cultural atributions. This extent research questions from the NOaA 1.0 can is due to diferences in geological history, the be answered at a synthetic level, drawing from nature of modern-day land use and spatial the published reports resulting from development, the nature of development-led development-led projects from the period 2005- feldwork, the nature of political decision 7 —

making, as well as archaeological conceptual enough critical mass to deepen insights into constraints. Any synthetic analysis of data typo-chronological developments. Partly, this derived from these reports has to take these has resulted in the confrmation of the biasing factors into account, most particularly established phasing of the Mesolithic. At the where it comes to interpretations with a same time, it is also evident that there is geographical component or the reconstruction considerable geographical variability. This may of long-term occupation histories. mirror diferent technological traditions, a Notwithstanding these restrictions, the analysis hypothesis that connects to the distinction made allows for a critical evaluation of existing between the Northwest Group and the Rhine- insights and ideas, as well as a (re)formulation of Meuse-Scheldt Group on the basis of typology hypotheses for future research. many decades ago. The technological modes of Particular problems were encountered with production that seem to be connected with regard to quantitative data, such as the mention these two groups – and perhaps more – have of total numbers of (categories of) fint or received litle atention. Yet, socio-cultural potery in the text, but no summary table; the transmission of technological knowledge and use of vague quantitative terms (such as much, practical expertise must underlie the persistent many, litle), instead of exact numbers; or the existence of these technological traditions, provision of percentages without clear indication which themselves will have been subject to of sample or population size. Other types of variation and change. It is more than worthwhile problems concern issues of classifcation and to consider just these aspects in future analyses. subsequent presentation of data, the Along the same lines, there is still much to presentation of radiocarbon dates and the gain regarding the study of ceramic availability of primary data. technological traditions. Even more so than lithic studies, ceramic studies seem to serve relative Analysis: people and materials dating purposes. It will require a major shif in When considering the spatial scale of archaeologists’ minds to step beyond this investigation of the relationship between people superfcial approach to ceramics. Research and materials, we were confronted with questions need to be more explicit regarding problems of representativeness. Mostly, results modes of production and use, the study of which from development-led research can only be can rely on various methods and techniques that assessed at the local scale. The sites that have are readily available. Because such data are as been subject to development-led excavation yet barely available, use of these techniques will were selected on the basis of what one might allow archaeologists to gain knowledge not only call ‘anticipated success’. Basically, this involves in the local and short-term domains, but also in the selection of sites for which, in consideration the regional and long-term domains. of survey data, chances are high of fnding the The possibility of increasing our insight into expected. Hence, the results from projects that the socio-cultural meaning of material were concerned with the investigation of (materiality) is not restricted to utilitarian ‘familiar’ contexts have produced insights that objects. Several projects have provided mostly ft or confrm the existing picture. On the information with regard to ritual or cultic other hand, the use-wear studies conducted to dimensions of objects and burial practices. date have demonstrated considerable variability Ofen, the manifestation of belief systems in tool use (fint and other rock) and have also through the treatment of objects and of the provided evidence for the working of materials deceased are disconnected in archaeological that themselves were not previously known studies, and development-led projects are no from the archaeological record (shell and jet in exception herein; in many cases, fnds categories the Mesolithic). In addition, these use-wear are decontextualised. The interpretation of analyses give archaeologists much to think particular fnds contexts as ritual depositions about concerning the relationship between tool may appear less straight forward than the typology and tool function. identifcation of graves, but this is certainly not With regard to long-term regional trends in (always) the case. In the example of Roterdam- Mesolithic tool types (microlithics), Beverwaard, the Mesolithic cremation graves development-led projects have generated were only identifed as such thanks to the 8 —

analysis of the small number of calcined bone are palimpsests, which complicate the fragments. The presence of a perforated cobble deciphering of the archaeological record at the mace head alone would not have led to the site level; second, there is a lack of application of same conclusion. However, it makes one wonder suitable spatial methods and techniques. In about comparable fnds contexts where bone most cases, spatial paterns within fnds remains are lacking. Hence, there is a need to distributions are mostly only subject to visual pay particular atention to fnds context in inspection of maps that present various general, even when objects could just as easy be atributes of fnds categories. Statistical explained as having been lef in a secular seting. approaches are almost lacking, as are multiscalar techniques, which permit archaeologists to Analysis: people and space investigate spatial paterning amongst diferent We are somewhat ambivalent regarding the gain fnds categories. Connecting fnds distributions of knowledge around the theme people and to features and the identifcation of built space. The regional dimension of space appears structures equally sufer from a lack of thorough to be one which can barely be addressed. analysis. The approaches are complementary, Potential new insights into such topics as for which reason we favour the application of setlement location choice and setlement multiple techniques – visual, statistical, systems can only be evaluated when hitherto multiscalar – to potentially increase insights into underrepresented landscape zones and short- and long-term dimensions of the use of geographical regions are investigated. However, locations. because survey strategies largely build on the known archaeological record through predictive Analysis: people and environment models, and sites selected for excavation ft The number of data points for the theme people those models, underrepresented zones and and environment is restricted. Some aspects regions are simply not investigated. have been shown to deserve atention, notably Nonetheless, development-led projects with respect to the use of animal and plant have permited to gain insight into long- resources. distance transport of raw materials, notably red Bone, tooth, and antler, mostly calcined and Heligoland fint (Late Palaeolithic), Wommersom mostly in small numbers, appear to be found quartzite (Mesolithic), and possibly fint from more ofen at Mesolithic and Neolithic sites than other sources along the coast of northern France one tends to think; the evidence adds useful and southern Scandinavia. Although the data are information to the overall still poor record scanty and observations are incidental, they dating to early prehistoric times. Furthermore, show that identifcation of potential raw animal bones whose anthropogenic origin is material sources provides new insights into the questionable can still provide information about geographical sphere of interaction in various the composition of the ‘animal landscape’ in time periods. Importantly, such insight draws terms of species diversity. This is a much needed our atention to the degree of mobility of groups source of information, as reference frameworks and to the social interaction between groups. for subsistence strategies are largely founded on Most of the acquired results relate to the actualistic models of animal behaviour and local use of space, that is, the intra-site level. communities. Yet other phenomena, such as Spatial information about fnds distributions and arrangements of wooden stakes in stream beds, features is provided in all of the reports included provide information about food procurement in this study, but analysis of data and subsequent techniques, notably fshing techniques. 14C-dates interpretations of paterns remain superfcial. of such stakes point to surprisingly signifcant Eforts to increase our understanding of site- time depths. formation processes and palimpsests are an Development-led research has made a exception, despite the fact that the overwhelming considerable contribution to our knowledge majority of sites excavated manifestly result from about hunter-gatherer plant resources. Here we long occupation histories and complex see that the application of modern techniques of depositional and post-depositional processes. analysis by specialists in the feld leads to There are several reasons for the present exciting results. Such apparently dull categories situation. First, the majority of sites investigated as charcoal and other charred plant remains can 9 —

provide information that was not easily ‘extinct landscapes’, which are so far only partly accessible until recent years. Insight into the represented in the fossil physical record. Another exploitation of a range of plants for food has source of information concerns stable isotopes increased considerably, most notably with – notably 15N and 13C. Although it has been regard to wetland species, the remains of which known for a considerable time that stable have been found in charred condition. Research isotope analysis can yield important information regarding the use of Mesolithic pit hearths has about human and animal diet and habitat, such resulted in the development of a new analysis is not conducted in the context of hypothesis, notably their potential function in development-led research. Ofen not even the the process of wood-tar production. 15N and 13C values are provided. A recent study of Notwithstanding the fact that several isotope data from Mesolithic human remains potential interpretations of the pit hearth data from the North Sea has shown that even are possible, the importance of the reported contextless fnds can provide meaningful results carries further than just the possible use information about human diet. of the enigmatic Mesolithic pit hearths. Because these features are prominently present on many The research agenda and ﬁeld practice sites north of the Rhine throughout the The reports on the more extensive projects Mesolithic, the question arises how diferences make clear that in-depth study of fnds in time and space can be explained, for instance, categories and excavation data can contribute to in connection with climate-driven models about the answering of research questions in the vegetation succession. Other related questions NOaA 1.0. We conclude that, for the most part, refer to the time-depth of locational use – this it is not so much the research questions can span more than 2000 years – in connection themselves that are the problem, but, rather, with such pits. their embedding in project outlines and The same holds, to some extent, for strategies. Ofen the solution needs to be found research into human impact on the in the frst step of the research process, namely, environment. Traditional models about the the project outline (Programma van Eisen). restricted infuence of early prehistoric people, Within an archaeological ‘market’ in which mostly hunter-gatherers, on their environment competition is primarily expressed in fnancial are now being questioned. Development-led terms, commercial frms are evidently hesitant research again has made a considerable to do more than is required. The quality of the contribution to this debate, mainly because of project outline therefore determines the quality an open-minded involvement of specialists. of the research project to a large extent. The methods applied are traditional, but the One characteristic of development-led frameworks of reference have changed under archaeological research in the Netherlands is the the infuence of an international debate. In fact, use of predictive models. The rationale is that in development-led work in the Netherlands is at the past resources were exploited using various the forefront of these new developments. The strategies and technologies based on (expected) publication of such results in scientifc papers by availability, experience and tradition. We professionals who work in development-led conclude that it will be necessary to make model archaeology is, however, frustrated by the expectations more explicit. Next, we will need to dictates of commercial continuity. establish to what extent results ft these explicit It is important to note that recent years expectations in order to improve our predictive have seen the development of innovative models. Diverging results can involve absence of techniques that use data sources that also are the expected or presence of the unexpected. (or can be) found in the Netherlands. Analysis of The later readily causes problems if identifed ancient DNA is potentially possible. A recent during feldwork, as a change of strategy and study of skeletal remains from Bell Beaker methodology is ofen needed, which in turn contexts in Europe – including the Netherlands – normally has implications for the overall demonstrates the potential for genetic studies. planning and budget. This factor in particular, Equally, the potential presence of aDNA in frequently lead to decisions that turn early buried palaeosols – sediment aDNA – may shed prehistoric remains into ‘by-catch’, which is light on animal, plant and human populations in reported on to only a minimal extent. Yet it is 10 —

precisely this by-catch which may represent low- strategy has become common in Dutch resolution phenomena (low-density fnds archaeology. In this strategy, excavation starts scaters, isolated pit hearths, isolated projectile with the excavation of a small grid cells with a points) which are otherwise underrepresented in relatively large distance between them. the archaeological record. Diferences in fnds numbers are used to In many projects, preliminary stages of delineate clusters. Low-density zones are mostly research include surveying by means of excluded from further research. Here, two augering. Here two issues demand our atention. concerns need to be addressed. The frst concern First, the augering depth is limited to the is that the use of grid-cell densities to predict anticipated depth of disturbance. As a variation in the density of the fnds scater is consequence of this standard procedure, one methodologically problematic. We must specifc group of archaeological remains remember that fnds scaters are of a systematically receives too litle atention: fundamentally diferent nature than distributions Middle and Late Palaeolithic remains covered by of soil or geological units. Whilst soils and younger Pleistocene sediments. As yet, geological units, such as sediment layers, are anticipated atention for Palaeolithic sites continuous three-dimensional phenomena, remains exceptional, and, as a consequence, artefacts as composite elements of sediment fnds fall systematically in the category ‘by- layers are not. It is important to note that Stone catch’. Many development plans do, however, Age sites that have been extensively excavated result in the disturbance of Pleistocene deposits, (e.g. Hoge Vaart-A27, Dronten-N23, Ede- for instance, due to piling for foundations of Kernhem and Schipluiden) demonstrate buildings. Any Middle and Late Palaeolithic sites substantial quantitative variation regarding the that are present are thus disturbed and number of fnds within high- (or low-) density inaccessible for archaeological research. Second, zones, even at short distances. This implies the detection of Stone Age sites by means of intrinsic uncertainty with regard to quantitative auger surveys faces many problems. Various interpolation of archaeological sample data. The statistical studies have demonstrated that only second concern is that the quantity-driven high-density scaters of at least a medium-sized selective approach to excavation results in patchy horizontal extent – that is, palimpsests covering and discontinuous excavation plans, in which more than 1000 m2 – can reasonably be clusters or concentrations of fnds are ofen only detected in a 20 × 25 m sampling grid. Therefore partially excavated. This leaves litle potential for auger sampling as part of a standard the investigation of spatial aspects. Many prospection strategy is a disadvantageous techniques for spatial analysis are available, but starting position for the discovery of the kinds of their applicability requires continuous areas of smaller, high-resolution sites (e.g. low-density fully excavated grid cells, preferably including ‘single event’ scaters) that we would like to empty zones to minimise edge efects. have more information on in order to answer With regard to the analysis phase of many research questions in the NOaA 1.0. The research, fve concerns are raised. The frst is only way to increase the chances of locating such that the description and analysis of cultural sites is to anticipate their potential presence material is not standardised, which is a serious (e.g. based on palaeolandscape information) in hindrance when one intends to use the results of situations where terrestrial deposits dating to an analysis for new research. The second is the the Pleistocene or Early-Middle Holocene are near lack of functional analysis of ceramic, fint reached. and stone tools. The third is a general lack of During excavation, more problems may dedicated chapters or sections on spatial arise. Stone Age sites are characterised by the analysis. The fourth is that projects in which presence of a scater of fnds, which comprise early prehistoric remains are found as a by-catch zones with higher and lower densities of see limited involvement of Stone Age specialists. materials. Within development-led projects, the The fnal concern is that regional comparison is focus is ofen on excavating the high-density lacking in many reports due to the simple fact zones, and not ‘wasting’ time and money on that datasets are ofen not available in the relatively empty zones. To maximise the digital repository, in contravention of the outcome of the excavation, a phased excavation regulations. 11 —

We propose that standards should also be knowledge gain), as well as analysis and developed for reports, which are currently highly interpretation? The answer to this question diverse in terms of data presentation. Standards should set the baseline for the improvement of are ofen considered as undesirable straitjackets, the NOaA 2.0, which certainly should and can but they should only be seen as a way to share serve a purpose in quality control. We are information in a structured fashion, so as to convinced that a more balanced and targeted set allow for comparative analysis. of research questions can come from the dialogue between the development-led community and Towards a new research agenda the academic community, and for which the While we were conducting our study, the Cultural present study may well form the starting point. Heritage Agency of the Netherlands launched the National Archaeological Research Agenda 2.0 Conclusions (NOaA 2.0). Although the NOaA 2.0 is intended to Our synthetic analysis has made clear that highlight the most important themes and to foster basically all NOaA 1.0 research themes and operationalisation in the feld by means of practical questions remain either unanswered or partially suggestions, many questions implicitly have a high answered. Several questions, particularly those in degree of complexity. Furthermore, empirical the ‘why’ category or which require synthetic aspects are interwoven with interpretation, which work at the regional scale, are well beyond the complicates their day-to-day use in the practice of possibilities of development-led projects. Any development-led work. Of course it is possible to future research agenda that is intended to inspire draf a new set of research questions based on our development-led work and is concerned with the analysis of reports resulting from development-led quality of the archaeological research needs to work. Afer all, we have shown what be beter aligned with the potential and practice development-led research has contributed to the of what can reasonably be achieved within such a answering of research question from the NOaA 1.0. context. Importantly, our study also leaves us Since this contribution is rather limited, we could with the impression that there is a lot of satisfy ourselves by stating that all questions potential to be tapped from all the work that has remain relevant. In a way they do. But it is equally already been done. To tap this potential, all data clear that several NOaA 1.0 question surpass the need to become available, but even once that scope of the development-led context of research, has been achieved, in many cases researchers will notably, questions in the ‘why’ category. Indeed, probably need to return to the fnds material we feel that such questions should be omited. itself. So paradise is not lost, at least not entirely. The issues addressed above lead to a more Yet, it is necessary to develop further practices fundamental question: What should be the role which lead to useful and accessible research of development-led and academic research in the results from development-led projects, in order broader context of archaeological feldwork to ensure that paradise doesn’t crumble much (which should be aimed at the collection of data further. This is the responsibility of the that can stand at the basis of question-driven archaeological community as a whole. 12 — Samenvating

Doelstelling NOaA 1.0 op een synthetiserend niveau Archeologisch erfgoed is belangrijk. De beantwoord kunnen worden op basis van Nederlandse overheid heef daarom wetgeving gepubliceerde rapporten die zijn voortgekomen gecreëerd voor de bescherming van uit ontwikkelingsgedreven ‘Malta-onderzoek’ archeologische waarden. De Nederlandse dat in de periode 2005-2014 is uitgevoerd. De erfgoedwet is sinds 2007 onderdeel van de resultaten hiervan dienen ter evaluatie van de zogenaamde Malta-wet, de uitwerking van het NOaA onderzoeks vragen en leveren bouwstenen Verdrag van Valleta (1992). In de wet is gesteld voor een voorstel voor de onderzoeksvragen in dat de ‘verstoorder’ de verplichting heef om de NOaA 2.0, welke recentelijk is opengesteld. voorafgaand aan de voorgenomen verstoring onderzoek te doen naar de (potentiële) Aanpak aanwezigheid van archeologische resten en We hebben het onderzoek ingericht op een wijze maatregelen te trefen om dit erfgoed in situ of die het mogelijk maakt om inzichten te ex situ veilig te stellen. Archeologisch erfgoed is verbreden en verdiepen onder drie de primaire verantwoordelijkheid van overkoepelende thema’s die ieder vanuit vier gemeentelijke of andere lokale autoriteiten. Als analytische perspectieven benaderd kunnen gevolg daarvan bestaat er geen nationaal beleid worden: (a) lokaal – korte termijn, (b) lokaal – ten aanzien van de uitvoering van de wetelijke lange termijn, (c) regionaal – korte termijn, (d) verplichting. Ten behoeve van de archeologische regionaal – lange termijn. De overkoepelende waardering en successievelijke selectie van thema’s, waaraan diverse NOaA 1.0 vindplaatsen die in aanmerking komen voor onderzoeksvragen zijn verbonden, hebben opgraving, vormde de Nationale betrekking op drie brede velden van het Onderzoeksagenda Archeologie (NOaA 1.0; ‘menselijk bestaan’: mens en materiaal, mens en 2006) het wetenschappelijk referentiekader voor ruimte, en mens en omgeving. Deze thema’s besluitvorming en het vertrekpunt voor hebben direct betrekking op de bestudering van onderzoek. menselijk gedrag in het verleden op de basis van Een recentelijk uitgevoerde evaluatie van de archeologische gegevens en de successievelijke efectiviteit van de monumentezorgcyclus onder interpretatie van patronen die daarin gevonden de Malta wetgeving bracht aan het licht dat worden. rapporten tekort komen ten aanzien van Om op een gestructureerde manier te kunnen interpretatie en synthetisering van zoeken naar relevante rapporten was het opgravingsgegevens in relatie tot de noodzakelijk om een fexibele database onderzoeksthema’s en –vragen uit de NOaA 1.0. structuur te ontwerpen die het mogelijk maakte Dit roept de vraag op wat geleerd kan worden uit om rapporten te selecteren die potentieel al dat onderzoek dat haar weg heef gevonden informatie bevaten over specifeke naar rapporten. Om hier een antwoord op te onderwerpen. Om een koppeling mogelijk te krijgen heef het Groninger Instituut voor maken tussen onderzoeksvragen en rapporten Archeologie (GIA) in opdracht van de Rijksdienst zijn door ons 50 ‘aspecten’ benoemd die tevens voor het Cultureel Erfgoed (RCE) een noodzakelijk zijn voor de beantwoording van de synthetiserende analyse uitgevoerd van vragen zelf. Deze aspecten betrefen niet alleen onderzoeksresultaten betrefende de vroege archeologische resten, maar ook prehistorie (Paleolithicum tot en met Midden- veldwerkmethoden en toegepaste analyses. Neolithicum). Het project is onderdeel van het De fltering van de rapporten op basis van deze Oogst voor Malta-project dat deel uitmaakt van aspecten geef tevens een indicatie van de het Kenniskaart Archeologie-programma. Het vragen die potentieel beantwoord kunnen doel van dit programma is de ontwikkeling van worden op grond van de informatie die in ieder instrumenten en toepassingen ter ondersteuning rapport is verwerkt. De koppeling tussen iedere van besluitvorming in de erfgoedzorg vanuit een onderzoeksvraag en het rapport is tevens nationaal perspectief. In het specifeke geval van opgeslagen in relatie tot NOaA 1.0 thema’s en de de onderhavige studie is het doel om vast te overkoepelende thema’s zoals die zijn stellen in welke mate onderzoeksvragen uit de gedefnieerd in het kader van onze studie. 13 —

Waardering van ‘bekende’ contexten inzichten opgeleverd De geselecteerde rapporten leverden een die in het bestaande beeld passen, of dit onevenwichtige dekking van vindplaatsen op bevestigen. Anderzijds heef ten aanzien van vertegenwoordigde gebruikssporenonderzoek aanzienlijke variatie geografsche regio’s, archeologische in het gebruik van werktuigen (vuursteen, landschappen, en chronologische en culturele natuursteen) aangetoond, en zijn er eenheden. Dit hangt samen met verschillen in de aanwijzingen gevonden voor de bewerking van geologische ontwikkeling, het karakter van materialen die niet eerder in het archeologisch modern landgebruik en ruimtelijke gegevensbestand waren aangetrofen (schelp en ontwikkelingen, het karakter van ontwikkeling git in het Mesolithicum). De resultaten roepen gedreven onderzoek, politieke besluit name, ook de vraag op over de relatie tussen evenals archeologisch-conceptuele beperkingen. werktuigtypen en –functie. Iedere synthetiserende analyse van gegevens Met betrekking tot lange-termijn regionale afomstig uit de rapporten dient rekening te trends in mesolithische werktuigtypen houden met deze vertekenende factoren, in het (microlieten) heef Malta-onderzoek voldoende bijzonder waar het interpretaties betref met kritische massa gegenereerd om het inzicht in een geografsche component of de reconstructie typo-chronologische ontwikkelingen te van lange-termijn bewoningsgeschiedenissen. vergroten. Deels hef dit geresulteerd in de Ondanks deze beperkingen is een kritische bevestiging van de vigerende fasering van het evaluatie van bestaande inzichten en ideeën Mesolithicum. Tegelijk is duidelijk geworden dat mogelijk, evenals de (her)formulering van er sprake is van aanzienlijke geografsche typo- hypothesen voor toekomstig onderzoek. technologische variabiliteit. Dit refecteert Uit ons onderzoek zijn specifeke problemen mogelijk het bestaan van verschillende naar voren gekomen met betrekking tot technologische tradities; een hypothese die kwantitatieve gegevens, zoals de vermelding aanhaakt op het onderscheid dat decennia van totalen van vondstcategorieën in de tekst bij geleden gemaakt is tussen de Noordwest Groep het ontbreken van overzichtstabellen, het en de Rijn-Maas-Schelde Groep op grond van gebruik van vage kwalifcaties (zoals veel, werktuigtypen. Technologische productiemodi weinig) in plaats van absolute aantallen, of de die met deze groepen – misschien zijn er meer vermelding van percentages zonder duidelijke – verbonden lijken te zijn, hebben echter nog vermelding van de populatieomvang. Andere nauwelijks aandacht gekregen. Toch moet socio- problemen betrefen classifcaties en culturele transmissie van kennis en praktische gegevenspresentatie, de presentatie van ervaring ten grondslag liggen aan de persistentie koolstofdateringen en de beschikbaarheid van van deze technologische tradities, die echter wel primaire gegevens. onderwerp zijn geweest van variatie en verandering. Het is uiterst zinvol om aan Analyse: mens en materiaal dergelijke aspecten aandacht te schenken bij Ten aanzien van het onderzoek naar de relatie toekomstige analyses. tussen mensen en materiaal werden we In dezelfde lijn is er nog veel winst te halen voor geconfronteerd met problemen van wat betref ceramisch technologische tradities. representativiteit in verband met de ruimtelijke Meer nog dan onderzoek van lithisch materiaal, schaal van onderzoek. Doorgaans kunnen de dient onderzoek van aardewerk primair het doel resultaten van Malta-onderzoek uitsluitend van relatieve datering. Er is een grote worden gewaardeerd op de lokale schaal. mentaliteitsverandering onder archeologen Vindplaatsen die in het kader van Malta zijn nodig om voorbij deze oppervlakkige onderzocht, zijn geselecteerd op wat benadering van ceramisch materiaal te geraken. ‘geanticipeerd succes’ genoemd zou kunnen Onderzoeksvragen betrefende productie en worden. Dit komt er feitelijk op neer dat gebruik moeten explicieter worden gesteld, vindplaatsen worden geselecteerd die op basis terwijl het onderzoek gebruik kan maken van van de resultaten van vooronderzoek een gerede diverse methoden en technieken die direct kans hebben dat gevonden zal worden wat beschikbaar zijn. Aangezien dergelijke gegevens wordt verwacht. Als gevolg daarvan hebben nog nauwelijks beschikbaar zijn, kan de inzet van projecten die gericht waren op het onderzoek deze methoden en technieken leiden tot 14 —

kenniswinst op alle schaalniveaus (lokaal, (Laat-Paleolithicum), Wommersomkwartsiet regionaal; korte termijn, lange termijn). (Mesolithicum) en mogelijk vuursteen afomstig De mogelijkheid om inzicht te vergroten in de van bronnen langs de kust van Noord-Frankrijk socio-culturele betekenis van materialen en Zuid-Scandinavië. Hoewel de gegevens (materiality) is niet beperkt tot gebruiks- fragmentarisch zijn en het incidentele voorwerpen. Diverse projecten hebben observaties betref, geven deze aan dat de informatie opgeleverd over rituele of cultische identifcatie van potentiële grondstofronnen dimensies van voorwerpen en het dodenritueel. nieuwe inzichten kunnen opleveren in de Veelal worden manifestaties van geloofs- geografsche dimensie van interactie in diverse systemen – de omgang met objecten en perioden. Belangrijk is dat dergelijke inzichten overledenen – uiteen getrokken in de aandacht vestigen op de mate van mobiliteit archeologische studies, en Malta-projecten en sociale interactie tussen groepen. vormen hierin geen uitzondering; in veel De verkregen resultaten hebben vooral gevallen worden vondstcategorieën betrekking op lokaal ruimtegebruik, dat wil gedecontextualiseerd. De interpretatie van zeggen op intra-siteniveau. Ruimtelijke bepaalde vondstcontexten als rituele deposities informatie over de verspreiding van vondsten en mag minder eenvoudig lijken dan de identifcatie sporen wordt in alle rapporten gegeven, maar de van graven, maar dat is zeker niet altijd het analyse van gegevens en de daaruit volgende geval. Te Roterdam-Beverwaard bijvoorbeeld, interpretaties van patronen blijven oppervlakkig. konden mesolithische crematiegraven pas als Pogingen om ons begrip te vergroten van site- dusdanig worden herkend na analyse van een formatieprocessen en palimpsesten vormen een kleine hoeveelheid verbrand bot. Uitsluitend de uitzondering, ondanks het feit dat het overgrote aanwezigheid van een Geröllkeule in één van de deel van de onderzochte vindplaatsen duidelijk kuilen zou een dergelijke interpretatie niet het resultaat zijn van lange gebruiks- hebben toegelaten. Dit roept de vraag op naar geschiedenissen en een complex geheel van het karakter van vergelijkbare vondstcontexten depositionele en post-depositionele processen. waar botresten volledig afwezig zijn. Het is Er zijn verschillende oorzaken aanwijsbaar voor daarom noodzakelijk om meer aandacht te deze situatie. Ten eerste zijn de meeste schenken aan vondstcontexten, zelfs wanneer onderzochte vindplaatsen palimpsesten, wat de de mogelijkheid bestaat voor een verklaring ontcijfering van de archeologische gegevens waarin objecten eenvoudigweg kunnen zijn moeilijk maakt; ten tweede is er een gebrek aan achtergelaten in een seculiere seting. de toepassing van geschikte methoden en technieken. In de meeste gevallen worden Analyse: mens en ruimte ruimtelijke patronen in vondstverspreidingen Ten aanzien van kenniswinst binnen het thema uitsluitend onderworpen aan visuele inspectie mensen en ruimte zijn we wat ambivalent. De van kaarten die verschillende atributen van regionale dimensie blijkt nauwelijks benaderd te vondstcategorieën vertegenwoordigen. kunnen worden. Potentieel nieuwe inzichten in Statistische methoden ontbreken vrijwel geheel, onderwerpen zoals nederzetingslocatiekeuze en evenals meerschalige technieken die het de nederzetingssystemen kunnen alleen worden archeoloog toestaan om ruimtelijke patronen verkregen indien tot dusver ondervertegen- onder meerdere vondstcategorieën te woordigde landschapszones en geografsche onderzoeken. Ook de koppeling tussen regio’s zijn onderzocht. Dit is echter niet het vondstverspreidingen en grondsporen, en de geval, aangezien prospectiestrategieën middels identifcatie van gebouwde structuren krijgt voorspellingsmodellen voortbouwen op de weinig analytische aandacht. De benaderingen bekende archeologische gegevens, met als zijn echter complementair, zodat wij er voorkeur gevolg dat uiteindelijk onderzochte vindplaatsen aan geven dat meerdere technieken – visueel, zullen passen binnen de bestaande modellen. statistisch, meerschalige technieken – Toch hebben Malta-projecten inzicht opgeleverd gecombineerd worden om de inzichten in korte betrefende langeafstand transport van en lange-termijn dimensies van ruimtegebruik grondstofen, zoals rode Helgolandvuursteen (potentieel) te vergroten. 15 —

Analyse: mens en omgeving gehele Mesolithicum, rijst de vraag hoe Het aantal data-punten voor het thema mensen verschillen in het voorkomen van kuilhaarden in en omgeving is beperkt. Sommige aspecten tijd en ruimte verklaard kunnen worden, verdienden echter aandacht, vooral met bijvoorbeeld in relatie tot klimaat gestuurde betrekking tot het gebruik van dierlijke en modellen van vegetatiesuccessie. Andere vragen plantaardige grondstofen. hebben betrekking op de tijdsdiepte van Bot, tand en gewei blijken vaker te worden locatiegebruik – deze kan meer dan 2000 jaar gevonden – dikwijls gecalcineerd en zwaar bestrijken – in relatie tot dergelijke kuilen. gefragmenteerd – in mesolithische en Hetzelfde geld tot op zekere hoogte voor neolithische contexten dan snel wordt gedacht; onderzoek naar menselijke beïnvloeding van de de gegevens voegen bruikbare informatie toe omgeving. Traditionele modellen over de aan de doorgaans beperkte gegevens waarover beperkte invloed van de vroeg-prehistorische we voor de vroege prehistorie beschikken. mens, vooral jager-verzamelaars, op de Verder kunnen dierlijke resten, waarvoor een omgeving staan nu ter discussie. Malta- antropogene context twijfelachtig is, informatie onderzoek heef ook op dit vlak een bijdrage opleveren over de samenstelling van het geleverd aan deze discussie, vooral vanwege de ‘dierlijke landschap’ in termen van betrokkenheid van ruimdenkende specialisten. soortdiversiteit. Dit laatste is een wenselijke De toegepaste methoden zijn traditioneel, maar bron van informatie, omdat referentiekaders de referentiekaders zijn veranderd onder invloed voor bestaansstrategieën vooral gefundeerd zijn van de internationale discussie. In feite staat op basis van actualistische modellen over dierlijk Malta-onderzoek in Nederland in de frontlinie gedrag en populaties. Weer andere van deze nieuwe ontwikkeling. De internationale verschijnselen, zoals confguraties van houten publicatie van dergelijke resultaten in staken in stroombeddingen, leveren informatie wetenschappelijke tijdschrifen door over technieken voor voedselvoorziening, vooral professionals die in de Malta-context werken, visvangst. 14C dateringen van dergelijke staken wordt echter in de weg gestaan door de wijzen op verrassend grote tijddiepten in de noodzaak van commerciële continuïteit. lokale exploitatie van bronnen. Het is belangrijk om vast te stellen dat Malta-onderzoek heef een aanzienlijke bijdrage gegevensbronnen die met recente, innovatieve geleverd aan onze kennis over het gebruik van technieken kunnen worden onderzocht, ook in plantaardige bronnen door jager-verzamelaars. Nederland aanwezig (kunnen) zijn. De analyse Hier zien we dat de toepassing van moderne van aDNA is in principe mogelijk. Een recent wetenschappelijke methoden en technieken onderzoek van menselijke skeletresten uit door specialisten belangrijke resultaten kunnen klokbeker-contexten in Europa – inclusief opleveren. Ogenschijnlijk saaie Nederland – toont de potentie van genetische vondstcategorieën als houtskool en andere studies aan. Ook de potentiële aanwezigheid verkoolde plantaardige resten kunnen van aDNA in sediment (sedaDNA) – begraven informatie opleveren die eerder niet toegankelijk paleosolen – kan licht werpen op dierlijke, was. Het inzicht in de exploitatie van plantaardige en menselijke populaties in plantaardige voedselbronnen is aanzien lijk ‘uitgestorven landschappen’ die tot dusver verbreed, vooral wat betref soorten uit alleen maar fragmentarisch vertegenwoordigd wetland-contexten en waarvan verkoolde resten zijn als fysieke resten. Een andere bron van zijn gevonden. Onderzoek naar het gebruik van informatie betref stabiele isotopen, zoals 15N en mesolithische kuilhaarden heef geresulteerd in 13C. Hoewel al geruime tijd bekend is dat de een nieuwe hypothese, namelijk een potentiële analyse van stabiele isotopen belangrijke functie in het proces van houteerproductie. informatie kan vrijgeven over het dieet en de Ondanks het feit dat diverse interpretaties van leefomgeving van mensen en dieren, wordt kuilhaardgegevens mogelijk blijven, reikt het dergelijk onderzoek niet uitgevoerd in de belang van de gerapporteerde resultaten verder context van Malta-onderzoek. Zelfs 15N en 13C dan uitsluitend de mogelijke functie van deze gegeven worden niet verstrekt, ondanks dat raadselachtige mesolithische kuilen. Omdat deze deze t.b.v. 14C-analyse worden gemeten. Een verschijnselen prominent aanwezig zijn op veel recente studie van isotopengegevens van vindplaatsen benoorden de Rijn gedurende het menselijke resten uit het Mesolithicum uit de 16 —

Noordzee heef laten zien dat zelfs vondsten ondervertegenwoordigd zijn in het zonder context bruikbare informatie kunnen archeologisch gegevensbestand. opleveren over dieet. In veel projecten bestaat het vooronderzoek uit prospectie met behulp van boringen. Wat dit De onderzoeksagenda en de betref vragen twee aspecten om aandacht. Op onderzoekspraktijk de eerste plaats wordt de boordiepte beperkt tot De rapporten van de meer omvangrijke de geanticipeerde diepte van de verstoring. Als projecten maken duidelijk dat diepgaande gevolg van deze standaardprocedure krijgt één studies van vondstcategorieën en specifeke groep archeologische resten opgravingsgegevens een bijdrage kunnen systematisch te weinig aandacht: midden- en leveren aan de beantwoording van laat-paleolithische resten afgedekt door jonger onderzoeksvragen uit de NOaA 1.0. pleistoceen sediment. Tot dusver is de We concluderen dat de onderzoeksvragen zelf geanticipeerde aandacht voor paleolithische – voor het grootste deel – niet zozeer het sites uitzonderlijk, met als consequentie van probleem zijn, maar de inbedding ervan in vondsten systematisch in de categorie onderzoeksplannen en -strategieën. In veel ‘bijvangst’ vallen. Veel ontwikkelingsplannen gevallen moet de oplossing worden gezocht in leiden echter tot verstoring van pleistocene de eerste stap van het onderzoeksproces, het afzetingen, bijvoorbeeld vanwege de fundering Programma van Eisen (PvE). In een van gebouwen op heipalen. Potentieel archeologische ‘markt’ waarin competitie aanwezige midden- en laat-paleolithische sites primair wordt uitgedrukt in fnanciële termen, worden dan ongezien verstoord en onbereikbaar aarzelen commerciële bedrijven om meer te vanwege overbouwing. Een tweede probleem doen dan wordt vereist. De kwaliteit van het PvE betref de opsporing van steentijdvindplaatsen beïnvloedt daarom in belangrijke mate de met behulp van boringen. Diverse statistische kwaliteit van het onderzoek. studies hebben laten zien dat alleen omvangrijke Eén kenmerk van Malta-onderzoek in Nederland (> 1000 m2 – vondstspreidingen met een hoge is het gebruik van verwachtingsmodellen. De dichtheid een redelijke kans hebben om te onderliggende gedachte is dat de exploitatie van worden gedetecteerd in een grid van 20 x 25 m. bronnen in het verleden met behulp van Het gebruik van boringen is standaard verschillende strategieën en technologieën prospectiestrategie levert dan ook een gebaseerd was op de (verwachte) ongunstige uitgangspositie op voor de detectie beschikbaarheid, ervaring en traditie. We van kleine sites met een hoge resolutie (lege concluderen dat modelverwachtingen explicieter dichtheid; eenmalig gebruikte locaties) die we moeten worden gemaakt. Vervolgens dient te juist willen hebben om veel vragen uit de NOaA worden vastgesteld dat in welke mate de 1.0 te kunnen beantwoorden. De enige manier verkregen onderzoeksresultaten aansluiten op om de kans op de detectie van dergelijke sites te de geëxpliciteerde verwachting om modellen te vergroten, is te anticiperen op de potentiële kunnen verbeteren. Afwijkende resultaten aanwezigheid (bijvoorbeeld gebaseerd op kunnen betrekking hebben op de afwezigheid paleolandschappelijke informatie) in situaties van het verwachte, of de aanwezigheid van het waar terrestrische afzetingen uit het Pleistoceen onverwachte. Dit laatste zorgt uiteraard voor tot vroeg Midden-Holoceen aan snee komen. problemen wanneer dit tijdens het veldwerk Tijdens opgravingen kunnen meer problemen wordt vastgesteld, omdat dan aanpassing van opduiken. Steentijdvindplaatsen worden de strategie of methodiek noodzakelijk is en wat gekenmerkt door een strooiing van weer gevolgen heef voor de planning en vondstmateriaal waarin zones met hoge en lage middelen. Vooral deze factoren leiden dikwijls dichtheid voorkomen. In de context van Malta- tot besluit name die resten uit de vroege onderzoek ligt de focus meestal op het opgraven prehistorie degraderen tot ‘bijvangst’ en van zones met hoge dichtheden materiaal en waarover minimaal wordt gerapporteerd. Maar wordt geen geld ‘verspild’ aan relatieve lege het is juist deze bijvangst die verschijnselen met zones. Om de opbrengst van een opgraving te een lange resolutie (strooiing met een lage maximaliseren wordt doorgaans een gefaseerde dichtheid; geïsoleerde kuilhaarden; geïsoleerde strategie toegepast, waarbij de opgraving start pijlspitsen) vertegenwoordigd en die met het uitgraven van kleine vakken (gridcellen) 17 —

die relatief ver uiteen gelegen zijn. Verschillen in worden gezien als een manier om op vondstaantallen worden vervolgens gebruikt om gestructureerde wijze informatie te delen om zo clusters te begrenzen. Zones met lage vergelijkende analyses mogelijk te maken. dichtheden worden meestal niet onderzocht. Dit brengt twee problemen met zich mee. Naar een nieuwe onderzoeksagenda Ten eerste is het voorspellen (interpoleren) van Tijdens de uitvoering van ons onderzoek werd variatie in vondstdichtheden op basis van de NOaA 2.0 opengesteld door de RCE. Hoewel monsterpunten methodisch problematisch. de NOaA 2.0 bedoeld is om de belangrijkste Vondstverspreidingen zijn van een fundamenteel onderzoeksthema’s naar voren te halen en om ander karakter dan bodemkundige of de operationalisering ervan voor veldwerk te lithologische eenheden; de laatsten zijn faciliteren middels praktische suggesties, 3-dimensionale verschijnselen, de eerste zijn dat hebben veel vragen impliciet een hoge graad van als samenstellende elementen van een complexiteit. Verder blijken empirische aspecten bodemkundige of lithologische eenheid niet. Het verweven met aspecten van interpretatie, die de is belangrijk om vast te stellen dat intensief dagelijkse uitwerking ervan in de praktijk van onderzochte steentijdvindplaatsen (bijvoorbeeld Malta-onderzoek ingewikkeld maakt. Natuurlijk Hoge Vaart-A27, Dronten-N23, Ede-Kernhem en is het mogelijk om op basis van onze analyse een Schipluiden) grote verschillen in vondstdichtheid nieuwe set onderzoeksvragen te formuleren. We laten zien, zelfs over korte afstanden. Dit hebben immers laten zien dat Malta-onderzoek impliceert intrinsieke onzekerheid met een bijdrage heef geleverd aan de betrekking tot de kwantitatieve interpolatie van beantwoording van NOaA 1.0 onderzoeksvragen. monstergegevens. Ten tweede leidt de op Daar deze bijdrage echter beperkt is, zouden we kwantiteit gerichte selectie tot ongelijke en ons tevreden kunnen stellen met de constatering discontinue opgravingseenheden waarin clusters dat alle vragen relevant blijven, wat in zekere zin of vondstconcentraties dikwijls ‘onvolledig’ zijn ook het geval is. Maar het is ook duidelijk dat onderzocht. Dit belemmert het potentieel voor diverse NOaA 1.0 vragen de reikwijdte van de analyse van ruimtelijke aspecten. Er zijn veel Malta-projecten overstijgen. Het betref vragen technieken voor ruimtelijke analyse beschikbaar, in de ‘waarom’ categorie en deze zouden uit de maar de toepassing ervan vraagt om continue onderzoeksagenda geschrapt moeten worden vlakken die bij voorkeur ook lege zones als deze gericht moet zijn op de uitvoering van omvaten om randefecten te minimaliseren. Malta-onderzoek. Voor wat betref de analysefase van onderzoek Deze punten leiden tot een meer fundamentele hebben we vijf problemen naar voren gebracht. vraag: wat zou de rol moeten zijn van Malta en Ten eerste is de beschrijving en analyse van academisch onderzoek in de bredere context cultureel materiaal weinig gestandaardiseerd, van archeologisch veldonderzoek (dat gericht wat het gebruik van resultaten voor nieuw zou moeten zijn op het verzamelen van data die onderzoek in de weg staat. Ten tweede is er aan de basis kunnen staan van vraag-gedreven beperkt aandacht voor de functionele analyse kenniswinst), analyse en interpretatie? Het van aardewerk en lithisch materiaal. Ten derde is antwoord op deze vraag zou het vertrekpunt er een gebrek aan specifeke hoofdstukken moeten zijn voor de verbetering van de NOaA betrefende ruimtelijke analyse. Het vierde 2.0, die zonder twijfel een rol zou moeten en probleem is dat in projecten waarin vroeg- kunnen hebben in de kwaliteitszorg. Wij zijn er prehistorische resten als bijvangst zijn van overtuigd dat een meer gebalanceerde en gevonden, de inzet van specialisten (vrijwel) gerichte set van onderzoeksvragen kan volgen ontberen. Op de laatste plaats ontbreekt het aan uit een dialoog tussen de Malta-gemeenschap regionale vergelijkingen, als was het maar en de academische gemeenschap; daartoe kan omdat datasets dikwijls ontbreken in digitale deze studie als basis dienen. depots ondanks de wetelijke regelgeving. We stellen voor dat standaarden ontwikkeld Conclusies worden voor rapporten, die nu nog uitermate Onze analyse heef duidelijk gemaakt dat vrijwel divers zijn voor wat betref de presentatie van alle NOaA 1.0 onderzoeksthema’s en vragen gegevens. Standaarden worden vaak gezien als onbeantwoord of deels beantwoord zijn onwenselijke dwangvesten, maar ze moeten gebleven. Diverse vragen, vooral in de ‘waarom’ 18 —

categorie en waarvoor synthetiserende analyse werk dat al is uitgevoerd. Om dat potentieel af op een minimaal regionale schaal nodig is, te kunnen tappen moeten alle gegevens liggen buiten de mogelijkheden van Malta- beschikbaar komen, hoewel dan nog in veel projecten. Een onderzoeksagenda die bedoeld is gevallen teruggegaan zal moeten worden naar om Malta-onderzoek te inspireren en gericht is het materiaal. Het paradijs is dus niet verloren, op de kwaliteit van archeologisch onderzoek, althans, niet volledig. Maar het is noodzakelijk dient beter gestroomlijnd te worden met het om kaders te ontwikkelen die leiden tot potentieel en de praktijk van wat redelijkerwijs bruikbare en toegankelijke onderzoeksresultaten in een dergelijke context bereikt kan worden. uit Malta-projecten, om te voorkomen dat het Maar het is ook belangrijk dat we uit ons paradijs verder afrokkelt. En dat is de onderzoek de indruk hebben gekregen dat er verantwoordelijkheid van de archeologische een enorm potentieel beschikbaar is in al het gemeenschap als geheel. 19 1 Research framework —

jointly develop such policies at the regional level. 1.1 Introduction Provincial authorities also have their own policies – mostly of a more general nature – to which local or regional policies need to adhere. The essence of archaeological research is that Finally, national authorities are involved in the we learn about the human past and that we development and monitoring of legislation, and develop an understanding of how the behaviour they have an advisory role in the context of of, and choices made by people in the past have projects initiated by (semi-)governmental led to where we are today. It is a scientifc quest organisations. In addition, the national for knowledge that delivers the ingredients for authorities engage in projects where the narratives about what people were doing and archaeology to be impacted is considered to be thinking in a socio-cultural and environmental of national importance. context that difers from our own. And this With local authorities bearing the primary knowledge needs to be based on the responsibility for archaeological heritage, there is interpretation of mostly scarce archaeological no nation-wide policy on how to deal with the remains and traces of ancient landscapes legislative obligations. Indeed, policies can difer encapsulated in the subsoil. The subsoil is, from one municipality to another, even when they however, far from stable. A myriad of processes are adjacent. Hence, the practical execution of the lead to continuous change in, for instance, disturber’s obligation to conduct archaeological chemical or biological characteristics below the investigation is highly variable, and the fact that present-day land surface. Mostly invisible, these policy is implemented at the sub-national level dynamics result in progressive alteration of the can even lead to diferent approaches for a single archaeological and palaeoenvironmental site when that site straddles a municipal records. In contrast, physical disturbances, such boundary line. Commercial frms that have to as construction, agriculture and natural erosion conduct these investigations operate within an lead to a directly observable impact on the established heritage management cycle of archaeological heritage. inventorying, assessing signifcance, selecting Archaeological heritage is protected under how to deal with the site (excavation/in situ Dutch law. Since 2007, the Dutch heritage preservation/no action), and reporting (fg. 1.1). legislation has been part of the so-called Malta These frms also have to work within the legislation, which represents the constraints of time and money set by developers, implementation of the Valleta Convention which may lead to conficting interests between (1992). Renamed Wet op de archeologische archaeology (research) and development (saving monumentenzorg (literally: law on the care for on costs).1 In the context of signifcance archaeological monuments) in 2007, the assessment and subsequent selection, version 1.0 heritage legislation is implemented by the of the national research agenda for archaeology Cultural Heritage Agency of the Netherlands (Nationale Onderzoeksagenda Archeologie; (Rijksdienst voor het Cultureel Erfgoed), which is NOaA) is intended to serve as a national scientifc part of the Dutch ministry responsible for reference framework for decision making and a education, culture and science (Ministerie van starting point for research design. However, there Onderwijs, Cultuur en Wetenschap).The 2007 law is a growing tendency towards the development states that the ‘disturber’ has the obligation to of regional or even local research agendas, which investigate the (potential) presence of potentially undermines the achievement of a archaeological remains prior to the planned common baseline for the assessment of disturbance, and to subsequently take measures archaeological signifcance. to safeguard this heritage in situ or ex situ. A recently conducted evaluation of the Archaeological heritage is primarily the efectiveness of the heritage management cycle responsibility of municipal or other local under the Malta legislation has revealed that the authorities, who all need to have policies in NOaA 1.0 has come to be a directive for project place in order to embed archaeological outlines, known in Dutch as Programma van investigations in the authorisation process for Eisen (PvE), which are intended to defne any development that involves disturbance of the goals, strategy and methodology of the subsoil. Sometimes multiple local authorities excavations. However, the evaluation also 1 Groenewoudt & Peeters 2006. 20 —

Decision Making: Known Quality Significance Verification • protection Resources Assessment Evaluation •research

Archaeological New Assessed Heritage Resources Resources Selection Policy: • scientific aims & •public aims •preconditions Unknown Predictive Prospection Predictive •priorities (Expected) Resources modelling Models

Fig 1.1 The cycle of archaeological heritage management in the Netherlands explicitly involves both the known and the unknown (expected) heritage (from Groenewoudt & Peeters 2006).

revealed that reports lag behind with respect to In the following sections, we provide a the interpretation and synthesis of excavation general outline of the analytical framework and results, relative to the research themes and practical approach that form the basis of this questions in the NOaA 1.0. This raises the study. Chapter 2 presents further details about question: What can we learn from all these the development of a knowledge system, the investigations that have found their way into assessment of reports, the geographical and reports? chronological representativeness of selected For this purpose, the Cultural Heritage reports, and data mining. The chapters 3, 4 and 5 Agency of the Netherlands (Rijksdienst voor het present the results in connection with the NOaA Cultureel Erfgoed; RCE) commissioned a project, 1.0 research questions. These chapters will, undertaken by the Groningen Institute of however, also include critical comments with Archaeology (GIA), which aims at a synthetic regard to problems encountered. Hence, we not analysis of (mostly) excavation results concerning only address the aspect of knowledge gain, but early prehistory (here defned as the Palaeolithic also the aspect of knowledge gaps. The critical to the Middle Neolithic). The project is part of the discussion will be continued in chapter 6, and RCE’s Valleta harvest (Oogst voor Malta) project, suggestions for improvement will be made in which resides under the ‘archaeological connection with the practice of development- knowledge kit’ programme (Kenniskaart led research. Chapter 7 draws general Archeologie).2 The aim of the programme is to conclusions from our study and presents a look develop tools or utilities to support heritage into the future of the national research agenda. management decision making from a national perspective. In the particular case of our study, the aim is to establish to what extent research 1.2 Analytical framework questions from the NOaA 1.0 can be answered at a synthetic level, drawing from the published reports resulting from development-led projects. As mentioned above, the aim of this study is to The results of our study will feed in to an bring together and analyse excavation results, in evaluation of the NOaA research questions and order to deepen and broaden our knowledge will eventually help inform a proposal for changes about the early prehistoric past of the to the research questions for NOaA 2.0, which has Netherlands, relative to the research themes and 2 Lauwerier et al. 2017. recently been launched (see section 7.2). questions in the NOaA 1.0. The broad range of 21 —

research questions listed in the NOaA 1.0 are of which, in turn, can be approached from four grouped under fve thematic headings, which analytical spatio-temporal dimensions can be translated as follows: (a) human (section 1.2.4). colonisation and early occupation history of the The overarching themes (fg. 1.2) connect to Netherlands; (b) land use and setlement three broad arrays of the ‘human realm of systems; (c) food economy, relationship between existence’: people and materials (section 1.2.1); humans and environment; (d) burial and people and space (section 1.2.2); and people and deposition of human remains; (e) cultural environment (section 1.2.3). All three themes traditions/social relationships and interaction. have a direct bearing on the study of human Within these themes, for the archaeological time behaviour in the past on the basis of period covered by this study, a chronological archaeologically recorded data and subsequent distinction is made between Middle Palaeolithic, interpretation of the paterning found within. Late Palaeolithic and Mesolithic, and Neolithic. It is important to note here that we use the In addition, a number of research questions refer division into archaeological periods (Palaeolithic, to particular archaeological phenomena (e.g. pit Mesolithic, Neolithic, and further subdivisions) hearths), feldwork strategies, or the analysis of in a strictly chronological sense, not to refer to data and fnds. sets of material culture or human lifeways (e.g. A synthetic analysis of the reported economy, socio-cultural relationships, ideology). excavation results in direct reference to the NOaA 1.0 research questions will without doubt lead to deeper insight into specifc questions, 1.2.1 People and materials yet also risks missing the connections between various aspects of interest. In fact, this is the crux of the problem with the body of reports The study of the relationships between people that have resulted from development-led and materials is perhaps the most obvious from archaeology: Questions are answered one by an archaeological perspective. Our primary one, but seldom in tandem or within a broader source of archaeological information consists of thematic perspective. In order to avoid such objects made out of a broad range of natural fragmentation of knowledge, we will approach resources, as well as features resulting from the problem in a way that permits us to human interference. Not surprisingly, fnds and simultaneously broaden and deepen our insights anthropogenic features receive a great deal of under three overarching diachronic themes, each atention in reports in terms of both description

WĞŽƉůĞ Θ DĂƚĞƌŝĂů

WĞŽƉůĞΘ^ƉĂĐĞ

WĞŽƉůĞΘŶǀŝƌŽŶŵĞŶƚ

Fig 1.2 The three overarching themes that are central in this study connect people to diﬀerent scales of interaction and are approached from multiple scales in terms of space and time. 22 —

and analysis. Mostly, this results in a typological, of behaviour.7 Studies aimed at increasing our technological and/or functional classifcation of insight into the ‘structure’ of palimpsests in (assemblages of) materials to support a order to develop a beter understanding of the chronological and functional characterisation of underlying processes of palimpsest formation a site or a complex of sites. However, other are few, but they ofer hopeful perspectives.8 aspects of ‘materials’ also need to be taken in The application of complementary analytical consideration – for instance, the role of potery techniques derived from diferent theoretical and grinding stones as socio-ideological factors and methodological concepts has recently in connection with the preparation of food.3 shown that surprisingly useful results can be Another example is the role of objects in ritual obtained, for instance, in the case of several Late contexts,4 or the role of particular objects – Neolithic setlement sites in the province of which, from an archaeological perspective, may Noord-Holland.9 So, instead of considering not even be considered ‘special’ – to connect palimpsest sites/distributions to be impossible geographically separated places in the to understand in terms of spatial behaviour, landscape.5 Thus, the ‘materiality’ of various eforts can be made to turn such confgurations aspects of people’s behaviour can take diferent into valuable sources of archaeological forms. However, the question is: Is the range of information. The fact that it may be difcult, if interpretive possibilities entirely considered not mostly impossible, to identify discrete within development-led archaeology at the spatial entities that can securely be considered research design stage (that is, in the Programma ‘contemporaneous’ does by no means imply van Eisen), or given the proper atention in that palimpsests are a dead-end street. reports? The answer is most probably ‘No’, because less straightforward interpretations ofen depend on contextualisation within a 1.2.3 People and environment synchronic and diachronic and spatial framework that exceeds that of the excavated ‘window’ (i.e. the site). Just as the previous two themes are the ‘bread and buter’ of archaeological research, the study of the relationship between people and 1.2.2 People and space environment is another theme that is prominent in Dutch archaeology, including within development-led archaeology. ‘Landscape’ The study of the spatial behaviour of people fgures as a structural element in many project involves analysis at the intra- and/or inter-site outlines, but it seems to be variably perceived level. Intra-site analyses focus primarily on the and investigated. Also, in the context of early identifcation of spatial structures (e.g. houses, prehistoric archaeology, the concept of 3 E.g. Devriendt 2014; Raemaekers et al. 2013. storage facilities, wind breaks) or zones with a landscape is frequently considered to be 4 E.g. Louwe Kooijmans 2003; Van Gijn distinctive functional signature (i.e. activity synonymous with natural scenery. Landscape 2010. 5 E.g. Kamstra et al. 2016. areas). Inter-site analyses aim at the thus has a connotation of a state of being, rather 6 Within this theme, we will only partly identifcation of functional diferences and than a dynamic whole. Consequently, litle address the inter-site level of use of space. The number of new data-points relationships between sites, for instance, to atention is paid to the interaction between that has been acquired from 2005 to reconstruct ‘setlement systems’ or draw people and the landscape. In our opinion, the 2014 is most probably too restricted to generate new insight into, for instance, inferences about mobility strategies. term environment seems more appropriate than choice of setlement location. In Development-led archaeology is particularly the term landscape in this context, frst, because addition, much development-led research is prompted by survey results concerned with the intra-site level of spatial people experience their surroundings as they that are rooted in existing models of behaviour.6 However, major problems emerge move and, second, because the environment is spatial behaviour at the (supra-)regional scale, hence leading to self-fulﬁlling from the palimpsest character of the great multidimensional in terms of agency, as expectations. Any new insights will majority of sites that are normally selected for perceived by individuals who operate within a therefore be of an anecdotal nature. 7 Bailey 2007. excavation. High-density scaters of material particular socio-ideological and ontological 8 Gordon 2009; Peeters 2007. remains and features typically result from context. By placing these interactions in the 9 Nobles 2016. Fokkens, Steﬀens & Van As 2016 have expressed their doubts about long-term, repeated use of a particular place, foreground, the spatial and temporal scales the interpretive possibilities of these and these are difcult to decipher and become more pertinent, as not all changes in the particular datasets (see section 4.3.1 for further discussion). understand in terms of meaningful components environment will have been of relevance at the 23 —

human scale. Millennial- and centennial-scale environmental dynamics, for instance, due to climate change, may be of archaeological Environment

interest for studying long-term paterns of Regional behavioural change, but they are probably less informative for studying regional paterns of Space subsistence at the scale of several generations. Also, the use of the environment is not simply bound to economic factors, but is also related to historic and cosmological factors, which can be Materials expressed through the duration and particular use of localities.10 The question, then, is: At which scales is landscape – or, as we prefer to term it, environment11 – considered in the context of development-led archaeology? Local 1.2.4 Spatio-temporal dimensions

The themes described above will be connected with an analytical framework that distinguishes four space–time dimensions: • Local and short-term (use of locations); • Local and long-term (use-history of locations); • Regional and short-term (cultural geography); Short-term Long-term • Regional and long-term (cultural biography). The local scale refers to the spatial units Fig 1.3 The themes and spatio-temporal dimensions combine to form 12 knowledge domains. (‘windows’) as investigated through excavation. The regional scale refers to some defned spatial entity which exceeds the local scale, and corresponds to an area the size of a so-called ‘archaeological landscape’. An archaeological 1.2.5 Knowledge domains landscape is defned by the national heritage agency as a landscape with particular geological histories and subsoil characteristics, or as a The connection of the three diachronic themes landscape containing archaeological phenomena and four spatio-temporal dimensions of analysis with characteristics that are related to the lead to twelve so-called knowledge domains, geological history of one particular landscape.12 each of which covers diferent aspects of These are, however, of variable size. For our behaviour taken from variable perspectives 10 Cf. Amkreutz 2013; Peeters 2007, 2009a, purpose, the synthetic analysis of excavation (fg. 1.3).13 The four dimensions, in theory, apply 2009b. 11 Landscape is a vague concept, referring to results at the two local dimensions involves to each overarching research theme. However, an undeﬁned stretch of land which is research questions that refer to the site and its specifc NOaA 1.0 research questions can be of characterised by a selected range of features (e.g. ‘mountain landscape’; direct surroundings. Synthetic analysis at the two relevance to more than one research theme, and ‘river landscape’; ‘hedgerow country’). regional dimensions involves research questions will thus contribute to more than one knowledge Environment refers to the broader range of phenomena that a landscape is that require a broader contextualisation of data domain. In the thematic chapters (chapters 3, 4 composed of, such as trees, animals, to permit interpretations at a non-local scale. and 5), the knowledge domains will not explicitly terrain features and water. These are also the phenomena that are studied more Hence, the various analytical scales provide a serve to structure the discussion. However, we directly, e.g. through the analysis of tool for investigating paterns within the will use these domains in our broader discussion botanical macroremains, pollen and microfauna and through relief mapping. archaeological record. The reality in the past no of what development-led archaeology has 12 Rensink et al. 2015. doubt consisted of a complex of interwoven contributed to our understanding of the past 13 For example: People and space – Local and short-term; People and space – space/time dimensions. (chapter 6). Local and long-term. 24 —

Although these quality-driving factors infuence 1.3 Report selection the quality of the research and the resulting reports, an analysis of these factors would require a diferent approach and would involve In agreement with the RCE request, we made a issues that are beyond the scope of this study. selection amongst the available reports for the purpose of our synthetic analysis. Only reports with the following characteristics were eligible 1.4 Practical approach for our analysis: • The report describes the results of a commercial, development-led feldwork project. This study was conducted in two phases. The frst • The report was published between 2005 and phase involved a survey of the available reports to 2014. evaluate their potential as a useful source of • The subject mater of the report includes information relative to (a) the research themes (as material from the Palaeolithic up to and noted above) and (b) a set of related NOaA 1.0 including the Middle Neolithic (i.e. early research questions that seemed viable for a prehistory). synthetic analysis. The second phase involved the The basis for the selection procedure was formed actual synthetic analysis including an in-depth by the national Archis database, in which evaluation of the reports as a source of information. information about all archaeological investigations In the frst phase, an initial list of questions and observations is stored (e.g. administrative was compiled by our specialist group, based on data; spatial, geological, and typological aspects; the questions provided in the NOaA 1.0. This archaeological interpretation).14 Reports about resulted in a list of no fewer than 72 questions, specifc feldwork projects are linked to the system, many of which were subsequently combined although in reality this appeared to not always be (being interrelated research questions) or the case. Missing reports could, in some cases, be removed based on whether we thought it was obtained from the DANS-EASY database, which feasible for them to be answered within the serves as the digital repository for research data context of development-led research. The later and documentation.15 category particularly involved broad questions The reports were screened for their potential that require regional-scale inter-site analysis. The usefulness for this study based on the three potential for research questions to be answered criteria above and, beyond that, based on depended, of course, on the information provided whether, as a corpus, they are able to provide in reports. The fnal list comprised 34 research coverage of the following: questions, each of which was connected with a set • survey versus excavation (trial trenches, of ‘aspects’ (i.e. keywords or terms such as ‘fint’, full excavation); ‘Wommersom quartzite’ or ‘grave’). With the • early prehistoric fnds anticipated versus not aspects linked to the publications and questions, anticipated (‘by-catch’); the reports could be easily sorted for their richness • academic involvement versus no academic of information (sorted by number of linked involvement; aspects) or for specifc categories of information 14 An offline version of the Archis database was made available by the Cultural • overall ‘quality’: (a) generic versus specifc (e.g. presence of human remains). Likewise, Heritage Agency (Rijksdienst voor het (data based); (b) clarity of presentation (by because we were then able to list the number of Cultureel Erfgoed [RCE]). This version allows for more rigorous and flexible subject, by excavation unit); (c) level of publications that were available in connection data analysis and querying than does synthesis (integrated research results, with each specifc research question, it was now the online front-end query environment (htps://archeologieinnederland.nl/ question-by-question answers). clear to us how answerable certain questions bronnen-en-kaarten/archis). Archis The assessment process is further described in would be. The process and outcome of this stands for Archeologisch Informatiesysteem (archaeological section 2.3. It should be mentioned here that the exercise are discussed in further detail in chapter 2. information system). assessment did not include an analysis of The second phase involved the retrieval of 15 DANS stands for Data Archiving and Networked Services. It is an institute of potential correlations with preconditions or information in connection with specifc research the Dutch royal academy of sciences restrictions set by the clients (e.g. municipal questions, and analysis of that information to the (Koninklijke Nederlandse Academie van Wetenschappen; KNAW) and the authorities, the state or private companies), the extent that the quality and quantity of data Netherlands Organisation for Scientific project outlines and budgets, and the allowed. The outcome of this synthetic analysis is Research (known by its Dutch acronym, NWO). commercial archaeological frms involved. presented in chapters 3 to 5. 25 2 Knowledge system and report — assessment

2.1 Introduction

The preceding chapter provided an outline of the analytical framework and a general characterisation of the reports resulting from development-led research, as well as a brief description of the practical approach. This chapter presents a detailed description of the Fig 2.1 Schematic representation of the aspect-related structure of the knowledge system, the process connection between research questions and of report assessment, data mining procedures, publications (reports). and problems encountered during data mining.

between our research questions and the reports, 2.2 Knowledge system we compiled a list of 50 ‘aspects’ that we also deemed to be prerequisites for formulating answers to the research questions (table 2.1). In order to enable structured searching of reports These aspects range from physical archaeological in connection with the analytical framework and evidence to feld methods and applied analyses. the interrelated NOaA 1.0 research questions, as The screening of reports on the basis of these outlined in chapter 1, it was necessary to build a aspects indicates which questions can potentially fexible database structure (fg. 2.1). The structure be answered based on the information provided was designed to allow for the selection of reports in each report. Because the database structure concerning sites that potentially provide permits querying in both directions, a query on a information on particular topics, such as raw particular research question generates a list of material procurement, or the function of pit potentially relevant reports, and the selection of hearths. To enable us to make connections a report generates a list of potentially answerable

Table 2.1 Aspects initially used to connect research questions to reports.

ID Aspect ID Aspect ID Aspect

1 artefacts: fint 18 features: storage 35 landscape data

2 artefacts: stone 19 features: post holes 36 dating: OSL

3 artefacts: ceramics 20 features: hearths 37 dating: 14C

4 artefacts: organic 21 features: waste pits 38 site function

5 artefacts: bone 22 features: agricultural 39 site function: activity areas

6 resources: fint 23 Structures 40 site function: setlement

7 resources: stone 24 spatial data: artefact distribution plot 41 site function: cemetary

8 resources: ceramics 25 spatial data: features 42 structures: houseplans

9 fint technology 26 structure: burial/grave 43 artefacts: other

10 fint typology 27 grave goods 44 collection method: trench

11 ceramic decoration 28 special deposition 45 collection method: grid

12 faunal remains 29 XRF analysis 46 collection method: point

13 human remains 30 micro-morphology 47 collection method: other

14 botanical remains: macro plant 31 organic residue analysis 48 spatial data: DANS

15 botanical remains: pollen 32 use-wear analysis 49 spatial data: further analysis

16 botanical remains: seeds, fruits 33 Palaeoecology 50 spatial data: palimpsest

17 botanical remains: wood 34 Palaeogeography 26 —

questions. Furthermore, the connection between each research question and report was stored in relationship to NOaA 1.0 research themes and the overarching themes, as defned under the umbrella of our analytical framework (see chapter 1). This permited further thematic sorting, which made it possible to generate lists of reports that are of potential relevance to thematic sets of research questions. The fnal set of 34 research questions that were distilled from the NOaA 1.0 variably connect to the overarching themes. Appendix I provides a listing of the NOaA 1.0 questions in Fig 2.2 Number of NOaA 1.0 research questions in connection with the research themes as defned relation to the overarching research themes. in chapter 1. As is apparent from fgure 2.2, the great majority refer to the theme ‘people and further confgurations are possible, since materials’, due to the fact that the analysis of research questions can be related to one or material culture is at the heart of many more of these analytical dimensions. The archaeological studies (see section 1.2.1). possibility to connect individual research The themes ‘people and space’ and ‘people and questions to more than one overarching environment’ each cover somewhat less than research theme, spatio-temporal dimension, or half of the research questions. When we break knowledge domain (fg. 2.3), demonstrates the these numbers down, we see that 19 questions difculty of providing a structured framework relate to more than one theme, which leaves for synthetic analysis based on the existing 15 questions unique to a single theme, mostly construct of the NOaA 1.0. We will discuss this that of ‘people and materials’. Combined with problem further in chapter 6 in connection with the spatio-temporal dimensions (section 1.2.4) the research agenda, feldwork and reporting.

2.3 Report assessment Long-term Local 1 1 1 2.3.1 Report selection and preliminary assessment 1 3

The complete list of reports in the Archis 6 6 10 database was frst fltered by publication date (between 2005 and 2014), which resulted in a list 3 of c. 750 reports. Next, entries that included a starting date of occupation that was later than the Middle Neolithic were deselected, because these fall outside the scope of this project. Regional Short-term Unfortunately, this did not eliminate all reports that were unsuitable for our analysis. Querying for early prehistoric sites only proved somewhat problematic, as information about the dating of a site and the archaeological periods actually 2 represented is neither specifcally nor systematically provided within Archis. Instead, only the starting period and end period are Method given. Provided a site covers only a restricted Fig 2.3 Number of NOaA 1.0 research questions in relation to the spatio-temporal dimensions. and well-identifable time range, this does not 27 —

lead to any problem in this instance. However, fool-proof solution, because a certain number of uncertainty about the dating of the phenomena survey reports got ‘trapped’ in the fnal uncovered ofen leads to a wide chronological selection. We kept them in the fnal selection as atribution. Entries of non-diagnostic fint a ‘random’ sample of survey reports. In addition, artefacts, for instance, are ofen given the broad we purposefully added reports on some rather range of Palaeolithic up to Iron Age. Other fnds sizeable surveys (e.g. Well-Aijen) because of categories, such as charcoal and stone artefacts their anticipated value for our study.16 made of rocks other than fint, are ofen given a Furthermore, we added some reports on date range from the Palaeolithic up to the post- archaeological watching briefs, e.g. for the medieval period. In other cases, excavations purpose of pipelines or channels (e.g. Olieveld may have produced remains from multiple Schoonebeek),17 because these potentially periods. Consequently, queries involving start provide insight into the archaeological dates within one or more of the periods covered signifcance of areas that are normally not by the category ‘early prehistoric’ are not excavated due to the absence of setlements, necessarily going to return results for reports which remain the prime concern of most covering early prehistoric remains. In order to development-led work. reduce the amount of ‘noise’ within the selected Appendix II lists the titles of the selected set of reports, manual screening was necessary. reports. In the following text, general references As the aim of the project is to conduct a to reports will consist if the publication synthetic analysis of results from development- identifcation number (PUBID) used in appendix led work in connection with the NOaA 1.0, it was II; where reference is made to specifc topics or decided to give priority to reports that present sections, we provide the full reference, which is the results of excavations. In contrast to surveys listed in the bibliography. The fnal list comprises (auger sampling, test trench or test pit 139 publications of varying size and subject sampling), excavations mostly will have covered mater – 85 on excavations, 30 on surveys, and extended and continuous surfaces and are 24 on archaeological watching briefs. A total of expected to have produced high-resolution 84 reports are explicitly concerned with early data, which potentially allows for more solid prehistoric archaeology. The remaining 55 interpretations. This, however, does not imply reports have a focus on a later periods; early that surveys provide no valuable information prehistoric archaeology is considered of from which to develop new insights. secondary importance. Observations can potentially add to the analysis A preliminary assessment of the reports for of, for instance, land-use paterns on a wider their information potential was conducted on geographical scale, although the interpretive the basis of the number of linked aspects possibilities are expected to be limited due to (fg. 2.4). These aspects involve archaeological the lack of sufcient data to put these in context. phenomena (materials, features) as well as However, to assess the potential information methods and approaches (e.g. 14C-dates, value of survey results within the context of this optically stimulated luminescence (OSL) dates, project, detailed manual examination of such pollen analysis, excavation method). Obviously, reports would have been required. Because we the number of linked aspects is no direct could not be certain a priori about the actual indication of information value: reports that do gain, we decided that manual screening of not contain many aspects can contain a lot of survey reports would use up too much of our information on just one aspect, for instance, budget. Consequently, reports on surveys were a site with a single-period fint assemblage deselected. Again, this did not eliminate all without many other preserved fnds categories. reports that were unsuitable for our analysis, Nonetheless, it can be argued that reports with because the two types of reports were not many linked aspects potentially ofer beter always correctly categorised. The solution was possibilities for contextualisation of specifc data to remove those entries from the list that due to the broader range of topics that can be contained keywords pertaining to surveying in covered. As expected, the number of linked 16 Appendix II: PUBID 592. the report title. Even this turned out to be not a aspects is generally higher for reports on 17 Appendix II: PUBID 482, 483, 485, 489. 28 —

Fig 2.4 Number of aspects per report/site. 29 —

Fig. 2.4 Continued number of reports per aspect (lef), and number of research questions per aspect (right). 30 —

Ϯϱ ϯϬ

Ϯϱ ϮϬ

ϮϬ ϭϱ ^ƵƌǀĞǇ ϭϱ ^ƵƌǀĞǇƐ ϭϬ ǆĐĂǀĂƚŝŽŶ ǆĐĂǀĂƚŝŽŶƐ ϭϬ

ϱ ϱ

Ϭ Ϭ ϭͲϱ ϲͲϭϬ ϭϭͲϭϱ ϭϲͲϮϬ ϮϭͲϮϱ ϮϲͲϯϬ ϯϭͲϯϱ ϯϲͲϰϬ ϰϭͲϰϱ ϭͲϱ ϲͲϭϬ ϭϭͲϭϱ ϭϲͲϮϬ ϮϭͲϮϱ ϮϲͲϯϬ ϯϭͲϯϱ

Fig 2.5 Number of reports of surveys and excavations per aspect Fig 2.6 Number of reports of surveys and excavations per aspect cluster frequency class. frequency class.

excavations compared with reports on surveys analysis and OSL dating, are missing in the survey and archaeological watching briefs (fg. 2.5). As reports. This leads to the conclusion that the fgure 2.5 shows, initially, the number of information potential of survey reports will excavation reports increases in proportion to the generally be lower than that of excavation number of aspects, but drops when more than reports, and that it will be restricted to material 20 aspects are linked. The number of survey fnds categories. reports drops when more than ten aspects are linked. In order to get a clearer picture of the relationship between the nature of feldwork 2.3.2 Geographical and chronological (excavation, survey) and the aspects in reports, coverage we clustered some aspects in order to reduce the number of aspects (table 2.2). Figure 2.6 shows a slightly diferent distribution in comparison with It became clear from our study of the 139 reports fgure 2.5, but the trends remain the same with that development-led feldwork conducted regard to survey and excavation. between 2005 and 2014 resulted in the discovery Because aspects are linked to the reports in of at least 139 early prehistoric sites; some our database, it was possible for us to assess projects (e.g. two large-scale projects near whether a report had the potential to help us Knooppunt Hatemerbroek,18 as well as the answer a specifc research question. The reports Well-Aijen survey19) cover complex palimpsest were sorted based on their inclusion of these confgurations of overlapping and merged sites. aspects – such as hearths or graves, dwelling For the purpose of the present study, it is structures, potery, botanical remains, faunal important to assess the potential of this body of remains, or human remains. A notable diference sites for answering research questions in a wide between survey and excavation reports is the geographical seting. The distribution of these absence of burials/cemeteries and built structures sites can potentially be of interest for the in the former (table 2.3). This is to be expected, development of models of variability in since burials and cemeteries, being relatively rare landscape use. In addition, the analysis of phenomena, have less chance of being recovered diachronic paterns of site distribution and the in the context of surveys, even in test trenches. nature of activities can inform us about Built structures are normally identifed once long-term dynamics of landscape use. larger surfaces have been investigated, but are in fact rather exceptional in early prehistoric Geographical distribution contexts. Hence, survey reports are primarily Figure 2.7 shows that the sites discussed in the connected with common fnds categories, such as 139 reports are unevenly distributed. The majority fint, stone, ceramics, macrobotanical remains, as are located in the eastern half of the Netherlands, well as categories of features, such as postholes with a cluster of sites located in the western part and hearths. With regard to methods and of the country, roughly between Leiden and approaches, less common specialisations, such as Roterdam. The geographical distribution of 18 Appendix II: PUBID 259, 260. 19 Appendix II: PUBID 592. X-ray fuorescence (XRF) analysis, organic residue these sites is primarily infuenced by three factors. 31 —

Table 2.2 Aspect clusters in relation to initial aspects.

Type Cluster ID Aspect cluster Aspect ID Aspect Evidence 1 fint 1 artefacts: fint 6 resources: fint 9 fint technology 10 fint typology 2 stone 2 artefacts: stone 7 resources: stone 3 ceramics 3 artefacts: ceramics 8 resources: ceramics 11 ceramic decoration 4 organic artefact 4 artefacts: organic 5 artefacts: bone 5 other artefact 43 artefacts: other 6 macrobotany 14 botanical remains: macro plant 16 botanical remains: seeds, fruits 7 wood 17 botanical remains: wood 8 faunal remains 12 faunal remains 9 human remains 13 human remains 10 burial/grave 26 structure: burial/grave 27 grave goods 11 cemetery 41 site function: cemetary 12 special deposition 28 special deposition 13 hearths 20 features: hearths 14 pit / feature 18 features: storage 21 features: waste pits 22 features: agricultural 15 post holes 19 features: post holes 16 activity area 39 site function: activity areas 17 structure 23 structures 18 house plan 42 structures: houseplans 19 setlement 40 site function: setlement 20 site function 38 site function Methods and approaches 21 ecology 15 botanical remains: pollen 33 palaeoecology 22 palaeogeography 34 palaeogeography 23 landscape analysis 35 landscape data 24 micro-morphology 30 micro-morphology 25 XRF analysis 29 XRF analysis 26 OSL dating 36 dating: OSL 27 14C dating 37 dating: 14C 28 organic residue analysis 31 organic residue analysis 29 use-wear analysis 32 use-wear analysis 30 collection method: trench 44 collection method: trench 31 collection method: grid 45 collection method: grid 32 collection method: point 46 collection method: point 33 collection method: other 47 collection method: other 34 display: distribution plot 24 spatial data: artefact distribution plot 35 display: features 25 spatial data: features 36 further spatial analysis 49 spatial data: further analysis 37 DANS data available 48 spatial data: DANS 38 spatial data: palimpsest 50 spatial data: palimpsest 32 —

Table 2.3 Frequency of excavation and survey reports per aspect cluster.

Aspect cluster ‘evidence’ N survey reports N excavation reports

Flint 42 79

Stone 20 60

Ceramics 12 43

Organic artefacts 2 6

Other artefact 0 4

Macrobotany 9 35

Wood 7 29

Faunal remains 9 23

Human remains 1 7

Burial/grave 0 5

Cemetery 0 4

Special deposition 1 2

Hearths 11 29

Pit/feature 9 45

Postholes 4 19

Activity area 8 27

Structure 0 10

House plan 0 7

Setlement 0 12

Site function 3 33

Aspect cluster ‘method/approach’ N survey reports N excavation reports

Ecology 9 27

Palaeogeography 37 76

Landscape analysis 18 57

Micromorphology 4 18

XRF analysis 0 3

OSL dating 1 9

14C dating 16 45

Organic residue analysis 0 5

Use-wear analysis 2 10

Collection method: trench 39 76

Collection method: grid 15 36

Collection method: point 3 17

Collection method: other 5 3

Display: distribution plot 19 47

Display: features 20 72

Further spatial analysis 1 16

DANS data available 22 42

Spatial data: palimpsest 3 13 33 —

Fig 2.7 Geographical distribution of all selected sites/reports (background: 'archaeological landscapes' as deﬁned by the Cultural Heritage Agency of the Netherlands). 34 —

9000 BC 5500 BC

3850 BC 2750 BC 1500 BC

500 BC AD 100 AD 800

AD 1500 AD 1850 AD 2000

Fig 2.8 Time series of palaeogeographical reconstructions of the Netherlands, showing the major landscape changes throughout the Holocene (from Vos 2015). The maps make clear that the present-day situation is the result of far- reaching natural and anthropogenic processes, which have largely inﬂuenced the formation of the archaeological landscape in terms of diﬀerences in the nature of initial patern formation (location, materials, features), as well as the subsequent transformation of initial paterns in terms of preservation and detectability. 35 —

Fig 2.9 Distribution of deselected reports and observations made between 2005-2014 (background: ‘archaeological landscapes’ as deﬁned by the Cultural Heritage Agency of the Netherlands).

The frst factor is the geological history of relevant for the central and western parts of the the Netherlands, which has a considerable efect Netherlands. On the other hand, processes of on site discovery. On the one hand, processes of large-scale erosion have resulted in the continuous sedimentation and peat growth have disappearance of early prehistoric land surfaces resulted in traces of early prehistoric activity and layers of sediment, notably in the south- having been covered with thick layers of clay, western (province of Zeeland) and north- sand and peat, rendering most prehistoric western (province of Noord-Holland) parts of archaeology inaccessible. This is particularly the Netherlands. This erosion is particularly due 36 —

to Holocene coastal dynamics (fg. 2.8), which (fg. 2.7) are not necessarily devoid of are partially infuenced by human activity.20 As a archaeological observations and feldwork result, traces of early prehistoric activity have (fg. 2.9). Here the history of landscape largely disappeared in these areas; however, exploitation and historical development – where this is not the case, early prehistoric land medieval and post-medieval peat reclamation in surfaces and archaeological layers are deeply the western part of Noord-Brabant, for buried, up to 20 m below Amsterdam ordnance instance – may (partly) explain the absence of datum (OD). For this reason, only a small sites in that area in our database.22 This is not to number of sites within our sample are located in say that none of the sites pre-dating these large- the provinces of Zeeland, Noord-Holland and scale reclamation activities have survived, only Utrecht or in the coastal regions of the provinces that there is a fair chance that many will have of Friesland and Groningen. In geological terms, been heavily disturbed.23 Archaeological surveys the presence of sites between Leiden and in such areas will ofen result in the discovery of Roterdam is mainly due to the presence of Late disturbed contexts, which easily leads to the Pleistocene and Early to Middle Holocene sand decision not to conduct an excavation, because dunes – river dunes and coastal dunes – in the knowledge gain is expected to be limited in Rhine-Meuse estuary and along the coast.21 these cases.24 The decision whether or not to These sand dunes are ofen found close to the excavate is taken by policy-makers, in many present-day surface and are within reach of cases at the municipal or provincial level, and archaeological and geological surveying not all of these policymakers are supported by 20 Vos 2015. methods. an archaeological advisor. 21 The sand dunes are known to be rich in The second factor infuencing the In addition, the process of archaeological archaeological remains (see e.g. Amkreutz 2013; Louwe Kooijmans geographical distribution of sites is the location inventory making in the Netherlands is strongly 2001a, 2001b) and are therefore subject of planned development. This is wholly to be infuenced by the use of predictive models, such to survey. 22 Leenders 1996. expected, as the focus of this project concerns as the map of archaeological potential 23 An example is the Mesolithic site of the results of development-led archaeology. (Indicatieve Kaart van Archeologische Waarden; Tilburg-Kunstcluster, where the heavily disturbed remains of a Mesolithic Linear paterns in the distribution refect the IKAW).25 This also plays a role in decision campsite were discovered within an area trajectories of railways and pipelines, as well as making. Zones with low potential values are showing traces of (post-)medieval reclamation activities (Dijk et al. 1994). foodplain adjustments. It is important to bear in frequently – although certainly not always – 24 A recently published overview of the mind that such projects are generally initiated selected for development purposes, in the hope archaeology of the western portion of the province of Noord-Brabant clearly and fnanced by the Dutch state. Archaeological of avoiding costly archaeological investigation. demonstrates that the overwhelming work in the context of these projects is ofen However, because most models of majority of investigated sites date to later prehistoric and early historic times subject to involvement of the national cultural archaeological potential rely on relationships (Ball & Van Heeringen 2016). Burial heritage agency. Examples are the construction amongst archaeological observations and structures, house plans and other structures (e.g. wells) that are of the A73 highway and the foodplain geomorphological features and soil characterised by deep features (pits, adjustments (known in Dutch as the characteristics,26 there is an inherent risk of false trenches, posts) form the majority of the recovered phenomena. Remains Maaswerken) north of Maastricht (province of quantitative correlations and model self- from early prehistoric times are Limburg), which have led to numerous fulflment due to a lack of independent model extremely scarce. 25 Groenewoudt & Peeters 2006. The IKAW excavations. The building of houses and the testing (validation).27 In some cases where rich map, version 3 of which was published development of business parks also form an archaeological sites are found during survey, an in 2007, is only sporadically used today, as it has been replaced by maps infuential factor. Not all regions are equally alternate location for development is chosen, or produced on a regional or a municipal dynamic in this respect, and not all local measures are taken to preserve the site(s) in scale. Because the responsability for archaeological heritage lies in the hands authorities are equally concerned about situ.28 of local authorities, this is logical. archaeological heritage. Nature reserves, such as This combination of factors – geological However, the basic principle of the IKAW, which distinguishes zone of low, the national parks De Veluwe (province of history, present-day development, and decision medium and high potential, has been Gelderland) and the Utrechtse Heuvelrug making – underlies the geographical distribution maintained. 26 Deeben et al. 2002. (province of Utrecht), are devoid of of development-led surveys and excavations, as 27 Kamermans et al. 2009; Van Leusen & development. Other areas primarily have an shown in fgure 2.7. The basic layer of the maps Kamermans 2005; Verhagen 2007. Moreover, the archaeological record agricultural or recreational function and see only shows the ‘archaeological landscapes’, as seems inherently unﬁt for the purpose limited urban development. distinguished by the National Heritage Agency, of validation of predictive models (Brouwer Burg et al. 2016; Lovis 2016; The third factor, that of decision making, which are defned on the basis of landscape Peeters & Romeijn 2016; Whitley 2016). renders the picture far more complex. Empty features (geogenesis and landforms) and 28 Van der Reijden et al. 2011. 29 Rensink et al. 2015. regions in the distribution map of selected sites archaeological characteristics.29 Within each 37 —

‘archaeological landscape’, further landscape the Meuse River terraces, which describe yet zones are distinguished, which basically another archaeological landscape. The valley is represent characteristic landforms.30 We will variable in width and sometimes deeply incised; briefy consider the distribution of surveyed and in its wider parts, the valley foor consists of an excavated locations in connection with four of undulating foodplain with remnant gullies and these ‘archaeological landscapes’ and the small river dunes. Local diferences in elevation biasing factors described above.31 amount to c. 5 m, whilst the overall diference in Rhine-Meuse delta. This landscape can be elevation decreases in northern direction, from described as a wide river valley composed of c. 50 m to 12 m Amsterdam OD. Sand and gravel typical foodplain and stream gully features, deposits are common, and these are frequently aeolian river dunes and peat. Surface elevation covered with loam. Extensive aggregate decreases from east to west, from c. 14 m to extraction of gravel and coarse sand, as well as −1.5 m Amsterdam OD, which results in foodplain adjustment for water management increasingly extended peat plains in the western purposes, has instigated most archaeological part of the delta. Typically, landforms at the surveying and excavation. This is particularly the present-day surface are of Holocene age, with case north of Maastricht. Because aggregate the exception of river dunes, which mostly extraction and water management works are formed during the Late Glacial (Younger Dryas), controlled by the national authorities, but continued to be formed during the Early involvement of the national heritage agency has Holocene.32 However, this landscape hides a had major infuence on development-led complex lithological and geomorphological feldwork. However, early prehistoric remains stratigraphy that is related to its long geological are relatively scarce, because the Meuse Valley history as a river valley in the Pleistocene and landscape is principally of younger Holocene the Holocene. The complexity of the sub-surface age. Landscape zones dating to the Pleistocene build-up – clay, loam, sand, gravel, peat – is consist of local outcrops, but are mostly largely due to the alternation of erosion and represented in the bordering Meuse terrace sedimentation at multiple spatial scales.33 In landscape. Early Holocene deposits occur general terms, the number of surveys and occasionally farther north. Archaeological excavations, which potentially could reveal early surveys and excavations that involve early prehistoric archaeology, increases to the east. prehistoric remains are confned to such relic This is partly due to intense economic activity features. and related construction work, and partly due to Campine sand area. Bordering the southern the occurrence of landforms dating to the Late limits of the Dutch coastal and river regions, this Pleistocene and frst half of the Holocene close landscape, a tectonic high, is characterised by a to the present-day surface. In the western half of gently undulating relief intersected by many the Rhine-Meuse delta, hardly any excavations brook valleys. The sandy nature of the have been conducted. Archaeological remains subsurface encountered today is to the result of that are early prehistoric in age can only be large-scale peat reclamation in medieval times, 34 found at considerable depths, making in particular in the western half of the area. This 30 Rensink et al. 2015. excavation exceedingly expensive. Hence, has caused major soil degradation and sand drif 31 At the time of writing, only these four archaeological landscapes have been excavation can only be considered in cases at various spatial scales. Elsewhere, Pleistocene deﬁned in greater detail (Rensink et al. where highly elevated palaeolandforms, such as coversand deposits are encountered at the 2015). 32 Vos & Cohen 2015. river dunes, are encountered. Typically, deeply surface in between the brook valleys, which 33 Busschers 2008; Hijma 2009. buried deposits that are of potential interest but themselves have an infll of peaty fuvial 34 Holocene deposits reach a thickness of c. 15 m near Roterdam, in contrast to that are expected to bear few archaeological deposits; Early Pleistocene deposits are present c. 1 m at the Dutch–German border. remains are ofen excluded from the heritage relatively close to the surface. Except for the 35 See, for instance, the notes to the archaeological base map of the 35 management process altogether. In addition, sand drifs, diferences in elevation from the municipality of Gouda (Groenendijk deeply buried sites are not considered valley foor to the sandy plateaus are between 2 2011). An exception is made with regard to deeply buried (>2 m) Holocene threatened because construction or disturbance and 5 m. The overall diference in elevation is meander belt deposits, for which a high will not reach the archaeological levels. from c. 40 m to c. 0 m Amsterdam OD in archaeological potential was established at Gouda-Westergouwe; Meuse valley. As the upstream extension of northerly direction. Despite the fact that archaeological survey is required where the Rhine-Meuse delta, the Meuse valley Pleistocene terrestrial deposits can be found at such deposits occur within development plans exceeding 100 m2, landscape describes a narrow zone bordered by or near the surface practically everywhere, even at depths greater than 2 m. 38 —

surveys are almost exclusively restricted to the middle and eastern parts of the Campine sand Table 2.4 Number of excavated sites and area.36 None has led to excavations involving periods represented. A single site can signifcant early prehistoric archaeology – not represent several periods. even in the case of major archaeological feldwork in connection with the construction of Period Excavation Total sites the high-speed railway (Hoge Snelheidslijn All sites 85 139 Zuid).37 Concentrations of Palaeolithic, Mesolithic Palaeolithic 10 17 and Neolithic fint artefacts collected at the surface do, however, atest to the early Late Palaeolithic 10 17 prehistoric use of the entire area, but these also Mesolithic 56 90 point to major disturbance of sites. In many Early Mesolithic 17 25 cases only the deeper features (postholes, pits Middle Mesolithic 25 31 and ditches) seem to have survived the anthropogenic landscape-forming processes Late Mesolithic 24 35 that have occurred since the end of the Middle Neolithic 58 78 Ages; most of these features are of late Early Neolithic 24 34 prehistoric or (early) historic age. Middle Neolithic 37 48 Northern loess area. This landscape, which covers only a small portion of the extreme south- eastern part of the country, is characterised by a strongly accentuated relief. Pleistocene loess somewhat fuzzy in connection with cultures, deposits (partly colluvial) and underlying alluvial groups and traditions. Figure 2.11 expresses how Meuse deposits are incised by brook valleys and defned socio-cultural entities overlap or succeed dry valleys. Surface elevation ranges from 40 m to one another in time. Diferences in the number of 180 m Amsterdam OD. Aggregate extraction in sites from one period to the next do not quarries has led to deep exposure of Pleistocene necessarily refect any ‘prehistoric reality’, such as (and older) deposits, but even construction works shifs in population density or occupation history. necessitating only superfcial disturbance can Second, most of the sites in this analysis comprise easily cut into Pleistocene deposits. Early material from more than one archaeological prehistoric remains are widely known from this period, based on either radiocarbon dates or area, dating to, amongst others, the Middle typo- and technological information. Third, the Palaeolithic, the Late Upper Palaeolithic, the Early possibilities of anchoring fnds assemblages Neolithic (LBK) and the Middle Neolithic chronologically are (partly) dependant on fnds (Michelsberg). Archaeological surveys and quantity and, related to that, the chances of excavations concentrate in the economically diagnostic elements being present, as well as the dynamic western part of the area, along the availability of associated and reliable dates. Meuse valley. Compared with other parts of the Assessing the potential information value of country, the number of excavations involving such chronological ‘divides’ is, however, not remains of early prehistoric occupation is high, easy, as the availability of chronological due to the particular geological situation. anchoring points is highly variable within the set of selected sites. Chronological atribution of Temporal distribution archaeological remains is frequently based on The temporal distribution of the sites selected for typological or technological arguments instead our synthetic analysis is summarised in table 2.4. of radiocarbon dates. As a result, the Prior to presenting any further discussion, we chronological framework remains extremely should note various problems that infuence the coarse. In exceptional cases, large numbers of assessment of the information potential. First, radiocarbon dates and stratigraphic arguments archaeological time is categorised in periods of permit more detailed analysis. Below we will variable lengths, with somewhat arbitrary therefore restrict ourselves to a broad outline of 36 It is not clear why this is the case. Possibly it coincides with diﬀerences in boundaries that sometimes overlap. Figure 2.10 the chronological coverage of the sites, and we the number of spatial developments. provides the chronology that is generally used in will discuss potential gaps and biases within the 37 Ball & Van Heeringen 2016; Kranendonk et al. 2006. the Netherlands. Clearly, the chronology becomes dataset. 39 —

Fig 2.10 Schematic chronology of the prehistory of the Netherlands (from Van den Broeke, Fokkens & Van Gijn 2005). 40 —

Fig 2.11 Chrono-geographical diagram showing the mosaic of socio-cultural groups/traditions in NW Europe during the Neolithic; colours reﬂect ‘stages’ of neolithisation (from Louwe Kooijmans 2008). 41 —

Palaeolithic surfaces are deeply buried. Occasional fnds of The selected reports do not involve any Middle Hamburgian and Federmesser Gruppen artefacts Palaeolithic or Upper Palaeolithic archaeology. along the west coast provide evidence for the Although reference is made to the Middle presence of the people associated with these Palaeolithic for Wezuperbrug-Oranjekanaal, artefacts.47 Because there are no apparent where remains of woolly mammoth were reasons for diferent paterns of land use, we found,38 this site was excluded due to the assume continuity of geographical distribution absence of evidence for human (i.e. paterns with the eastern half of the Neanderthal) intervention. A handful of Middle Netherlands. Hence, the only marked diference Palaeolithic artefacts have been reported from concerns the Hamburgian, which is, as yet, only the site of Nederweert-Rosveld.39 As a single dot represented north of the River Rhine. on the map, representing only a few artefacts The majority of Late Palaeolithic sites originating from a heavily disturbed context,40 discovered in the context of development-led this site has also been excluded from our feldwork occur in ‘archaeological landscapes’ selection. Finally, charcoal features that are which were already known to contain such interpreted as hearths have been reported from remains. Typically, these are the ‘boulder clay Bergeijk-Keersop, and a radiocarbon date of area’, the ‘northern sand area’, and the ‘Ruhr c. 29,430 BP (GrA-42737) would place them in valley graben’. More striking are a number of the Early Upper Palaeolithic.41 However, the sites in the ‘lower Meuse terrace’ and ‘Meuse anthropogenic nature of these features cannot valley’ landscapes. Due to alluvial dynamics, be confrmed, and fnds, such as fint artefacts land surfaces of Late Palaeolithic age have been are lacking (there is one unworked chunk of subject to erosion and have a patchy occurrence. poor-quality fint). Because the archaeological The small number of sites discovered in this nature of this site is highly questionable, it has landscape context can potentially inform us been excluded.42 about facets of land use that are not represented The earliest sites represented in our in the other areas. The same may be true for a selection of reports date to the Late Palaeolithic Hamburgian site discovered at the border of the and are located in the eastern half of the country ‘IJssel valley’ and ‘northern sand area’. (fg. 2.12). Finds from these sites have been atributed to one or more cultural groups: Mesolithic Hamburgian, Federmesser Gruppen and Sites dated to the Mesolithic are well Ahrensburgian. Seven sites have an unclear represented in the reports (fg. 2.13). The great cultural signature but are likely to be of Late majority is situated in the eastern half of the Palaeolithic age in view of technological and Netherlands. In the western half, a small ‘cluster’ stratigraphic considerations. There is no of sites is reported in the surroundings of 38 Aalbersberg et al. 2013. mention of fnds atributable to the Roterdam;48 these are amongst the most 39 Hiddink 2005. 40 Research into site taphonomy was Magdalenian, but this may be a coincidence, as western Mesolithic sites that have been conducted by the national heritage the Magdalenian is sparsely represented in the investigated in the Netherlands thus far. agency (Deeben et al. 2010). 41 Appendix II: PUBID 65. Dutch archaeological record in general.43 The Although related to river dune contexts, which 42 It should be noted that a ﬁnd such as distribution of these Late Palaeolithic sites are landscape features known to be of this would have been of major importance had the anthropogenic conforms to our current knowledge of the archaeological signifcance, these sites can nature of the features been established; geographical occurrence of these cultural potentially also inform us about unknown facets it would have provided the ﬁrst evidence for the presence of groups. Hamburgian sites are not found south of of land use and habitation. The ofshore location anatomically modern humans in the the Rhine,44 whilst Federmesser Gruppen sites at Roterdam-Maasvlakte 2, at a depth of −20 m north-west European sandy plain during the Early Upper Palaeolithic, afer the are encountered throughout the Netherlands.45 Amsterdam OD is exceptional in this respect. We disappearance of Neanderthals (cf. Ahrensburgian sites have previously been can identify no explicit geographical diferences Roebroeks 2014). 43 Rensink 1993, 2010. encountered in the northern and southern parts in site occurrence between the early, middle and 44 Rensink & Niekus 2016. of the eastern half of the country.46 In a general late phases of the Mesolithic that might 45 Deeben & Niekus 2016a. 46 Deeben & Niekus 2016b. sense, the presence of Late Palaeolithic hunter- corroborate or reject current models of 47 Stapert 1981; Veerman 1971. gatherer groups in the western half of the increased use of river valleys towards the Late 48 The cluster of sites is a consequence of the fact that the archaeological unit of Netherlands is, as yet, poorly known due to the Mesolithic.49 Several seem to represent the municipality of Roterdam is aware geological history of that part of the country. If multiphase sites, where locations were used of the potential of the region for the presence of Stone Age sites. they are preserved, Late Pleistocene land throughout the Mesolithic. 49 Crombé et al. 2011; Verhart 2008. 42 —

Fig 2.12 Distribution of the selected sites/reports mentioning Upper/Late Palaeolithic ﬁnds (background: ‘archaeological landscapes’ as deﬁned by the Cultural Heritage Agency of the Netherlands).

The sites occur in a wider range of landscapes. Several occurrences – Well-Aijen in ‘archaeological landscapes’ than do the Late particular – in the ‘Meuse valley’ landscape Palaeolithic sites discussed above. The setings are still under investigation, and have, distribution of Mesolithic sites resulting from as yet, only been reported upon in the context of development-led feldwork conforms to the surveys. This preliminary information will general picture, but again with the exception of probably shed only limited new light on our the ‘lower Meuse terrace’ and ‘Meuse valley’ knowledge of the Mesolithic. 43 —

Fig 2.13 Distribution of the selected sites/reports mentioning Mesolithic ﬁnds (background: ‘archaeological landscapes’ as deﬁned by the Cultural Heritage Agency of the Netherlands).

Neolithic distinct concentration of Early and Middle As it is the case for the Mesolithic, a considerable Neolithic sites in the southern loess area of number of reports involve Early or Middle Limburg. A breakdown by ‘cultural assignment’ Neolithic remains (fg. 2.14). Typically, these are suggests that the sites are somewhat more situated more ofen in the wetland areas of the bound to a single phase of cultural context. This central and western Netherlands, compared with is partly due to the well-identifed LBK sites on the sites dating to earlier periods. Also, there is a the loess plateaus, which are characterised by 44 —

Fig 2.14 Distribution of the selected sites/reports mentioning Early-Middle Neolithic ﬁnds (background: ‘archaeological landscapes’ as deﬁned by the Cultural Heritage Agency of the Netherlands).

clear house plans and pits. Furthermore, this the site may contain material belonging to other could well be a self-fulflling efect of cultural cultural entities that is simply not diagnostic. In atribution of distinctive fnds and not a this respect it is important to bear in mind that refection of prehistoric reality. For instance, the the archaeological distinction of cultural groups positive identifcation of a single Michelsberg or traditions in the Neolithic is based on potery. potery fragment will lead to the ‘cultural In the absence of potery fragments, cultural labelling’ of a site as Michelsberg, even though atribution of fnds assemblages is ofen 45 —

difcult. For instance, it is impossible to distinguish between the Late Mesolithic and the 2.4 Thematic data mining Early Neolithic Swiferbant Culture on the basis of fint artefacts.50 To an extent, such problems may also hold With the database fnished and the relationships for the relationship between cultural assignment amongst questions, aspects and publications set, and geographical occurrence (fg. 2.15); the new we started our thematic data mining. This proved data reveals litle to no deviation from the less straightforward than we had hoped, as the established ‘cultural territories’. LBK is reported manner in which data is presented is mostly from the south-east (province of Limburg); highly fragmented or incoherent (see examples Swiferbant shows up in the Rhine-Meuse delta below). This became particularly apparent when and the northern Netherlands; Michelsberg and we atempted to create automated thematic Stein are identifed in the south-east (province overviews of the available data.53 Whilst many of Limburg) and along the Meuse valley; and the reports present the data in an exemplary manner, Hazendonk and Vlaardingen are confned to the others lack clear data summaries or were found Rhine-Meuse delta.51 However, distinguishing to be incomplete. Particular problems were amongst these cultural labels is far from easy, encountered with regard to quantitative data. even for specialists in this domain. What would • Mention of total numbers of (categories of) fint be called Stein in the south-east would be called or potery in the text, but no summary table; Vlaardingen in the west. Indeed, much is due to • Use of vague quantitative terms, such as diferences in regional research history and to much, many, litle, instead of exact numbers; conceptual diferences in how to interpret • Provision of percentages without clear variability in material culture at various spatial indication of sample or population size (totals scales.52 On the other hand, the application of and subtotals not given). these archaeological constructs does help to This last problem – which we also saw in identify noteworthy paterns. For instance, some graphical quantifcations, such as pie and bar sites in the Rhine-Meuse Delta are labelled TRB charts – was encountered most ofen in reports (the German abbreviation of Funnelbeaker where early prehistory was so-called by-catch. Culture) in the Archis database, whilst most TRB A more rigorous approach to quantitative data sites are situated in the north and east of the presentation, which follows basic conventions, country, where one would expect to fnd them. is the least we can ask for. Clearly, this prompts questions about diagnostic Other types of problems concern issues of viability. classifcation and subsequent presentation of data. Examples concern the raw material Taken together, the selected reports provide characterisation of fint and the cultural unbalanced coverage in terms of the atribution of potery. The characterisation of representation of geographical regions, fint is highly variable and appears clearly related archaeological landscapes, and chronological and to diferences in the level of expertise amongst cultural atributions. This is due to diferences in researchers. The breakdown of origin in terms of geological history, the nature of modern-day land extraction context (primary, secondary); use and spatial development, the nature of geographical origin (northern/southern, development-led feldwork, the nature of political Scandinavian/Limburgian); and type assignment decision making, as well as archaeological (e.g. Rijckholt, Valkenburg, Banholt) varies conceptual constraints. Any synthetic analysis of amongst reports, whereas diferent terms may data derived from these reports has to take these be used to characterise the same source (e.g. biasing factors into account, most particularly Limburgian/southern/Rijckholt fint). With where it comes to interpretations with a regard to ceramics, data appear to be presented geographical component or the reconstruction of according to variable groupings. Data from a long-term occupation histories. Notwithstanding single site can be presented according to cultural 50 Cf. Peeters 2007. these restrictions, the analysis allows for a critical group assignment for one category of fnds and 51 For a discussion about the status of these cultural groups, see Beckerman evaluation of existing insights and ideas, as well for the assemblage as a whole for another and Raemaekers (2009). as a (re)formulation of hypotheses for future category of fnds. For example, the potery from 52 Cf. Ten Anscher 2012. 53 This forced a manual data-mining research. Hazerswoude-Rijndijk was analysed by cultural approach. 46 —

Fig 2.15 Distribution of the selected sites/reports referring to speciﬁc socio-cultural groups/traditions of the Neolithic (background: ‘archaeological landscapes’ as deﬁned by the Cultural Heritage Agency of the Netherlands). 47 —

Fig 2.15 Continued. 48 —

Fig 2.15 Continued. 49 —

assemblage, whereas the fint was considered at • Monuments, reports and independent the level of the entire site. For both fint and observations are to be entered into the ceramics, it is difcult to be certain about which national information system, which should be data from diferent reports are comparable. accessible to anyone (emphasis in the Another problem concerns the presentation original); however, the relevant agencies of radiocarbon dates. It is not unusual to fnd retain the rights of authorship and the only calibrated 14C-age estimates, without database rights (Article 5.12). mention of the measured radiocarbon age or the In the context of our synthetic study of research calibration sofware version used. Clearly, this is results, we frequently wanted to turn to the undesirable, as diferent calibration sofware original data. The unavailability of this data (in versions may use diferent calibration curves, digital or printed form) frustrated any atempts hence producing diferent dates; uncalibrated to take further steps and increase the potential dates should always be presented according to to build on research results from accepted conventions.54 development-led projects. Publication of A fnal problem to be addressed here primary data and specialist databases would be concerns the availability of primary data. For 72 more in line with Article 8 of the Valleta treaty.55 out of the 139 projects, additional data, such as In the present situation, data is not as easily fnds databases or additional maps could be accessible for interested parties and researchers found online (Archis, DANS-EASY). For the as it should be, and this situation poses a risk to remaining 61 projects, only the writen report the archaeological resource. was available online. This situation is in confict Notwithstanding the problems and with Dutch legislation, which includes the restrictions mentioned above, we were able to 54 Waterbolk 1983. following requirements for excavating agencies extract useful data from the set of selected 55 Article 8 of the Valleta Convention states that ‘each party (country) (translation ours): reports. This required screening each individual undertakes to (i) facilitate the national • During excavation, actions and archaeological report. It turned out that the majority of reports and international exchange of elements of the archaeological heritage for fnds are documented; the fnds are conserved present data of limited use. The following professional scientiﬁc purposes, whilst and a report is made that describes the results chapters will demonstrate that the contribution taking appropriate steps to ensure that such circulation in no way prejudices of the performed actions (Article 5.4). that can be made to the various research the cultural and scientiﬁc value of those • The fnds, the report and additional questions under the umbrella of the themes and elements; and to (2) promote the pooling of information on documentation are to be presented to the dimensions as defned in chapter 1 is highly archaeological research and excavations government within two years afer the variable and unbalanced. in progress and to contribute to the organisation of international research excavation is completed (Article 5.6). programmes.’

51 3 People and material —

transmission as a mechanism of hereditary 3.1 Introduction socio-technological traditions.66 The broad range of approaches to material culture that developed over the years have As we briefy indicated in section 1.2.1, found their way into Dutch archaeology. relationships between people and material are Generally speaking, traditional typo- complex. The archaeological study of material technological culture-historical and processual culture has a long tradition and, indeed, goes perspectives are dominant. The study of back to the very early days of the discipline.56 artefacts and structural features predominantly From a culture-historical perspective, material serves as a means to interpret sites from a remains have been studied as a means to defne functional perspective – as part of a ‘setlement and distinguish between ‘cultures’ and system’ –and to contextualise remains within ‘traditions’ in the past, as well as to build established chronological and cultural evolutionary and chronological frameworks.57 frameworks. The typo-chronological analysis of New questions and interpretations of material material remains represents, without doubt, the culture arose with the paradigm shif in the overwhelming majority of studies, especially in 1960s, when archaeologists became critical of the context of development-led research. As the culture-historical approach.58 To these standard procedure, material remains are processual archaeologists, variability and change classifed according to commonly used in material culture should, frst of all, be typologies. Particular aspects, such as the use of understood in terms of functionality and raw materials, and secondary atributes, such as adaptation to diferent and changing completeness and burning, are taken in environmental conditions. In the 1970s, new consideration. Technology is mostly approached approaches, such as microscopic use-wear from a techno-typological perspective, where analysis and refting, opened new windows, objects are classifed into technologically leading to re-interpretations of an otherwise defned categories based on certain atributes. inert source of information.59 Soon afer, and Use-wear analysis of lithics has become triggered by ethnographical studies,60 ferce increasingly popular, also in development-led debates about how to distinguish between archaeology.67 In connection with use-wear ‘function’ and ‘style’ made clear that the reality analysis of stone tools, studies into the meaning of what material culture actually involves is far of material culture have also become an more complex.61 This discussion was paralleled important line of research,68 although maybe not by yet other perspectives on material culture. as much in the context of development-led Structuralist approaches – mostly a French projects. The same holds for the study of ‘school’ infuenced by the work of Leroi- ceramics.69 Gourhan62 – developed anthropological The study of material culture as it is perspectives on technology, where the concept conducted in the Netherlands, and to a variable of the chaîne opératoire (operational sequence) extent echoed in development-led research, provides a connection between technological covers the following three main themes: action, gesture and mind.63 Post-processual • Craf activities; 56 Bordes 1968; Klejn 1982. 57 See, for example, Bordes 1968. archaeologists put more explicit emphasis on • Socio-cultural meaning; 58 Binford & Binford 1968; Bordes & meaning and perception of material culture • Frames of reference. De Sonneville-Bordes 1970. 59 Van Noten & Cahen 1978. (‘materiality’), and they ascribe a central role to The frst theme puts the object central; studies 60 Wiessner 1983. the human mind.64 Building on the post- focus on the reconstruction of modes of 61 Sacket 1982; Wiessner 1983. 62 Schlanger 1994. processual perspective on how material culture production and the use and discard of items. 63 Lemonnier 1992; Pétrequin & Pétrequin should be understood, phenomenological The second theme is more concerned with the 1990. 64 Hodder 1986. approaches seek to overcome problems of contextualisation of material culture; studies 65 Ingold 2000; see Conneller 2011, p. 30. dualism (mind<>mater; material<>form), and focus on the socio-cultural role of objects and on 66 Jordan 2015. 67 This is particularly so for projects with emphasise the act of production arising from the understanding culture change. The third theme involvement of the Faculty of interaction of person and material.65 Yet another aims at the establishment of frameworks for the Archaeology at Leiden University, which hosts the Laboratory for Material perspective on material culture, and particularly characterisation of the archaeological record; Culture Studies (now Foundation LAB). on the dynamics that underlie variability and studies focus on the classifcation of cultural 68 Van Gijn 2010; Wentink 2008. 69 Raemaekers, Kubiak-Martens & (dis)continuities, emphasises the role of cultural remains and the development of chronologies. Oudemans 2013. 52 —

Below, we will evaluate what contribution comparable fracturing characteristics, such as development-led projects have, or could have, phtanite and fne-grained quartzite. Artefacts made to these lines of research. made of other types of rock (sedimentary, metamorphous and igneous) have long received limited atention, and mostly it is only tools 3.2 Craf activities made of these rocks that have been analysed. However, it seems that this situation is gradually changing and that lithic artefacts made out of Traditionally, objects are studied within rock other than fint are increasingly atracting categories that are defned on the basis of the the atention of researchers. This is particularly raw materials used. In the context of early the case with regard to use-wear analysis. prehistory, lithics feature as the dominant In general terms, the analysis of lithics is category in the NOaA 1.0. Within this main approached from a rather classifcatory and category, a systematic distinction is made descriptive perspective. In reference to the between fint and other rock.70 Other categories research question mentioned in the NOaA 1.0, are bone/horn/antler, wood, potery, and artefacts are assigned to a pre-defned typo- miscellaneous materials, such as amber, ochre morphological or technological category. Further and wood tar. As appendix I shows, specifc characterisation of artefacts on the basis of research questions as listed in the NOaA 1.0 are particular atributes – for instance, length, basic and can be summed up in more general width, thickness and completeness – is at the terms. Most of these questions are included in category level. Typically, this results in overview project outlines (the PvE) in one form or another. tables and graphs, which summarise the Ofen, questions are specifed according to the typological/technological composition of the raw material categories that are expected to be assemblage or atributes of artefact categories. (potentially) preserved on the site that is to be The results of use-wear analysis are mostly investigated. However, the way in which these presented separately. In the end, this leads to a research questions are approached in practice is rather static picture of the assemblages studied, variable and less straight forward than one might instead of a beter understanding of the imagine. This is partly because certain lines of behavioural dynamics that underlay the enquiry – the chronological resolution of certain formation of these assemblages. It appears that material categories, for instance – necessitate the analytical concept of the chaîne opératoire, contextualization beyond the site level and partly which as an integrated approach permits the because conceptual issues of analysis – for development of insight into just these example, the chaîne opératoire, as we will see dynamics,71 has not been picked up by lithic below – are not necessarily shared or understood specialists and consequently does not make it in similar ways by researchers. Although both into the project outlines. these issues are probably inherent to Below we take a closer look at what archaeological research, these are issues to keep contribution development-led archaeology has in mind in the context of the study we present made to NOaA 1.0 research questions concerning here. In the following sections, we will focus on a the production, use and discard of stone tools number of topics that have received considerable (fint and other types of rock). We will do so with atention in development-led projects, and we another question in mind: What insight can will evaluate how this atention has contributed (potentially) be gained into the behavioural to our understanding of craf activities in early dynamics that resulted in technological prehistory. variability in craf activities?

Tool manufacture and raw materials 3.2.1 Production, use and discard of With some exceptions where Middle Palaeolithic lithic tools fnds are involved, development-led projects have produced assemblages dating to the Late 70 Other rock is what is referred to as Palaeolithic, Mesolithic and/or Neolithic. Other “natuursteen” in the Dutch text. The analysis of stone artefacts is strongly than the more sporadic production of core tools 71 Pélegrin et al. 1988; Schlanger 1994, 2005. focused on fint and other rocks with (e.g. axe blades), two conceptual modes of 53 —

production were applied for the crafing of most to the selection of raw materials and tools during these time periods. One involved manufacturing (methods and techniques) of the production of tool elements out of blades, blanks and tools. the other, the production of tool elements out of fakes. These are very diferent technologies, Late Palaeolithic. Late Palaeolithic (Hamburgian, which demand diferent knowledge and practical Creswellian, Federmesser Gruppen and skills in their execution. To maintain such Ahrensburgian) assemblages rarely feature in knowledge and skills, information had to be the reports in our database.77 Artefacts that are shared amongst group members, whilst skills considered to be of this age are atributed to a had to be acquired though practice.72 Of course, particular cultural tradition on the basis of in reality things were probably not that black typological arguments, or, more ofen, on the and white. During blade production, fakes are basis of technological characteristics, the also detached, and were used as tools. But the presence of a particular type of raw material, or deliberate choice of producing blades or fakes the geological context. as a basis for tool manufacture involved The few fnds atributable to the diferent conceptualisations. Whereas blade Hamburgian come from the sites of Knooppunt technologies were dominant during the Late Hatemerbroek and Epse-Olthof,78 but only the Palaeolithic through Early Neolithic, the relative later produced enough material to draw 72 Lemonnier 1992. importance of fake technologies seems to inferences about stone tool manufacture.79 In 73 Van Gijn 2010. increase during the Middle Neolithic, when line with the current state of knowledge, we see 74 De Grooth 1987. 75 Noens 2011, 118. blades become more scarce.73 This shif is the systematic production of slender, slightly 76 Tixier 1984; Tixier et al. 1980. probably related to changing socio-cultural curved blades out of fne-grained ‘northern’ It should be noted that, even when a technological definition is applied, contexts, where we start to see a distinction fint – amongst which red Heligoland fint (see production schemes are not always between ‘domestic technologies’ and ‘specialist section 4.2.2) – through the application of direct aimed at producing highly regular blanks or blades. Mesolithic Coincy- technologies’.74 However, as yet, we have limited sof percussion (fg. 3.1). Blades were detached style production schemes are relatively knowledge about how various technologies and from two opposite striking platforms and a irregular, but there is still a tendency to systematically produce elongated blanks the use of tools were embedded within early single production surface. One core was (Walczak 1998). prehistoric socio-cultural traditions, and how originally prepared on a large fake, but the core 77 Late Upper Palaeolithic (Magdalenian) finds which could be expected to be and when change occurred. was dramatically reduced in size due to a present in the south-eastern part of the The distinction between blade-based and plunging blade that came of in the very frst Netherlands are absent. 78 Knooppunt Hatemerbroek covers a fake-based technologies is systematically made stage of core reduction; the remaining core was large area involving two projects in development-led studies. In this respect, it is further used to produce small blades and was (appendix II: PUBID 259, 260, 749), which made use of the same project important to take a look at what distinguishes a subsequently abandoned. A few scrapers and a outline. Epse-Olthof involves a series of blade from a fake. As mentioned in connection fragment of an impact-damaged Havelte (?) projects conducted over more than ten years by the archaeological unit of the with the study of the assemblages from point suggest maintenance activity and hide municipality of Deventer; the results on Leeuwarden-Hempens, blades have a working.80 Core reduction debris, some of which the early prehistoric occupation have been published in a single volume length:width ratio of at least 2:1, although it is could be refted, provides evidence for on-the- (appendix II: PUBID 733). acknowledged that many other atributional spot tool manufacture out of high-quality fint 79 It should be noted that the first indications for the presence of criteria (e.g. parallel edges, regularity of dorsal that was brought in from elsewhere (there is Hamburgian artefacts were reported in scars) can be added.75 Indeed, the execution of a none available in the vicinity of the site). the context of development-led research in which the Groningen pre-conceived scheme of production to obtain Although the Hamburgian assemblage from Institute of Archaeology was involved relatively long and regular blanks is considered Epse-Olthof is fragmentary, it is likely the result (Peeters et al. 2015). At a later stage, fieldwork was continued in the context to be fundamental to this technology by of multiple short-lived visits by small groups of of the annual fieldschool for specialists in fint knapping.76 Yet, the simple, hunter-gatherers. The resulting picture fts the archaeology students of Saxion University of Applied Sciences in arbitrary metrical criterion is picked by most idea of a technological system that combines Deventer, which resulted in the recovery researchers for practical purposes, neglecting on-the-spot core reduction and tool of more, undisputably Hamburgian, material, again through the the problems this creates for technological manufacture/maintenance, the abandonment of involvement of the Groningen Institute analysis that aims to go beyond descriptive blades and utilised tools that were brought in of Archaeology (Hoebe 2015; Peeters & Rensink in prep.). classifcation. But let us see what development- from elsewhere, and the transport of cores 80 Since the tang is almost entirely led studies contribute to our general between encampments.81 missing, the atribution of the point to the Havelte type is uncertain. understanding of lithic technology with regard 81 Johansen & Stapert 2004. 54 —

Fig 3.1 Various types of flint from the Hamburgian assemblage of Epse-Olthof. Most of the flint is of the Senonian type (translucid without fossil inclusions); the red blade (with two refited blade fragments) is made of red Heligoland flint (photo by P. Hoebe). Scale bar in centimetres.

Assemblages atributed to the Federmesser Federmesser Gruppen,83 but the spatio-temporal Gruppen are equally rare. In several cases (e.g. integrity of this material is mostly questionable Hasselo, Almen-Het Asseler), we seem to be due to the scatered distribution patern and dealing with ‘by-catch’ of isolated objects, such unclear stratigraphical context. Only one of as backed points (Tjonger points, Federmesser these assemblages represents a clear points), which are easily identifable.82 Several concentration, said to have been found in a assemblages collected at Knooppunt shallow ‘pit’. It contained a group of 79 artefacts 82 Appendix II: PUBID 645. 83 Appendix II: PUBID 259. Hatemerbroek can possibly be atributed to the originating from the same nodule, some of 55 —

Fig 3.2 Blade production waste from the Late Palaeolithic concentration at Knooppunt Hatemerbroek (from appendix II: PUBID 749). The length of the core (top lef) is 66 mm.

which could be refted: fakes, blades and one Remains of short-lived encampments blade core with two opposite striking platforms atributable to the Ahrensburgian have been (fg. 3.2). Initially, this material was atributed to documented at the sites of Wijchen- the Federmesser Gruppen.84 The authors state Kraanvogelstraat and Alverna (also in Wijchen). that an atribution to the Hamburgian or even Although the lack of points at both of these sites Ahrensburgian is possible in reference to the makes it difcult to confrm an Ahrensburgian absence of hard hammer blade production, afliation, such an afliation is probable in view which is typical of the Federmesser Gruppen.85 of the technological characteristics of the blades, Since the description in the report is the morphology of the burins and scrapers, as insufciently detailed,86 and since some well as the use of high-quality fint. For Alverna, drawings and a photograph published in the the presence of light grey Belgian fint was earlier evaluation report lack the detail that reported;88 it concerns several blades and a might permit the reader to establish the cultural scraper, but no knapping debris. This type of fint afliation,87 the material should be looked at is known from well-documented Ahrensburgian again. Whatever the cultural atribution and sites in the southern Netherlands. Production precise dating, we seem to be dealing with a debris and refted artefacts indicate on-the- Late Palaeolithic, short-lived fint-knapping spot manufacture of blades at the Alverna site workshop, where only the exhausted core, (fg. 3.3). Several blades and fakes could be 84 Hamburg & Knippenberg 2005, 58. preparation fakes and rejected blades were lef refted to the core. The absence of decortication 85 Verbaas et al. 2011, p. 344. 86 Verbaas et al. 2011. behind. Given the absence of raw material fakes suggests that the initial preparation of the 87 Hamburg & Knippenberg 2005, 55-58. sources in the direct environs of the site, it is core, from a cobble, took place elsewhere. Blade 88 Appendix II: PUBID 742. The assemblage was studied by J. Deeben (Rijksdienst clear that the assemblage must be seen in the production was executed from two opposite voor het Cultureel Erfgoed). Deeben (†) context of a technological system that involved striking platforms and a single production had a profound knowledge of the Late Palaeolithic and Mesolithic of north- transportation of larger nodules of fint. surface; the morphology of the core was west Europe. 56 —

Fig 3.3 Core with reﬁted blades and ﬂakes from the Ahrensburgian assemblage at Alverna (from appendix II: PUBID 742). Scale bar in centimetres. 57 —

Fig 3.4 Ochre-stained coversand at A2-Aalsterhut with some ﬂint artefacts in situ (from appendix II: PUBID 216).

controlled by means of detachment of 2-3 m wide strip could be investigated, it is not correction fakes. The few formal tools (n=8, clear how representative the lithic assemblage is c. 10% of the assemblage) – mostly scrapers and of the total distribution. Both blades and fakes burins – indicate a narrow activity spectrum. The were used as blanks for the manufacture of assemblage from Wijchen-Kraanvogelstraat tools. Bigger blades atest to a bidirectional core contains a higher percentage of tools (n=39, reduction scheme. Most of the production waste c. 20% of the assemblage), but the range of tool and tools are of Belgian fint (amongst which types is comparable.89 The presence of several Haspengouw fint) and fint of south-eastern tested nodules, cores and various core Dutch origin (Rijckholt, Rullen) collected from rejuvenation pieces, and of many artefacts with river deposits. However, some artefacts that cortex, shows that here, in contrast to the bear relatively fresh cortex may have been Alverna assemblage, all steps of the production collected at outcrops in the southern part of the sequence are represented. As in the previously province of Limburg.91 The overall assemblage discussed Hamburgian and Federmesser composition and the presence of some quartzite contexts, these Ahrensburgian assemblages ft a artefacts – including one showing abstract linear technological system that involved decoration – seems to point to a multipurpose transportation of nodules of fint, as well as of encampment. As was the case at other prepared cores, blanks and tools. Ahrensburgian sites in the vicinity, red ochre was A third Ahrensburgian assemblage comes part of the behavioural context of the activities from the site of A2-Aalsterhut.90 The site is part conducted at the site (fg. 3.4). of a legally protected complex of Late Although restricted in number, the Late Palaeolithic (Federmesser Gruppen, Palaeolithic assemblages described above Ahrensburgian) and Mesolithic sites that has provide evidence for technological systems been (partly) excavated since the 1950s. More within which (long-distance) transportation of than 1500 artefacts were collected, amongst raw materials and objects in various stages of which 39 tools, such as points, scrapers, burins, exploitation and use was common. Diferences and retouched/utilised fakes and blades. This is in assemblage composition point to inter-site 89 Heirbaut 2010, 136. 90 Appendix II: PUBID 216. a rather broad spectrum, but because only a variability in stone tool production. 91 Vroomans et al. 2014. 58 —

Mesolithic. Lithic tool assemblages of Mesolithic An Early Mesolithic assemblage is age are frequently encountered in the context of represented at Knooppunt Hatemerbroek. At development-led feldwork. Both tools made of this site, assemblage 6.12 comprises blade cores fint and tools made of other types of rock (e.g. with uni- and bidirectional reduction schemes. quartzite, sandstone, granite) are encountered. The majority (75%; n=15) are fake cores (fg. 3.5). The later category will be discussed in the Maintenance of cores is atested by rejuvenation section on tool typology and function. With and correction fakes, whereas blades show regard to fint and other types of rock with evidence of platform edge preparation by means comparable fracture properties, Mesolithic of retouch. Typically, about 60% of the blades technology is generally characterised by the (n=85) lack a clear bulb of percussion, c. 25% manufacture of (micro)blades and fakes as show a weakly developed percussion cone, and blanks for tools. However, the modes of blade another 25% show a lip. These characteristics production have been shown to change over indicate the use of sof hammers (stone, antler), time. Early Mesolithic blades are, to some which supports the idea of continuity from Late extent, produced in a Late Palaeolithic tradition, Palaeolithic technological traditions.95 An but they are nonetheless diferent from assemblage dated to the Early-Middle Mesolithic Ahrensburgian blades. Middle Mesolithic blade ‘boundary’ from Dronten-N23 (assemblage 5) production is strikingly diferent, and shows a shows an equal representation of fake and blade clear focus on microblades. In the northern cores, but the later are probably Netherlands, microblades were possibly – but underrepresented due to their use for the not exclusively – produced by means of the production of fakes in the fnal stage of pressure technique;92 no such indication exists reduction. Bipolar cores reduced by the hammer- for the central and southern Netherlands. The and-anvil technique, which occur in Middle and Late Mesolithic, in contrast, is characterised by Late Mesolithic assemblages from the same site, the production of relatively long, regular blades are lacking in assemblage 5. On the other hand, by means of indirect percussion (the punch assemblage 6 from Dronten-N23, also dated to technique); this mode of production is continued the Early-Middle Mesolithic ‘boundary’, does into the Early Neolithic. Although this shif has comprise bipolar cores.96 Middle Mesolithic been described in the literature for some time,93 assemblages 3, 7, 8 and 9 show a dominance we think that it is worthwhile to see how (>70%) of fake cores. The remaining cores are assemblages excavated in the context of blade cores and bipolar cores; the later are development-led projects ft this picture, and reduced by means of the hammer-and-anvil how these assemblages can inform us in greater technique. Further technological information is detail about fint tool manufacture and lacking in the report. associated use of raw materials. One problem Late Mesolithic assemblages are we have to face is the palimpsest nature if the represented at Dronten-N23 and Leeuwarden- great majority of sites and the mixing of Hempens.97 Assemblages 1 and 2 from materials dating to more than one occupation Dronten-N23 include more than 200 cores, of phase. The most extensive Early to Middle which 60% fake cores, 26% blade cores, and Mesolithic assemblages come from the sites of 14% bipolar cores. Flake cores have one or more Dronten-N23, Knooppunt Hatemerbroek, platforms and are ofen irregularly shaped. Roterdam-Yangtze Harbour, Epse-Olthof and Amongst the blade cores, unidirectional Ede-Kernhem.94 Without exception, these sites reduction schemes and single platform strategies were used on multiple occasions and, based on are dominant. Bipolar cores have been reduced radiocarbon dates, over many centuries –and by means of the hammer-and-anvil technique. hence by many generations of hunter-gatherers. No further technological information is given Nonetheless, several distinct clusters within with regard to fake and blade production, so it 92 Musch & Peeters 1993. some of the wider fnds distributions have cannot be confrmed whether indirect percussion 93 Pélegrin 2000; Tixier et al. 1980. 94 Appendix II: PUBID 259, 462, 538, 733, permited the archaeologists to isolate relatively was used. The use of indirect percussion seems, 745. discrete chronological assemblages, which have however, to be likely – based on the drawings 95 Verbaas et al. 2011, 364. 96 Niekus et al. 2012. been dated on the basis of 14C-dates of provided98 – at Leeuwarden-Hempens, where 97 Noens 2011; Niekus et al. 2012. associated charred hazelnut shell and microlith long and regular blades were found (fg. 3.6). The 98 Noens 2011, ﬁg. 68-70 (120, 122-123), ﬁg. 74 (130). typology. 14C-dates on charred hazelnut shell do, however, 59 —

Fig 3.5 Mesolithic blade cores from Knooppunt Hatemerbroek and Dronten-N23 (from appendix II: PUBID 259, 538). Scale bar in centimetres. 60 —

Fig 3.6 Mesolithic ﬂake and blade cores and blades from Leeuwarden-Hempens (from Noens 2011). Scale bar in centimetres. 61 —

Fig 3.6 Continued. 62 —

indicate activity throughout the Middle and Late people who also exploited the Roterdam area. Mesolithic, which makes it difcult to confrm a Late Mesolithic age for these blades. Several Neolithic. The Neolithic is generally considered to notched blades (Montbani blades) and numerous have been a period of technological change, trapezoid points made out of such regular blades including the appearance of new types of tools, suggest a Late Mesolithic age. The same site has new methods of tool manufacture, and methods produced regular microblades, as well as small of raw material acquisition. A considerable prismatic cores showing regular blade negatives number of assemblages have been excavated in that may indicate pressure faking. This is also the context of development-led archaeology in the case for the ‘handle cores’ and ‘bullet core’ almost all parts of the Netherlands. The mentioned in the report. Again, it is difcult to chronological and cultural context of these establish their age, but triangular points and assemblages is diverse, however, which makes it backed bladelets made out of regular possible to potentially obtain deeper insight into microblades make a Middle Mesolithic age more variability in stone tool manufacturing probable (see section 3.4.1, on the typo- technologies. Within the Early Neolithic of the chronology of the Late Palaeolithic and the Netherlands, several socio-cultural contexts are Mesolithic). distinguished: LBK, Rössen and Swiferbant. The In general terms, the raw materials used in Middle Neolithic comprises Michelsberg, Stein, the Mesolithic seem to be of local origin. This Vlaardingen, Hazendonk, and TRB. means nothing more than that there is no Several sites have yielded assemblages with concrete evidence for fint varieties collected in a considerable amount of material from LBK areas at considerable distance from the sites contexts: Sweikhuizen-Geverik, Beek-Kerkeveld, investigated. And what it does not mean is that and Rijksweg A2-Stein Heidekampweg.101 All the raw materials were collected at the sites derive from setlements characterised by typical themselves. Only in the case of Leeuwarden- LBK house plans and associated pits and ditches. Hempens can such an assumption be made From a technological perspective, the fint – to some extent – because the locally occurring material fts the LBK picture (fg. 3.7). Tools glacial till in the substrate contains nodules of mostly comprise scrapers, borers, and retouched fint. However, there is no evidence for local and notched/denticulate pieces made out of mining of these deposits, which did not reach blades and fakes. Points occur on all sites, but in the Mesolithic land surface.99 Hence, ‘local’ low numbers. As far as can be judged from the seems to refer to the exploitation of secondary published drawings and photographs – no sources of fint, the source of which is unknown, quantifed measurement data are provided – indeed impossible to know. As an alternative, blades are relatively wide and regular, and have it has been proposed to use the term ‘non- been detached from prismatic cores. The territorial’ for materials originating from distant denticulate edge of a platform rejuvenation fake sources, which implies that local sources are (tablet) from Beek-Kerkeveld indicates the located within a territory.100 But what, application of indirect percussion for the specifcally, is meant by this term? Does territory production of blades, which is common for the equate with home range or does it refer to the LBK. At Rijksweg 2-Stein Heidekampweg, entire area exploited over a yearly cycle of however, prismatic cores and tablets are rare in mobility? And how big is a territory and how is a comparison with more opportunistically particular site positioned within it? Even if one exploited fake cores. At all sites, the waste can pinpoint a unique source area, for instance, products represent the entire sequence of in the case of Wommersom quartzite (see production, from initial core preparation, section 4.2.2), it does not follow that this source through full exploitation and maintenance, to is to be considered non-territorial when discard. Several blade cores were used as artefacts of this material occur, for instance, hammer stones once they were not longer useful at sites in Roterdam. Depending on one’s for blade production. The dominant raw material 99 Noens 2011, 156. defnition of territory, Wommersom (Belgium) is fint of the Lanaye (Rijckholt) type, which could 100 Van Gijn & Houkes 2006. 101 Appendix II: PUBID 053, 528, 537. might just as well be within the territory of the have been collected at outcrops in the region, 63 —

Fig 3.7 Flint artefacts from various LBK sites. Beek-Kerkeveld: (a) blade core used as a hammer stone, (b) point, (c) blade with sickle gloss, (d) scraper; Stein-Heidekampweg: (e) production waste; Sweikhuizen-Geverik: (f ) point, (g) scraper (from appendix II: PUBID 53, 528, 537). 64 —

e.g. in the vicinity of Banholt. The Banholt of trapezoid points. Frequently, blades were used variant, which has been reported for both Beek- without any, or only limited, modifcation. Tools Kerkeveld and Rijksweg 2-Stein Heidekampweg, such as scrapers were produced out of fakes. has a characteristic brown–red iron oxide Longer blades (length exceeding 5 cm, no dorsal infltration band underneath the chalky cortex.102 cortex) were produced by application of indirect Other fint varieties – Rullen, Simpelveld and percussion on cores with a unidirectional Valkenburg – from the Limburgian and Belgian production surface. The lack of voluminous cores chalk districts are less well represented in these in the excavated assemblage suggests that these three assemblages. None of the selected sites blades may have been manufactured elsewhere. has produced evidence for specialised fint Small cores still bear cortex and mostly show knappers;103 stone tool manufacture was negatives of fakes and smaller, less regular probably organised at the household level. blades. The fint is mostly of northern origin; Few lithic assemblages atributable to the a limited number of artefacts are made of Swiferbant Culture have been studied since the Rijckholt-type fint and possibly light grey Belgian major excavations at Hardinxveld-Giessendam fint, both of which can be found in Meuse gravel and Hoge Vaart-A27, more than 15 years ago.104 deposits. One artefact is made of white-speckled Roterdam-Groenenhagen yielded an assemblage black fint, which is believed to originate from the that shows technological afnities with the early Cap Blanc Nez (Calais, France) area. Another stage of the Swiferbant Culture (fg. 3.8).105 The artefact – heavily burned – may be of manufacture of tools was based on regular Wommersom quartzite. blades, which provided blanks for the production

102 Brounen & Peeters 2009. 103 The recently excavated site of Maastricht-Cannerberg has produced evidence for supra-local production, which may indicate specialised knapping (Knippenberg 2016). 104 Peeters et al. 2001; Van Gijn, Beugnier & Lammers-Keijsers 2001; Van Gijn, Lammers-Keijsers & Houkes 2001. 105 Niekus & Machiels 2013. It should be noted that early Swiferbant Culture ﬂint technology is in fact a continuity of Late Mesolithic technology and therefore cannot safely be distinguished from it without further contextualisation and dating evidence. In the case of Roterdam-Groenenhagen, potery atributable to the Swiferbant Culture Fig 3.8 Blades from Roterdam-Groenenhagen. The specimen in the lower right corner is retouched and slightly was found in association with the ﬂint assemblage (Bloo 2010). denticulate (from appendix II: PUBID 468). Scale bar in centimetres. 65 —

Fig 3.9 Flint cores of the Michelsberg Culture. Iteren-Voulwames: (a) core used as a hammer stone, (b) polished axe reused for ﬂake production, (c) pick-shaped core; Brunssum-Molenbeek: (d) coarse blade core (from appendix II: PUBID 95, 309). Scale bar in centimetres.

Middle Neolithic assemblages come from a Hatemerbroek, Groningen-Europapark, number of sites; they particularly provide insight Groningen-Helpermaar and Leeuwarden- into technological aspects of Michelsberg (Iteren- Hempens).110 Typically, these assemblages are Voulwames and Brunssum-Molenbeek),106 dominated by the production of fakes. With the Hazendonk (Schipluiden, Rijswijk-Ypenburg and exception of Michelsberg assemblages, blades are Nijmegen-Oosterhout),107 Vlaardingen rare. The Michelsberg assemblages still present a (Hazerswoude-Rijndijk, Hellevoetsluis- degree of standardised core reduction from a 106 Appendix II: PUBID 95, 309. Ossenhoek, Roterdam-Beverwaard),108 Stein single platform for blade production, whilst fake 107 Appendix II: PUBID 338, 463, 732. (Iteren-Hoeve Haertelstein, Hof van Limburg and cores are reduced from 1-3 platforms, yet not in an 108 Appendix II: PUBID 272, 463, 740. 109 Appendix II: PUBID 219, 253, 308. Haren-Groenstraat),109 and TRB (Knooppunt opportunistic or random fashion (fg. 3.9). 110 Appendix II: PUBID 260, 741, 750, 752. 66 —

Fig 3.10a Points with invasive retouch from Schipluiden (from appendix II: PUBID 732). Scale bar in centimetres.

Retouched tools (pointed blades, scrapers) are However, the majority of lithics connect to a less ofen large (>5 cm) and produced from wide standardised reduction of (beach-)rolled pebbles blades and fakes. A comparable picture emerges that may have been collected in the vicinity of the from the Hazendonk assemblages, which contain setlement locations. In the context of Stein, big retouched blades and fakes; other typical Vlaardingen and TRB, blades become scarcer, and elements are short and elongated points with normally these are small and not produced in a invasive retouch (fg. 3.10). These elements are systematic manner; it would be beter to describe mostly made out of what is described in the report these as long fakes or blade-like fakes. Evidence as non-local fint, notably originating from sources for blade production, however, is present in the in Belgium (Haspengouw?), southern Limburg Stein assemblage of Haren-Groenstraat (fg. 3.11), 111 Van Gijn et al. 2006; Houkes 2008. See section 4.2.2 for further discussion. and, possibly, northern France (Cap Blanc Nez).111 where several tools made out of blades were 67 —

Fig 3.10b Macro-tools of the Michelsberg Culture (a) and Hazendonk group (b-j). Iteren Voulwames: (a) pointed blade; Schipluiden: (d) pointed blade, (b, c, e) pointed ﬂakes, (f, i-k) scrapers, (g) borer, (h) retouched ﬂake (from appendix II: PUBID 309, 732). Scale bar in centimetres. 68 —

Fig 3.11 Cores (scale 1:2) and tools (scale 1:1) of the Stein Group from Haren-Groenstraat (from appendix II: PUBID 253). 69 —

Fig 3.12 Cores from Vlaardingen and TRB assemblages: (a, b) Hellevoetsluis-Ossenhoek, (c-i) Knooppunt Hatemerbroek (from appendix II: PUBID 272, 260). Scale bar in centimetres.

found, notably transverse arrowheads (which Hardinxveld-Giessendam Polderweg and De were also produced on fakes). At this site, there is Bruin.112 Owing to the positive results and the also evidence for the application of direct sof involvement of the national heritage agency, percussion. Vlaardingen and TRB assemblages several subsequent projects have paid atention appear to be dominated by opportunistic fake to use-wear analysis to obtain insight into craf production from small erratic nodules – probably activities within setlements. In many of the collected from outcrops in the vicinity – by means reports in our sample, typological classifcations of direct hard hammer percussion, and are used to draw conclusions about activities. sporadically by bipolar percussion (fg. 3.12). But use-wear analysis has shown that the Typically, tools are not very standardised. relationship between typology and tool function is not straightforward. Although use-wear Tool typology and function analysis has not been executed as a standard In the early days of development-led within development-led projects, valuable commercial archaeology, some 15 years ago, insights have been obtained. This is not only the few lithic assemblages were subject to use-wear case for tools made out of fint, but also for tools analysis of any reasonable intensity. These made out of other types of rock. There is analyses came from the Late Mesolithic and frequent cross-referencing to research results, 112 Peeters et al. 2001; Van Gijn, Beugnier & Lammers-Keijsers 2001; Van Gijn, Early Neolithic sites of Hoge Vaart-A27 and but this is mainly due to the fact that nearly all Lammers-Keijsers & Houkes 2001. 70 —

use-wear analysis has been performed by one the most comparable, as they were excavated in small group of researchers.113 The obvious more or less similar ways (sieving of 50 × 50 cm advantage thereof is analytical and interpretive grid cells). Clearly, the percentage of analysed coherency, as well as an increase of knowledge artefacts from these sites is low, with the most base upon which subsequent projects can build. extreme case being Groningen-Europapark. Nonetheless, problems of representativeness The bigger samples, from Roterdam-Yangtze may occur, owing to both sample size and Harbour and Schipluiden, are the result of the sampling strategy. researchers making an explicit choice to pay Table 3.1 provides an overview of fint more atention to use-wear as an important assemblages that have been subject to use-wear source of information. Furthermore, the analysis. The relative importance of the sample sampling criteria difer amongst sites. With size for the sites is difcult to compare, as the these diferences, it is impossible to say to what fgures do not take into account diferences in extent the results obtained are representative of excavation strategy and site taphonomy. variability at both the intra- and the inter-site Dronten-N23, Ede-Kernhem, Leeuwarden- level. Hempens and Groningen-Europapark are maybe

Table 3.1 Sample size and selection criteria for use-wear analysis.

Site N fint N tools Use-wear Selection criteria; analyst

Roterdam-Yangtze 2976 114 170 • intentional retouch Harbour (5.7%) • straight working edge ≥ 5 mm • point prodruding far enough to be functional • straight, blunt working edge • absence of heavy post-depositional damage • analyst: Verbaas

Dronten-N23 25040 2216 243 • selection based on assessment results (1.0%) • analyst: Verbaas

Leeuwarden-Hempens 19144 1729 354 • representation of each typological category (1.9%) • sample of unretouched artefacts • equal spatial distribution • analyst: pilot study by Schreurs; expanded study by Beugnier

Ede-Kernhem 23292 769 50 • selection of retouched artefacts with the exclusion of (0.2%) microlithic points • unretouched fakes and blades • analyst: Beugnier

Knooppunt Hatemer- 10151 1187 236 • representation of each typological category broek section Hanze- (2.3%) • sample of unretouched artefacts lijn • exceptional objects • sample from each cluster complex • absence of heavy post-depositional damage • analyst: Verbaas

Knooppunt 6925 645 203 • representation of each typological category Hatemerbroek (2.9%) • sample of unretouched artefacts section Bedrijven- • exceptional objects terrein (TRB context) • sample from each cluster complex • absence of heavy post-depositional damage • analyst: Verbaas

Schipluiden 2666 1123 373 • representation of each typological category (14.0%) • preferential treatment of blades • analyst: Van Gijn, Van Betuw, Verbaas

Rijswijk-Ypenburg 15515 356 133 • intentional retouch (0.9%) • regular edge morphology • representation of each typological category • sample of unretouched artefacts • analyst: Van Gijn, Verbaas

Hellevoetsluis- 2823 214 81 • retouched tools Ossenhoek (2.9%) • analyst: Metaxas 113 Most are connected to the Laboratory of Groningen- 10513 310 10 • selection criteria unknown Material Culture Studies at Leiden Europapark (0.1%) • analyst: Verbaas, Garcia-Diaz University (now Foundation LAB). 71 —

Typo-morphology and function. The typological classifcation of tools represents the basic exercise on which reports expand. Indeed, in many cases the ‘analysis’ is restricted to the presentation of a frequency count of tool types and brief comments on each of these types. In those cases where use-wear analysis has been performed, a more in-depth discussion is presented, including whether use-wear analysis confrms or repudiates a correlation between tool typology and function. Confrmed correlations are particularly found for Mesolithic scrapers, which normally appear to have been used for working fresh or dry hides and are, indeed, found to have been used in a scraping motion (Knooppunt Hatemerbroek, Roterdam-Yangtze Harbour, Epse-Olthof, Dronten-N23 and Ede-Kernhem).114 Hide scraping also seems to have been the dominant function of Neolithic scrapers, yet there also is evidence for the working of bone or antler, plant material, and mineral or potery – and even for their use as a strike-a-light (Schipluiden, Hellevoetsluis-Ossenhoek, Groningen-Europapark, Knooppunt Hatemerbroek).115 Another positive correlation pertains to borers and piercers, which were indeed used for perforation of various materials, such as bone, bark, sof wood, mineral (potery?) and occasionally shell and jet (Roterdam- Yangtze Harbour, Dronten-N23, Ede-Kernhem, Groningen-Europapark).116 More surprising is the less clear correlation between arrowheads and function. The broad and variable category of Mesolithic microlithic ‘points’ – thought to have Fig 3.13 Example of a sloted point (17.5 cm in length) functioned as projectiles (i.e. related to with hafed microliths from the Kongemose at Tågerup, subsistence activity) – appears to include other Sweden (from Karsten & Knarrström 2003). Lef: the functions, notably drilling and piercing, as well object as it was found (in a grave), middle: the structure as cuting and scraping (Roterdam-Yangtze of microlithic inserts, right: reconstruction. Harbour, Dronten-N23, Knooppunt Hatemerbroek).117 Neolithic arrowheads, in contrast, seem to have been used as projectile mind, however, that the typological categories points only (Schipluiden, Hellevoetsluis- are defned by archaeologists. Mesolithic Ossenhoek and Knooppunt Hatemerbroek).118 microliths represent a broad category, mostly Although it is possible that diferences consisting of variably shaped ‘insets’ for between the Mesolithic and Neolithic relate to compound projectile heads, whereas Neolithic sample size (representativeness), there is arrowheads are single-element projectile heads nothing to suggest that this is the case here. (fg. 3.13). The categorisation of individual There seems to be a ‘crossover’ in functional projectile elements can potentially lead to 114 Appendix II: PUBID 259, 464, 538, 733, range of Mesolithic and Neolithic scrapers and confating similarly shaped items that had 745. 115 Appendix II: PUBID 259, 260, 272, 732, arrowheads, which may suggest diferences in diferent functions. On the other hand, the 752. the formal relationship between particular chances that a scraper will be confused with a 116 Appendix II: PUBID 259, 745, 752. 117 Appendix II: PUBID 259, 462, 538. functions and tool type. We have to bear in projectile point will be (close to) zero. 118 Appendix II: PUBID 259, 260, 272, 732. 72 —

Nonetheless, we have to remain aware of a potential functional variability represented degree of mismatch between typological within assemblages, which has implications for categories and function, especially in the case of the interpretation of site function, or of even atypical tools.119 To stick to the above examples, broader views on socio-cultural relationships the fact that Mesolithic scrapers were between people and tools (or technology). predominantly used for scraping hides does not imply that this was also the case in the Neolithic. Tool use and activity. In view of the above, And elements we call ‘arrowheads’ may have questions about the use of stone tool and the served another function. activities performed with them cannot be The correlation between typological answered on the basis of typology alone. categories and function amongst tools made out Use-wear analysis performed in the context of of rocks other than fint (or out of related development-led projects provides important siliceous rocks) has barely been investigated, for new insight. This work has focussed not only on two reasons. First, for a long time, existing ‘formal’ tool types – intentionally modifed or typologies were crude; they only became more shaped tools – but also on the use of unmodifed diferentiated over the past 15 years.120 Second, artefacts, such as unretouched fakes and blades. use-wear analysis on non-fint tools is a Table 3.2 summarises the main fndings for relatively recent development. Although in Mesolithic and Neolithic assemblages. general terms a correlation is found between The uses to which unretouched and tool type and function – hammer stones were retouched fakes and blades were put not seem used for pounding and crushing, grinding stones to have difered, which could imply that the for grinding or polishing, querns for grinding – activities involved did not require specifc types the results from development-led research point of tools. However, to some extent people to variability within these functional groups (see engaged in purpose-specifc selection of the next section). Also, it has been observed that particular blanks. In Mesolithic and Early individual tools may have had multiple Neolithic contexts, for instance, unretouched functions, for instance, pounding as well as regular blades are frequently used for a specifc grinding. The use of broad typological categories way of plant working (Dronten-N23, Roterdam- therefore comes at the risk of neglecting the Yangtze Harbour, Ede-Kernhem, and

Table 3.2 Established functions and contact materials of ﬂint tools from Mesolithic and Neolithic contexts.

Category Mesolithic Neolithic

Unretouched fakes varied use (cuting, scraping, drilling) on plant varied use (cuting, scraping, spliting) on hide material (incl. wood, bark, …) hide and even and plant material (incl. wood) anorganic material and fsh

Retouched fakes varied use (cuting, scraping) on hide, wood, bark varied use (cuting, scraping, wedge) on hide, and bone plant material (incl. wood) and bone

119 It should be noted, however, that we Unretouched blades varied use (cuting, scraping, drilling) on plant varied use (cuting, scraping) on hide and plant may be dealing with a problem of material (incl. wood, bark, …), hide and even material researcher-related bias. Extensive use- anorganic material, bone/antler and fsh wear analysis on microliths from the Retouched blades varied use (scraping, drilling, butchering) on hide, varied use (cuting, scraping) on hide and plant Belgian sites of Verrebroek-Dok and plant material (incl. wood), fsh and bone/antler material (incl. wood) Aven Ackers by Valerie Beugnier only identiﬁed a projectile function. She Scrapers hide working hide working, working of bone, wood, plant excluded microliths from her material and mineral subsequent analysis of the Ede- Kernhem assemblage from the Borers perforating bone, wood and mineral perforating bone, wood, plant material and Netherlands because she didn’t expect mineral to ﬁnd anything new. The Dutch analyses that show a more diverse Piercers perforating and engraving bone and plant n.a. functional range amongst microliths material have mostly been undertaken by Arrowheads hunting, and working of organic material hunting Annemiek Verbaas. 120 See e.g. Van Gijn & Houkes 2006; Axes n.a. heavy and light duty wood working Devriendt 2014. 73 —

Fig 3.14 Examples from Hoge Vaart-A27 of highly reﬂective gloss caused by transverse working of a plant material of unknown nature (from Peeters, Schreurs & Verneau 2011). Scale bar in centimetres.

Leeuwarden-Hempens).121 Although the a dryland site, shows that this suggestion may distinctive wear traces have been identifed by not be correct (fg. 3.14). But there are some use-wear analysts for more than two decades,122 difculties here: many ‘dry’ sites are located next it is still unclear what plant material was to or near river gullies/valleys, and stone tools involved and how it was worked. It has been from these sites may well bear signs of wetland suggested that this particular function was exploitation. In other words, tools may have related to wetland sites, but that it disappeared been discarded at other places than where they once crop cultivation became part of the were used, and processing of – in this case – wetland subsistence economy.123 However, the plant material, even material originating from occurrence of such use-wear traces on blades wetland environments, may have occurred at from Ede-Kernhem, which is a dryland site, and some distance from the source area. 121 Appendix II: PUBID 462, 538, 745, 776. 122 Jensen 1994. from Dronten-N23, which can also be considered 123 Van Gijn 2010, pp. 65-66. 74 —

Fig 3.15 Use-wear traces of the working of shell (a) and jet (b) found on ﬂint tools from Roterdam-Yangtze Harbour (from appendix II: PUBID 462). Object drawings scale 1:1.

Whereas the above case shows that use- development-led projects has opened an wear analysis can raise questions without unexpected window on Mesolithic craf activities. providing answers, use-wear analysis can also Along similar lines, use-wear analysis, ofer insight into activities that otherwise would combined with residue analysis, has shed light remain invisible. At Roterdam-Yangtze Harbour, on activities that involve the use of non-fint evidence was found for the Mesolithic working of tools (table 3.3). Litle was known about the shell and jet (fg. 3.15).124 The working of shell actual use of various tools that are described in a involved piercing (using a borer), engraving, and generic fashion as, for instance, ‘hammer stone’, scraping (using a burin–scraper combination ‘retouchoir’, ‘anvil’, ‘quern stone’ or ‘boiling tool), but no shell was found. The evidence for stone’. In fact, artefacts made out of rock types the working of jet in the Mesolithic is even more other than fint – or other rocks with comparable surprising. Cuting and scraping was executed fracture characteristics – have received hardly with a retouched fake and the lateral side of a any atention whatsoever. This atitude is burin, but jet objects are absent in the Mesolithic changing, and we can now refer to some good record of the Netherlands and adjacent countries. examples of what information can be drawn In contrast, during the Neolithic, beads made out from this, to some, unexciting fnds category. of jet are known from several sites (e.g. The majority of information concerns the Swiferbant-S22, Schipluiden, Rijswijk- Neolithic rather than the Palaeolithic and Ypenburg).125 The absence of both shell and jet in Mesolithic, for which functional interpretations the Mesolithic record of the Netherlands may, are still based on intuition. however, be the result of preservation A few Late Palaeolithic sites (Geldrop- 124 Niekus et al. 2015. conditions.126 The dry and acidic sandy soils of the Aalsterhut, Knooppunt Hatemerbroek)127 have 125 Devriendt 2014; Van Gijn 2006, 2008. A recently excavated Swiferbant site at Netherlands are unfavourable; both materials do produced non-fint tools, but the Mesolithic Utrecht-Nieuwegein yielded a big survive under waterlogged and alkaline assemblages of Dronten-N23 and Ede-Kernhem pendant made out of jet. 126 Some caution is perhaps advisable, as conditions, but few waterlogged Mesolithic sites account for most of the data to date, perhaps there may be a researcher-related bias. have been investigated to date. Use-wear (partly) due to the extent of the excavations 127 Appendix II: PUBID 216, 259. 128 Appendix II: PUBID 538, 745. analysis conducted in the context of (fg. 3.16).128 Hammer stones of variable shapes 75 —

Fig 3.16 Mesolithic stone tools. Ede-Kernhem: (a) rubbing stone, (b, c) hammer-/rubbing stone, (d) hammer stone/ anvil; Dronten-N23: (e) fragments of plaquetes, of which the smallest shows traces of woodworking, (f ) fragment of a hammer stone used as a scraper (from appendix II: PUBID 538, 745). 76 —

and sizes were used for (fint) knapping as well represented in the northern Netherlands, where as crushing (hazel-)nuts. The later activity was they occur in moraine deposits. also executed by means of so-called ‘hand Although use-wear and residue analysis is not stones’ and ‘nether stones’.129 Nether stones, ofen performed (see table 3.3), it has revealed the however, were principally used for the polishing grinding of grain, peas and other plant materials of sof materials, which was also done with on LBK saddle querns and other tools from Stein- plaquetes, used for wood polishing and Heidekampweg. One quern showed evidence for processing of plant material. With the exception the grinding of cereal grains on the fat side, and is of plaquetes, which consist of thin sandstone coated with red ochre on both sides. Since no slabs, the afore-mentioned tools consist of traces of ochre grinding were found, it has been cobbles, mainly of quarzitic sandstone, quartz, suggested that the ochre coating was deliberate.137 and sandstone. Several sites in the northern Evidence for grinding of grain and other Netherlands (Finsterwolde-Ganzedijk, processing of plant materials was found on querns Groningen-Meerstad) have produced hammer and grinding stones at the Middle Neolithic site of stones of granite.130 It is likely that cobbles were Schipluiden (Hazendonk 3), as well as on querns at selectively collected on the basis of size, shape the Middle Neolithic sites of Hazerswoude-Rijndijk and raw material so that they could be used in (Vlaardingen) and Knooppunt Hatemerbroek unaltered condition. (TRB).138 Whetstones were used not only to grind The presence of fakes and debitage at a or polish bone tools and fint axe blades, but also number of Mesolithic sites (Dronten-N23, Ede- to polish sof materials, possibly for the Kernhem, and Roterdam-Yangtze Harbour)131 production of (amber) ornaments (for a discussion indicate knapping of quartzite and quarzitic of ornaments, see section 3.3.1). Woodworking sandstone in particular, although it cannot be was mainly done with axes and adzes, although excluded that part of this material did not result some non-fint stone fakes were used as well. from deliberate actions, since fakes and Only at Swiferbant Culture sites have non-fint fragments are sometimes detached from stone fakes and blades been modifed into hammer stones by accident. However, the retouched pieces and scrapers (Roterdam- presence of fake negatives on larger fakes atest Groenenhagen, Swiferbant-S3).139 129 Only for Dronten-N23 (Knippenberg & Verbaas 2012) was a clear distinction to repeated actions, that is, serial detachment of made between actively used tools (hand fakes from a core. Some fakes show evidence of stones, grinding stones) and passively used tools (stationary stones/querns, having been used, e.g. for scraping 3.2.2 Organic resources for artefacts whetstones). (Dronten-N23) or cuting (Schipluiden).132 130 Appendix II: PUBID 210, 255. 131 Appendix II: PUBID 462, 538, 745. However, it remains unclear how systematic the 132 Appendix II: PUBID 538, 732. knapping of these types of rock was. Deliberately Artefacts of other than lithic raw materials are 133 This tool was used as a hammer stone on a medium-hard material. Traces of shaped tools remain an exception in Mesolithic likely underrepresented in the archaeological hafing were found inside the contexts: Knooppunt Hatemerbroek and record due to diferences in preservation. hourglass-shaped perforation. 134 Appendix II: PUBID 260, 732, 740. Roterdam-Beverwaard both yielded a complete Uncharred objects made of bone, antler, wood 135 A word of caution is nonetheless in mace head (known in German as a Geröllkeule),133 and plant fbre, for instance, can only survive in order, as non-flint lithic artefacts seem to have been analysed superficially, if at which in both cases had broken in two.134 wet contexts and particular dry sediments, such all. Systematic analysis of such remains Non-fint stone tools seem to be more as calcareous loess; these kinds of matrix are may lead to a different picture. 136 Depending on visible traces and use- common on Neolithic sites.135 These tools mostly almost absent in the Netherlands. Sites such as wear analysis, but also on the researcher concern querns and grinding stones, whetstones, Hardinxveld-Giessendam (Polderweg and De and which typology is being employed, either the more generic term ‘grinding polishing stones, hammer stones, anvils, adzes Bruin), Hoge Vaart-A27, and Schipluiden stone’ is used, or specific terms are used and axes (fg. 3.17).136 In contrast to the situation demonstrate the rich spectrum of artefacts to define more specialised tools (querns, whetstones). in the Mesolithic, deliberate shaping of such made out of such organic materials.140 It is more 137 Knippenberg & Verbaas 2012, 117 tools was common, and was conducted by than likely that the lithic component, which is 138 Appendix II: PUBID 260, 732, 740. 139 Appendix II: PUBID 732; Devriendt 2014. faking and pecking – hence the frequent overwhelmingly present in the archaeological 140 Laarman 2001; Louwe Kooijmans, occurrence of fakes and debitage. There tools record, in fact represents a relatively small Oversteegen & Van Gijn 2001; Louwe Kooijmans, Vermeeren & Van Waveren were predominantly made out of sandstone, portion of the full range of artefacts that was 2001; Louwe Kooijmans, Hänninen & quarzitic sandstone and quartzite. Gneiss, quartz actually produced, used and discarded. Any Vermeeren 2001; Van Gijn 2006; Louwe Kooijmans & Kooistra 2006. and granite are less common, but they are well possibility of expanding our knowledge about 77 —

Fig 3.17 Neolithic stone tools: Schipluiden: (a-e) querns, (f ) rubbing stone, (l) axe, (m) axe fragment; Knooppunt Hatemerbroek: (g) rubbing stone, (h, i, k) querns, (j) hammer-/rubbingstone, (n) axe (from appendix II: PUBID 260, 732). Scale 1:4. 78 —

Table 3.3 Established functions of non-ﬂint stone tools.

Site Con- Quern Grinding stone Plaquetes Polishing Whetstone Hammer Hammer Anvil Axe / Fire Flakes / text (maalsteen) (wrijfsteen) stone (slijpsteen) stone stone / (aam- adze / cracked debitage (polijst- (klop- grinding beeld) chissel rocks steen) steen) stone Geldrop- Ab no analysis no analysis Aalsterhut Knooppunt LP no analysis no analysis Hatemerbroek Dronten-N23 EM- crushing crushing (hazel) polishing / polishing / pounding MM middle hard nuts grinding[1] grinding sof (middle) material and plant material stone hard (hazel) nuts (wood?) material Roterdam- EM- polishing / grinding (?) cooking no analysis Yangtze Harbour MM wood, crushing stones plant material Meerstad-Site 2a MM no analysis no analysis cooking / hearth stones Finsterwolde- M no analysis no cooking / Ganzedijk analysis hearth stones Ede-Kernhem EM no analysis no analysis no analysis no cooking no analysis analysis stones Hoogezand- MM- no analysis no analysis Vosholen LM Olieveld MM- no analysis hearth stones no analysis Schoonebeek- LM Site 14 Casteren-Wagen- LM no analysis broeksloopje Stein- LBK processing of processing grain processing middle hard no analysis crushing plant woodwor- (?) cooking no analysis Heidekampweg grain, colou- and other plant peas or plant material material? king stones red with red material similar plant (wood?), stone ochre material (red ochre?) Beek-Kerkeveld LBK no analysis no analysis no analysis Iteren- MC (?) no (?) no (?) cooking Voulwames analysis analysis stones Roterdam-De Sw (?) no analysis no analysis no no analysis Zwanen Rietpark analysis Roterdam- Sw also scraper, Groenenhagen yet no analysis Nijmegen- Hd3 no analysis no analysis no cooking / no analysis Oosterhout analysis hearth stones Schipluiden Hd3 processing of pounding, plant polishing fint pounding wedge in temper and cuting and grain processing and bone woodwor- cooking / scraping king hearth stones wood Iteren-Hoeve St no analysis no analysis no analysis no analysis Haertelstein Hof van Limburg St (?) no (?) no analysis analysis Haren- St (?) no (?) no analysis (?) no (?) no analysis no analysis temper and Groenstraat analysis analysis cooking / hearth stones Beverwaard- Hd3 no temper no analysis Tramremise analysis Hellevoetsluis- Vl no analysis no analysis no analysis Ossenhoek Hazerswoude- Vl processing of grinding bone, pounding used for no traces temper and Rijndijk grain fint and wood hard middle of use cooking material hard stones material (nuts?) Groningen- TRB no analysis no analysis no analysis no analysis no analysis temper no analysis Europapark Helpermaar TRB no analysis no analysis no analysis no analysis temper no analysis Hatemerbroek- TRB processing of processing cereals, polishing pounding crushing or cooking / Bedrijventerrein grain plant material and sof stone hard bruising plant hearth stones middle hard (amber?) material material and material (middle) hard material Leeuwarden- TRB no analysis Hempens

[1] In Dutch the term slijpen is used. 79 —

the use of organic artefacts is therefore of major Bruin) and Hoge Vaart-A27,147 the Middle importance if we are to build a somewhat more Neolithic assemblages mentioned above are representative record of early prehistoric socio- rather impoverished with regard to the range of technological traditions at a geographical scale tool types. Their earlier, wetland counterparts beyond the Netherlands. With a few exceptions show more variety with the inclusion of of projects that have employed use-wear matocks, adzes, axe shafs, and other analysis, the contribution that development-led implements for which no clear defnitions exist. research has made in this respect is limited This contrast cannot be explained by structural despite the potential. diferences in preservation. At the same time, it Artefacts of organic material can be divided does not signify a trend of decreasing importance in various categories, ranging from construction of bone and antler tools. An extremely rich and material, to implements, to jewellery. Organic varied assemblage from Emmeloord-J97, which material represents material that was still alive may date from the Middle to Late Neolithic/ or had ‘recently’ died at the moment of Bronze Age, suggests otherwise.148 Here a variety procurement. We therefore exclude amber and of needles, fshhooks, spatulas, matocks and a jet from the category organic material. Although possible dagger were found, in addition to chisels both amber and jet have an organic origin, they and awls. represent fossilised forms of resin and Whilst these artefacts are generally carbonised (petrifed) wood, respectively, that described in detail and ofen illustrated in the were collected in that form as a raw material for reports, interpretation of their function is ofen artefacts. Like petrifed wood, we classify amber limited. Use-wear analysis of tools from and jet as rock. Below we discuss the use of Schipluiden suggests that awls may have been organic material for the production of used in the production of basketry/fabrics and in implements and jewellery. hide working; bone chisels were used for woodworking.149 Indeed, it is suggested that Implements of bone and antler bone and antler tools should be considered a Bone and antler represent important raw structural part of toolkits that also included materials for the production of a wide range of stone tools.150 Furthermore, in consideration of implements. Adzes, axes, chisels, points and awls production waste and other traces of tool are the main categories, and items from these manufacturing, it would seem that the categories have also been found in the context of technology involved is a continuation of a 141 Increasing numbers of bone and antler development-led projects. Mesolithic tools Mesolithic tradition.151 tools dating to the Mesolithic are being collected on Dutch beaches. These remain exceptional, as few sites with well- originate from oﬀshore sand extraction preserved organic remains dating to this period Artefacts of wood and other plant material zones. The sand is used for beach replenishment. As yet, there is litle have been investigated.141 An example is a The number of sites that have produced archaeological monitoring of this complete chisel made out of a metatarsus of an artefacts of wood and other plant material is commercial extraction activity. 142 The report claims a Late Mesolithic age aurochs found at the site of Mildert-Tungelroyse also limited.152 Small fragments of rope for this object based on a 14C-date Beek (fg. 3.18).142 A fragment of an antler consisting of twisted plant fbre – plant species (appendix II: PUBID 395, p. 37), but the date itself is not provided. tool – an adze or chisel – is reported from the undetermined, due to the small size of the 143 Zeiler & Brinkhuizen 2015. Early to Middle Mesolithic site of Roterdam- fragment – of Late Mesolithic age were found in 144 De Vries 2008; Van Gijn 2006. 145 Van Dijk 2009. Yangtze Harbour.143 The remainder of bone and an auger sample at Roterdam-Emplacement 146 Van Dijk 2009. antler tools comes from the Middle Neolithic CS.153 These fragments possibly belong to a 147 Laarman 2001; Louwe Kooijmans, Hänninen & Vermeeren 2001; Louwe Hazendonk Culture sites of Schipluiden and fshnet or fsh trap, as has also been suggested Kooijmans, Overstegen & Van Gijn 2001; Rijswijk-Ypenburg144 and the Vlaardingen Culture for fragments of rope found at the Late Louwe Kooijmans, Vermeeren & Van Waveren 2001. site of Hellevoetsluis-Ossenhoek,145 mostly Mesolithic site of Hardinxveld-Giessendam 148 Bulten et al. 2002. chisels and awls. The later site also produced Polderweg.154 Fish traps made out of shoots and 149 Van Gijn 2006. 150 Van Gijn 2006. two hollow tubes of unknown function made of rope are not unusual in wetland contexts, and 151 Van Gijn 2006. goose long bone.146 Production waste points to well-preserved examples are known from 152 For a recent overview, see Lange 2017. 153 Brinkkemper 2007. local production. Compared with the rich Late Neolithic sites such as Hoge Vaart-A27 (early 154 Louwe Kooijmans, Vermeeren & Van Mesolithic and Early Neolithic assemblages of Swiferbant), Bergschenhoek (Swiferbant), and Waveren 2001. 155 Bulten et al. 2009; Hamburg et al. 2001; Hardinxveld-Giessendam (Polderweg and De Emmeloord-J97 (TRB, also Late Neolithic).155 It is Louwe Kooijmans 1986. 80 —

Fig 3.18 Bone and antler artefacts from Mesolithic and Neolithic sites. Roterdam-Yangtze Harbour: (a-b) fragments of antler adzes; Schipluiden: (c) cut-oﬀ distal end of red deer metatarsal, (d, h) antler waste products originating from tool production, (e) bone awls produced out of metapodia, (f ) fragment of an antler perforated base axe, (g) bifacially worked proximal part of an antler beam and adhering pedicle(from appendix II: PUBID 462, 732). Objects a, b and e scale 1:2; objects c, d, f, g and h scale 1:4. 81 —

unclear how to interpret nine strips of bark deriving from Pomoideae (which includes hawthorn and apple) that were found in a well at Schipluiden.156 Wooden implements were found at the Hazendonk group sites of Schipluiden and Rijswijk-Ypenburg (table 3.4).157 The bow fragment found at Schipluiden is made out of juniper wood; this is rather surprising, because Mesolithic and Neolithic bows from the Netherlands are usually made out of elm and yew.158 It has been suggested that the later two types of wood were not locally available, and that juniper wood should be considered a second-best option: people did not see the need to go to a whole lot of trouble in order to acquire other, more suitable wood types. This is all the more remarkable, since paddles made out of ash wood were found at the same site. Again, there is no evidence for local stands of ash, which suggests that the wood or maybe even the paddles were been brought in from elsewhere. In contrast, the axe hafs from both sites are made out of various types of wood, but not ash, although ash is known to have been preferentially used for axe hafs. Typically, the Schipluiden axe hafs are made of branches instead of sturdier trunk wood.159 As stated in the report, this may point to local – and ad hoc (?) – production of wooden implements and technological choices that did not involve external contacts or long-distance forays, with the exception of paddles. One crescent-shaped plank of ash wood whose function has not been determined possibly represents a canoe partition (fg. 3.19).160 If correct, this suggests that waterborne technology involved selective use Fig 3.19 Functionally unidentiﬁed wooden artefacts from of wood. Schipluiden (from appendix II: PUBID 732). Scale 1:4.

Table 3.4 Wooden implements and used wood.

Implement/tool Schipluiden Rijswijk-Ypenburg

Bow 1 juniper -

Paddle 2 ash -

Axe haf 8 spindle; hazel; alder; willow; Pomoideae 1? willow

Sickle haf? - 1 oak

Pointed object 11 buckthorn; spindle; hazel; juniper; honeysuckle; Pomoideae 2 juniper*; buckthorn 156 Louwe Kooimans & Kooistra 2006. Functionally not interpretable 10 ash; sloe; alder; Pomoideae; cherry/plum? - 157 Kooistra 2008; Louwe Kooimans & Kooistra 2006. 158 * This object somewhat resembles a leister prong, also known from Bergschenhoek and numerous sites in southern Louwe Kooijmans & Kooistra 2006, 228. 159 Louwe Kooijmans & Kooistra 2006, 234. Scandinavia. 160 Louwe Kooijmans & Kooistra 2006, 235. 82 —

Jewellery/ornaments Items that have been interpreted as jewellery or ornaments made out of organic material are rare. Only one example of a (calcined) single tubular bead made out of a hollow long bone of a bird has been reported from the Early to Middle Mesolithic site of Roterdam-Yangtze Harbour (fg. 3.20).161 As mentioned in section 3.2.1, this site also yielded evidence of the processing of shell, notably piercing, engraving and scraping. This activity was most likely related to the production of ornaments. The site itself has not produced any shell ornaments; nor has any other Mesolithic site in the Netherlands. Shell ornaments are, however, known from several Mesolithic sites elsewhere in north-west Europe. Middle Neolithic ornaments made out of Fig 3.20 Mesolithic and Neolithic bone beads: organic material were retrieved from the (a) Roterdam-Yangtze Harbour; (b, c) Schipluiden (from Hazendonk Culture sites of Schipluiden and appendix II: PUBID 462, 732). The broken bead from Rijswijk-Ypenburg.162 Schipluiden yielded two Roterdam-Yangtze Harbour measures 6.7 mm in beads similar to the one from Roterdam-Yangtze diameter; the beads from Schipluiden are scale 1:1. Harbour; both beads come from a child’s grave. At Rijswijk-Ypenburg, a bone fnger ring was found in a child’s grave, and a fragment of a not fully perforated tusk of wild boar from the occupation 3.3 Socio-cultural meaning layer may represent an unfnished pendant. At the Vlaardingen Culture site of Hellevoetsluis- Ossenhoek, three perforated teeth, of wild boar, The second theme to be addressed concerns the oter and dog, were found.163 The fnds from these meaning of materials in their socio-cultural sites show that ornaments are not exclusively context. Objects made out of various sorts of encountered in graves. This is not only the case for materials had particular ‘utilitarian’ functions late Early Neolithic and Middle Neolithic contexts, within specifc behavioural setings, and involved but also for earlier phases. Perforated tusks of wild a myriad of craf activities. Traditionally, boar, as well as perforated incisors of horse, catle, archaeologists tend to connect this ‘utilitarian’ oter and canines of dogs, are known from dimension of materials to the domestic and Swiferbant contexts.164 Interestingly, fnds from subsistence domains. However, objects also had the early Swiferbant Culture context of Hoge ‘non-utilitarian’ functions, of a more symbolic Vaart-A27 suggest that tusks of wild boar seem to nature. For certain objects, such as fint axes of have been systematically extracted from the Scandinavian origin, this may seem more obvious mandibles,165 possibly in order to make pendants than for objects that do not stand out from the out of them. ‘average’. Nonetheless, such objects could very Despite the fact that few ornaments made out well have had other intrinsic meaning to the of organic materials are known, the above people who used them, meaning which is simply examples demonstrate that there is a potential for difcult to recognise archaeologically. Part of the them to be found. Ornaments from burial contexts problem is us archaeologists: in trying to obviously have a beter chance of being recognised. understand the archaeological record, we break it Small items and fragments from fnds/cultural down into entities that are comprehensible to us. layers may, however, not easily be detected. An However, our interpretive framework need not 161 Zeiler & Brinkhuizen 2015. eye-opener is the discovery of a pierced plum stone have any connection whatsoever with the 162 Van Gijn 2006, 2008. in a screened archaeobotanical sample from the prehistoric perceptive frameworks within which 163 Van Dijk 2009. 164 Prummel 2016. site of Swiferbant-S4, which is interpreted as a these objects functioned. Another part of the 165 Laarman 2001. bead.166 Although this fnd falls in the category problem concerns the manifestation of the socio- 166 Personal communication Dr Mans Schepers, 4 September 2017. ‘lucky shot’, it shows that an eye for detail pays of. cultural meaning of materials. As this is an issue 83 —

Table 3.5 Summary of early prehistoric depositions.

Site Socio-cultural context Object(s) Depositional context

Dronten-N23 Late Mesolithic (?) A total of 41 artefacts; nodules, cores and tested blocks. Splinters (n=16), fakes and At the limits of a fint scater blades (n=9) and two tool fragments are considered to represent ‘setlement noise’.

Siddeburen-Foxhol Mesolithic A total of c. 500 fint artefacts; mainly knapping debris (splinters, fakes, blades, Two pits; separated c. 2m rejuvenation pieces; no cores) and some retouched pieces. from each other

Beek-Kerkeveld LBK Complete grinding set; traces of haematite only present on the grinding surface. Pit

Leur-Hernense Meer Swiferbant 115 fragments of a single (?) pot. Pit

Haren-Groenstraat Stein/Vlaardingen Two voluminous fint cores (eluvial Lanaye fint). Possibly in a pit

Knooppunt Hatemerbroek TRB large part of a pot (rim, neck, and body); part of a baking plate Posthole

Lomm-Hoogwatergeul Middle Neolithic? Complete polished fint axe (dünnackige Flintovalbeile). Transitional zone: river terrace to gully

Rijswijk-Ypenburg Middle Neolithic Complete pot + granite grinding stone Pot found in a pit; grinding stone on top of the pit

of context, the difculty is to determine which or caches. Of course, neither of these suggestions criteria should be met. can be excluded: people do lose things, and The NOaA 1.0 draws atention to ‘non- people do cache items, for instance, for future use utilitarian’ functions of materials and phenomena, or to protect valuables from looting. However, the in particular in connection with burial practices. contextualisation of such fnds, which are mostly Other non-utilitarian aspects concern ‘ritual’ or known from low-lying zones in the landscape ‘cultic’ depositions of objects, such as stone axe (ofen wetlands), has shed new light on the blades, ceramic vessels and skeletal remains. nature of these phenomena. Hierarchic paterns Context is primarily considered in reference to the in the nature of objects and their occurrence in localisation of such items in particular landscape the landscape or specifc features (e.g. burials) zones – brook valleys or wetland environments – suggests underlying principles of structuration or their occurrence outside setlements. From this that seem to be related to cosmological or perspective, the NOaA 1.0 states that “‘anomalous’ ideological considerations.168 From this features should be treated with considerable care: in a perspective, depositions of single or multiple later stage of research it could turn out that it concerns objects are frequently interpreted as burials or cultic depositions” (translation ours).167 manifestations of ritual activity. However, no explicit research questions Depositions represent a rather problematic concerning these aspects are included. category in development-led research. First, In the next few sections, we will consider such depositions are not easily noticed due to several manifestations of behaviour that shed their particularly localised occurrence, as an (some) light on the socio-cultural meaning of isolated object or a grouping of several objects. materials. For this purpose, we will not limit Second, the ofen ‘secular’ or ‘ordinary’ nature – ourselves to the more obvious, cosmology-induced simple knapping waste, a piece of bone, a aspects of burial and ritual. Cultural transmission of wooden stick – of objects frequently obstructs craf traditions and the act of exchange of materials interpretations outside the utilitarian domain. and knowledge, for instance, contribute just as Third, most feldwork is conducted on much to an establishment of socio-cultural setlement sites: depositions of objects need to relationship between people and materials. be distinguished from ‘plain setlement waste’. Finally, locations or landscape zones where depositions are likely to occur – e.g. brook 3.3.1 Materiality: the meaning of objects valleys, peatlands – are less ofen investigated. Despite these problems, several reports make explicit mention of such fnds (table 3.5). The Finds of single, isolated objects and hoards of nature of the depositions fts the known 167 Deeben et al. 2006, 32. objects have long been interpreted as lost items variability, but merits some further discussion. 168 Cf. Fontijn 2002. 84 —

The two depositions dated to the Mesolithic consist of fint-knapping waste. Although the Dronten-N23 deposition has been described as such, the lack of detailed information concerning the exact circumstances of this fnd makes it difcult to establish the behavioural context. The small concentration of fints will certainly have been deposited deliberately, but it remains unclear whether we are dealing with a cache – read, storage of fints for future use – or a dump of waste material (fg. 3.21). In the case of Siddeburen-Foxhol we are clearly dealing with simple knapping debris, but here we see deliberate deposition in two small pits. One wonders why a ‘random’ collection of production waste was put in a pit, whilst considerable quantities of such waste were found scatered over the surface. A similar situation was encountered at the previously mentioned site of Hoge Vaart-A27, where a random collection of knapping waste (100 fakes) was deposited in a small pit. Two other depositions at this site consist of several exhausted cores and conjoining large fakes, in one case, and tested nodules, in the other. All three depositions occurred in a swampy environment, at some distance from the sandy ridge on which vast quantities of setlement waste were lef during an early Swiferbant phase of repeated occupation.169 Another example comes from the Mesolithic site of Nieuwe Pekela 3, where burnt fint cores were found in a pit.170 This context has been interpreted as a cache, but this seems an unlikely option to us, as burnt fint – in contrast to that subjected to deliberate heat treatment (thermo- preparation) – is no longer usable. Yet another example has been described from Uddel- Uddelermeer, where a deposition of Late Mesolithic or Early Neolithic fint cores and blocks was found.171 In this case, the cores conform to the ‘rules’ of systematic punched- blade production, but the fint is of such bad quality that any fint knapper could have Fig 3.21 Possible ritual depositions of LBK quern stones anticipated failure. 169 Hogestijn & Peeters 2001; Peeters 2007. with traces of red ochre from Beek-Kerkeveld and Stein- The deposition by Mesolithic and Early 170 Groenendijk 2004. 171 Groenewoudt et al. 2006. Heidekampweg (from appendix II: PUBID 53, 528). Neolithic hunter-gatherers of ‘production waste’ 172 It should be repeated here that we use in small pits seems to somehow be connected the terms Mesolithic and Neolithic in a purely chronological sense. If one were with the process of fint tool production.172 The to use the terms in connection with Middle Neolithic deposition from Haren- particular behavioural (socio-economic, ideological) setings, the early Groenstraat can also be ft into such a context, Swiferbant phase could just as well be and could be seen as a continuity of a tradition termed the ‘ceramic Mesolithic’ (see e.g. Peeters 2010). that had started much earlier. The essence is 85 —

Fig 3.21 Continued.

maybe caught in the act of ‘extracting’ tools out From a broader perspective, the deposition of a block of raw material. Simple fakes, blades, of both pots and stone tools, such as complete unusable cores and chunky blocks are and fragmented axe blades and querns, are a materialisations thereof. They are the remains of common practice in Neolithic north-west a process of material transformation in which Europe. To what extent ritual deposition of deliberate ‘destruction’ – for example, of the objects can be understood as an expression of a burnt cores from Nieuwe Pekela 3 and maybe single belief system is difcult to say. also of the Mesolithic pierced mace heads Did deposition of deliberately – or even discussed in section 3.2.1 – may have played a accidentally – fragmented objects have the same part.173 However, complete objects are also meaning as deposition of complete objects? represented, for instance, at Beek-Kerkeveld and And how are depositional practices connected Lomm-Hoogwatergeul. Whether the deposition with places in the landscape? The recognition of of pots is part of the same ritual is difcult to (potential) ritual depositions in a setlement say. The possible deposition of a pot at Leur- context is important, as it extends the Hernense Meer, whilst rare, is not an exception. archaeological insights into the diversity of the Other examples from Swiferbant contexts come depositional landscape, which is, as yet, mostly 173 In this context, further reference can be made to the suggested deliberate from Hardinxveld-Giessendam De Bruin, Urk-E4 known from ‘of-site’ (non-setlement) contexts. destruction of ﬂint axes through and Bronneger.174 The large fragments of a pot As the majority of development-led projects are burning in southern Scandinavia (Larsson 2000) and the deliberate and baking plate at Knooppunt Hatemerbroek concerned with setlement contexts, there is an fragmentation of quern stones can maybe be interpreted as a construction important potential to be tapped. (Devriendt 2014). 174 Kroezenga et al. 1991; Peters & Peeters ofering, which is another seting again. 2001; Raemaekers 2001. 86 —

head reminds us of a Middle/Late Mesolithic pit 3.3.2 Death and ritual hearth at Knooppunt Hatemerbroek, which also yielded a portion of such a tool; a second, conjoining part was found outside the pit.178 The treatment of the dead ofers an explicit No cremated bone was found in the pit, and perspective on cosmological belief systems in unburnt bone would not have survived. The early prehistory. Development-led feldwork has same is true for a Late Mesolithic pit at yielded a considerable number of sites with Swiferbant-S22, which also contained a evidence for mortuary practices (table 3.6). fragment of a perforated mace head, and which Because these sites tend to draw the atention could also be fted to a fragment from the fnds of archaeologists, synthetic analysis has already layer.179 integrated most of the data over the past few Although in the case of both Knooppunt years, notably in two papers published in 2007.175 Hatemerbroek and Swiferbant-S22 it is However, more recent discoveries permit us to impossible to tell whether the pits are related to enlarge the body of data, and deepen insights. mortuary practice, it is important to consider the As to the Mesolithic, the dataset remains situation at the Mesolithic site of Mariënberg.180 limited. Afer the discovery of inhumations at Here, a group of six cylindrical pits containing a the sites of Hardinxveld-Giessendam Polderweg bright red, ochre-like compound were and De Bruin, evidence for mortuary practices interpreted as burial pits.181 Again, no bone was was found at two additional sites, Roterdam preserved, but particular soil paterns support Beverwaard, Dronten-N23. At Roterdam- the suggestion that bodies buried in a siting Beverwaard, three Middle Mesolithic pits with (fexed or crouched and upright) position were cremated human remains were discovered originally present. Apart from small numbers of (fg. 3.22).176 Hitherto, Mesolithic human fint artefacts (relatively many punched blades, ‘cremation’ in the Netherlands was only known which are of Late Mesolithic age), two burial pits from Oirschot V and, possibly, Dalfsen.177 yielded sandstone shaf polishers. It would seem Amongst the cremation remains at Roterdam- that objects like these, as well as perforated Beverwaard are also some fragments belonging mace heads, formed part of the mortuary ritual. to animals. Each pit also contained small Possibly, such objects were explicit individual amounts of fint artefacts and points made out belongings and bound to one’s personhood in of Wommersom quartzite. One pit yielded a life and in death. It cannot be excluded that broken yet complete perforated mace head and other pits containing such items, but not a slab (grindstone?), both made out of quarzitic containing remains of human bone, also sandstone. The presence of a perforated mace represent graves.

175 Louwe Kooijmans 2007; Raemaekers et al. 2007. 176 Appendix II: PUBID 463a; Niekus et al. Table 3.6 Sites with graves and loose skeletal remains. 2016. 177 Arts & Hoogland 1987; Verlinde 1974. Site Archaeological period Age (yrs) Gender Type PUBID Mesolithic cremations are a rare phenomenon in the Low Countries Beverwaard Middle Mesolithic 10-40 Unknown Cremation 463 altogether (see Niekus et al. 2016); other sites are Heﬃngen-Loschbour in 12-40 Female? Cremation Luxemburg and La Chaussée-Tirancourt in northern France (Ducroq, Bridault & 10-34 Unknown Cremation Munaut 1991; Toussaint et al. 2009). Dronten-N23 Late Mesolithic 35-45 Female Inhumation 538 178 The two parts together form a complete mace head. Schipluiden Hazendonk group 38-45 Male Inhumation 732 179 Devriendt 2014; Drenth & Niekus 2009; Price 1981. 59-65 Male Inhumation 180 Verlinde & Newell 2006, 2013. Louwe Kooijmans (2012) questioned several 46-49 Male Inhumation aspects – e.g. the nature of ‘ochre’ 41-50 Male Inhumation colouring (but see Huisman & Van Os 2013) – of the interpretations made by 25-40 Male Inhumation Verlinde and Newell (2006), but nonetheless concluded that the features 8 Infant Inhumation must be interpreted as burial pits. 181 Huisman & Van Os 2013. 2 Infant Inhumation 87 —

Fig 3.22 The Mesolithic grave at Roterdam-Beverwaard with the stone mace head (b) and slab (a), and the mace head (c) from Knooppunt Hatemerbroek (from appendix II: PUBID 259, 463). Object drawings scale 1:2.

Grave goods are, however, scarce in the developed into freshwater swamp and marsh. Dutch Mesolithic. Indeed, the Late Mesolithic It is in this context that the grave had been dug, inhumation graves at Hardinxveld-Giessendam possibly several centuries afer the last visit, Polderweg and De Bruin contained no grave when the location served other purposes. The goods at all, but there was some red ochre in exact timing is, however, difcult to determine, one burial at Polderweg. And absence of grave as no secure date can be given for the burial; it goods – at least those made of non-perishable may be of Late Mesolithic age, but could also be materials – is also noted at Dronten-N23, where of Early Neolithic (Swiferbant) age. a single inhumation grave was found on top of a The situation at Dronten-N23 seems to be river dune (fg. 3.23). The grave, which contained comparable to that at other river dune sites in the poorly preserved remains of a woman buried Flevoland. At Swiferbant-S21, S22 and S23, as in supine position, marks the fnal use of the well as at Urk-E4, burials with (poorly) preserved location, which had previously been visited by bone were found on top of the dunes.182 These many generations of Mesolithic hunter- dunes were used for dwelling and subsistence gatherers, over the course of three millennia (see purposes during the Mesolithic, but evidence for section 5.3.1). During this long time span, the occupational continuity in the earliest part of the environment became increasingly wet due to the Neolithic – the early phase of the Swiferbant structural rising of the groundwater table, and Culture – is lacking. The burials, however, seem 182 Raemaekers et al. 2007. 88 —

Fig 3.23 The inhumation grave at Dronten-N23 (from appendix II: PUBID 538). Top lef: excavation plan showing the burial pit that intersects several Mesolithic pit hearths. Top right: preserved molars in situ. Middle: overview of the burial. Botom: the burial pit in various stages of excavation. 89 —

to date to the classic phase of the Swiferbant graves discovered at Knooppunt Culture, post-dating the last Mesolithic activity Hatemerbroek.189 by several centuries. If the available dates refect We must bear in mind that our knowledge prehistoric reality to a reasonable extent, this about Mesolithic and Neolithic burial practices suggests the re-use of locations in a context may be heavily biased due to the research focus wherein mortuary practices played an explicit on setlement sites. The possibility of burial role. This could be seen as a means to outside setlements, or in locations other than historically connect people to the landscape, former setlement sites, must be kept in mind. whereby ancient setlement ground is used for Indeed, considering the fact that only limited burial of the dead.183 numbers of burials and ‘isolated’ human The fnds from Roterdam-Beverwaard and remains are known in the frst place, and that Dronten-N23 demonstrate how development-led there is considerable variability in mortuary research can make a valuable contribution to our practices within that relatively small ‘sample’, understanding of mortuary practices. Considering we can but wonder what we are missing. the small number of Mesolithic graves known Certainly there is a problem with regard to from the Netherlands, the variability in treatment diferences in preservation conditions, but this of the dead is considerable: inhumation and cannot account for all of the paterning observed cremation, extended and siting position, isolated thus far. From this perspective, it is important bones mixed with setlement waste, graves with that we archaeologists keep our eyes open, and and without grave goods, bowl-shaped and all the more so whenever we have an cylindrical pits, and graves with and without opportunity to investigate of-setlement ochre. In Swiferbant contexts, inhumation and contexts. burial in extended position of single and, sometimes, multiple individuals seems to be the rule, followed by the occurrence of isolated bones 3.4 Frames of reference mixed with setlement waste, an occasional skull burial, and a siting grave.184 This picture changes with the Middle Neolithic, when burial in fexed Although material culture can be considered to position in rectangular or oval pits is frequently have a mediating role in the context of socio- seen, in addition to burial in extended position cultural dynamics, material culture has and the occurrence of isolated bones mixed with traditionally received a lot of atention to setlement waste.185 Mostly, graves contain the characterise spatio-temporal paterns in the remains of a single individual. At Schipluiden, one archaeological record. This is also the case in the grave contained two individuals (males) buried in NOaA 1.0, which addresses questions concerning opposed orientation, one of whom seems to have the typo-chronology of artefact categories, fint sufered a violent death.186 tools and ceramics in particular (see appendix I). 183 Amkreutz 2013; Peeters 2007, 232. Although diferences in the treatment of the 184 Louwe Kooijmans 2007; Raemaekers dead are apparent from the archaeological et al. 2007. 185 Louwe Kooijmans 2007. record, it is also becoming clear that grave goods 3.4.1 Typo-chronology of the Final 186 Smits & Louwe Kooijmans 2006. remain rather scarce in the Dutch Middle 187 Smits & Louwe Kooijmans 2006; Van Palaeolithic and the Mesolithic Gijn et al. 2006. Neolithic. When they are present, they tend to 188 Van der Velde & Bouma 2016. consist of personal ornaments (beads and 189 Appendix II: PUBID 2059, 260. The Late Neolithic falls outside the scope of the pendants, mostly of amber, and sometimes jet) Whereas chronologies of the Palaeolithic and present study. or implements (e.g. strike-a-lights with a pyrite the Neolithic are usually characterised in terms 190 This is not to say that no problems exist with regard to the deﬁnition and nodule in grave 2 at Schipluiden).187 of successions of ‘well-defned’ cultures or chronological anchoring of these Although amber preserves relatively well, these technological traditions,190 the Mesolithic has ‘cultures’ and ‘traditions’. In many cases, these ‘cultures’ and ‘traditions’ remain rather exceptional. A recently excavated always been somewhat problematic, not only have inherited from culture-historical TRB cemetery near Dalfsen (province of because chronologies are traditionally based on classiﬁcations the assumption of spatial and chronological interrelationships, Overijssel) has, however, produced a grave with microliths – small parts of composite tools – which ofen lack underpinning with a complete necklace consisting of tens of amber with limited stylistic resolution (fg. 3.24), but reliable context and dating. The ‘clear’ picture that the Palaeolithic and beads.188 The use of amber for personal also because of problems of independent Neolithic successions seemingly exhibit ornamentation continues in the Late Neolithic, chronological anchoring of fint tool is, without doubt, more complicated and problematic than is thought by notably in the Corded Ware and Bell Beaker assemblages. The later problems are related to many. 90 —

Fig 3.24 Overview of the most important typochronological tool categories for the Late Palaeolithic through Middle Neolithic (adapted from Amkreutz, Verhart & Van Gijn 2016; Brounen 2016; Deeben & Niekus 2016a, 2016b, 2016c; De Grooth 2016; Houkes 2016; Niekus, Kramer & Deeben 2016; Peeters 2016; Peeters & Devriendt 2016; Rensink & Niekus 2016; Schreurs 2016). Scale bars in centimetres. Hamburgian (top) and Creswellian (botom). 91 —

Fig 3.24 Continued. Federmesser Gruppen (top) and Ahrensburgian (botom). 92 —

Fig 3.24 Continued. Mesolithic. 93 —

Fig 3.24 Continued. LBK (top), Rössen (middle) and Michelsberg (botom). 94 —

Fig 3.24 Continued. Swiferbant. 95 —

Fig 3.24 Continued. Hazendonk (top), Vlaardingen (middle lef), Stein (middle right), and TRB (botom). 96 —

the problem of palimpsest formation and the contemporaneity. At Meerstad, 14C-dates on assumed, but unestablished, chronological wood charcoal and hazelnut shell sampled from association of lithic assemblages and dated the same pit hearth gave slightly divergent yet features (mostly pit hearths).191 Especially the close results, which both seem to match the frequently used typo-chronology developed by expected age of the lithic assemblage.199 In view Newell sufers from this problem,192 although of the fact that charred hazelnut shell is other classifcations share similar generally associated with surface hearths and shortcomings.193 The use of diferent typo- fint scaters,200 it is likely that the date on the chronologies potentially leads to difculties in hazelnut shell provides a reliable age for the comparing reported research results. In the lithic assemblage of Meerstad. Pit hearths northern Netherlands, reference is particularly represent other behavioural contexts, which made to Newell’s scheme and to additional may not be related to that of surface hearths considerations by others regarding its viability.194 and lithic scaters. In the southern Netherlands, reference is made A fnal problem that needs to be mentioned to Arts and an adapted scheme published by here concerns the defnition of types of Newell, as well as Belgian classifcations.195 microliths, and the reliability of type assignment One way to solve problems of chronological by analysts. Indeed, it is not always generally resolution of Mesolithic typology is to focus on clear, for instance, what constitutes the diference fnds occurrences that resulted from single, between a lancet point and D point, what defnes short-lived activity phases, comparable to those a needle-shaped point, and where to draw the excavated in Verrebroek-Dok (Belgium).196 This line between C points and triangular points. This will lead not only to more fne-grained ambiguity leads to noise, or fuzziness, in chronologies, but also to deeper insight into classifcations and increases the risk of erroneous (sub-)regional variability. In the context of typological assignments. At the same time, development-led projects, a restricted number however, we must bear in mind that any of such chronologically discrete assemblages archaeological typology is intrinsically fuzzy, as have been excavated, for instance, at Zutphen- there is no such thing as fully mutually exclusive Ooijerhoek, Dronten-N23, Ede-Kernhem and types within a class of closely related items. Afer Well-Aijen. Some other assemblages have been all, it is the observer – the archaeologist – who recorded more recently, at Soest, Sumar and decides which characteristics need to be taken in Kampen-Reevediep, but these have not yet been account, and what degree of overlap between published and therefore fall outside the scope of types is acceptable. Hence, the concern should, the present study. Despite the availability of frst of all, be to establish to what extent the such valuable assemblages, the full potential defned types have meaning, be it socio-cultural cannot always be exploited, mostly due to a lack or chronological. 191 Lanting & Van der Plicht 1998. 192 Newell 1973. of 14C-dates. At Hoogezand-De Vosholen, for Based on previous insights and new 193 Arts 1988. See Peeters and Niekus (2005) instance, datable charred hazelnut shell and information derived from development-led and Verhart and Groenendijk (2005) for a discussion. charred bone were found, but these have not projects, a more variegated typo-chronological 194 Lanting & Van der Plicht 1998; Newell been dated; in this case several clusters of lithics framework is gradually taking form (table 3.7). 1973; Peeters & Niekus 2005; Verhart & Groenendijk 2005. have been dated on the basis of microlith This framework is built not only on sites from 195 Newell 1975; Arts 1988; Crombé, Van typology and the tree taxa (which are assumed the Netherlands, but also on key sites in Strydonck & Boudin, 2009. 196 Crombé, Van Strydonck & Boudin, 2009. to have chronological meaning) represented in Germany and Belgium. The typo-chronological 197 Appendix II: PUBID 737. the wood charcoal.197 In other cases assemblages framework is, however, not one of diachronic 198 Appendix II: PUBID 524, 617. 199 Arnoldussen et al. 2012. have been dated on the basis of radiocarbon succession, but, rather, an increasingly complex 200 Compare Crombé, Groenendijk & Van dates on charcoal from pit hearths,198 but here patern of spatio-temporal tendencies. Strydonck 1999 ;Crombé et al. 2013; Sergant et al. 2006. too, as above, is the problem of assumed 97 —

Final Palaeolithic to Mesolithic transition Early Mesolithic For the Ahrensburgian, which is normally Within the Belgian Early Mesolithic, a number of considered to represent the fnal Palaeolithic ‘groups’ have been defned on the basis of the tradition, it is clear that tanged points were relative importance of microlith types;209 these gradually replaced by B points and trapezoidal typological groups may also be of relevance for points. The later are in fact B points with basal the Netherlands. The so-called Neerharen retouch, and they are sometimes (erroneously) Group, dated c. 8550-8100 cal. BC, is dominated called Zonhoven points. Although tanged points by points without a retouched base (A and B are more frequent in the early Ahrensburgian, points, >50%), followed by isosceles and scalene they do not necessarily have to be present in triangles, and smaller numbers of crescents and sites belonging to this stage, e.g. as a result of other points. The Verrebroek Group, dated c. diferences in site function. The late 8350-8000 cal. BC, is dominated by points Ahrensburgian sporadically comprises other without a retouched base (A and B points), types, such as Tjonger points or Federmesser followed by scalene triangles, and smaller and A points.201 numbers of crescents and points with a The ‘transitional’ phase from Ahrensburgian retouched base (C points). Finally, the Ourlaine to Early Mesolithic, between c. 9250-8400 cal. Group – although mostly present in Belgium and BC, is rather obscure.202 One of the few sites northern France – dated between c. 8500-7550 known to date, Zwolle-Oude cal. BC, is characterised by crescents (35-45%) in Deventerstraatweg, has a date on calcined bone combination with points without a retouched of c. 9000 cal. BC.203 The assemblage is base; other microliths, such as points with a incomplete due to partial destruction of the site, retouched base and triangular points, occur but comprises several Late Palaeolithic incidentally. elements, such as large burins and scrapers, as As yet, the Neerharen Group is potentially well as fragmentary backed points (Tjongerian), only represented in the Netherlands at a tanged or trapezoidal point, and a B point. Posterholt.210 Cluster 5 at Dronten-N23 has been Several blades show splintered bulbs of dated at c. 8350 cal. BC and has been atributed percussion (esquillements de bulbe), a feature that to the Verrebroek Group on the basis of the is ofen observed in Ahrensburgian assemblages microlith spectrum, although the presence of and possibly indicates the use of sof stone both some isosceles and scalene triangles would hammers.204 Securely dated lithic assemblages permit an atribution to the Neerharen Group.211 from the period 9000-8400 cal. BC are lacking, The assemblage of Haelen-Broekweg, with with the exception of a small assemblage dates between c. 8350-8250 cal. BC, compares – mostly debitage – recovered from alluvial well with the Dronten-N23 cluster.212 Several 201 Niekus et al. in prep. deposits at Zutphen-Ooijerhoekse Laak.205 One clusters at Zutphen-Ooijerhoek, with dates 202 Also see Crombé, Deeben & Van cluster of lithics at Knooppunt Hatemerbroek between c. 8350-8000 cal. BC, have been Strydonck 2014. 203 Niekus et al. in prep. most probably dates between c. 8900 and atributed to the Verrebroek Group.213 The 204 Compare Weber (2012), who described 8400 cal. BC based on the assemblage of assemblage of Slochteren-Hooilandspolder falls similar features for the Hamburgian. 205 Groenewoudt et al. 2001. However, this points.206 It is composed of B points, elongated within the same age range,214 but compares research falls outside the scope of the trapeze or trapezoidal points, one isosceles and beter with Duvensee because of a clear present study. 206 Cluster 6.12 (Niekus et al. 2012) has been one scalene triangle, and one crescent. The presence of elongated trapeze points. partly excavated and has not been dated assemblage shows clear resemblance with directly. 207 Bokelmann 1971, 1991; Holst 2007; Duvensee 2 (c. 8600 cal. BC) and Haverbeck (c. Late Early Mesolithic Tolksdorf et al. 2009. 8950-8700 cal. BC) in Germany.207 It is possible Slightly younger are clusters 7a, 7b and 9 at 208 Kortekaas & Niekus 1994; Nieuwe Pekela is largely unpublished (Groenendijk that elongated trapeze points are characteristic Ede-Kernhem, with dates of c. 7850-7700 cal. 2004). of the Early Mesolithic in the northern BC; the majority of microliths are triangular 209 Crombé Van Strydonck & Boudin 2009. 210 This site has been investigated within an Netherlands; other sites where these have been points (>70%) with additional points with and academic context (Verhart 2000). found are Slochteren-Hooilandspolder (c. 8100 without retouched bases, as well as crescents.215 211 Niekus et al. 2012. 212 Bats et al. 2010. cal. BC) and Nieuwe Pekela 9 (c. 7850 cal. BC). Hence, the Ede-Kernhem assemblages difer 213 The site was excavated by the Rijksdienst They are lacking in the lithic assemblages of from both the Neerharen Group and the voor het Oudheidkundig Bodemonderzoek (now Rijksdienst voor Zutphen (c. 8300-8000 cal. BC) and Ede- Verrebroek Group. Although the dominance of het Cultureel Erfgoed) in 1999, but it has Kernhem (c. 7850-7700 cal. BC) farther south.208 triangular points corresponds to the not yet been published. 214 Appendix II: PUBID 773. Possibly this indicates regional diferentiation. characteristics of the Chinru Group, which is 215 Devriendt 2015. 98 —

Table 3.7 Typochronological table for the Late Palaeolithic to Neolithic (from Amkreutz et al. 2016). Magdalenian Early Hamburgian Late Hamburgian (Havelte phase) Creswellian Federmesser-Gruppen Early Ahrensburgian Late or Epi Ahrensburgian

Tjonger point / Federmesser / curved backed piece - - + + ++++ + + Gravete point - - + - ++ + - Kremser point --- + - - Azilian point ----++- Creswell point - - + ++++ + + - Cheddar point - - - + ? - - Penknife point ----+-- Malaurie point ----+--

Ahrensburgian (tanged) point -----++++ + Shouldered point - ++++ +-+-- Havelte (tanged) point - + ++++ ----

A point / Unilaterally retouched point / point w. unretouched base ----++++ B point / Obliquely truncated point / point w. unretouched base ----+++++ ++++ Long B point - - - + + - -

Small asymmetric trapeze ------Broad asymmetric trapeze ------Rhombic trapeze -- -+-++ Elongated trapeze -----+++ (Asymmetric) right-angled trapeze ----+++

Isosceles triangle -----++ Scalene triangle -----+-

Backed bladelet ----++++ ++ - Unilaterally retouched backed bladelet +++ - - - Bilaterally retouched backed bladelet + - - - Triangular backed bladelet ----

Lacan burin + - - - ? - - Borer / Zinke -++++++----

Legend ++++ frequent +++ fairly common ++ present + rare - absent unknown ? unclear 99 —

Table 3.7 Continued. Early Mesolithic southern Netherlands Early Mesolithic northern Netherlands Middle Mesolithic southern Netherlands Middle Mesolithic northern Netherlands Late Mesolithic southern Netherlands Late Mesolithic northern Netherlands

Tjonger point / Federmesser / curved backed piece +--

A point / Unilaterally retouched point / point w. unretouched base +++ - ++ + ++ ++

B point / Obliquely truncated point / point w. unretouched base ++++ ++++ ++ + +++ +

C point / Tardenois point / point w. retouched base +++-++++++

D point / Bilaterally retouched point / point w. unretouched base ++ - ++ ++ + +

Crescent +++ ++ - + + -

Isosceles triangle +++ ++ ++ +++ + +

Scalene triangle +++ ++ +++ ++++ ++ ++

Lancete point - +---

Needle-shaped point ++

Symmetric trapeze +-- +++

Asymmetric trapeze + ++++ - ++++ +++

Rhombic trapeze - - +++ +

“Mistletoe leaf” point + - +++ + ++ -

Invasively retouched triangle - +++ - + -

Invasively retouched point w. rounded base + - +++ - ++ +

Broken point w. surface retouch - +++ + + -

Transverse point --++

Sauveterre point +?

Backed bladelet +++ - ++++ ++++ +++ ++++

“Montbani-style” blade / denticulated blade ++

(Perforated) macehead / Geröllkeule ++++++

Legend ++++ frequent +++ fairly common ++ present + rare - absent unknown ? unclear

N.B. Late Mesolithic asymmetric trapezes in the southern Netherlands include LBK points (exchange) 100 —

Table 3.7 Continued. Linear Potery culture Linear Potery Rössen culture Swiferbant culture Michelsberg culture Hazendonk group culture Funnelbeaker Vlaardingen culture Stein group / Corded Ware Single Grave culture culture (Bell) Beaker

Small symmetric trapeze + Small asymmetric trapeze + Broad symmetric trapeze + Broad asymmetric trapeze + Elongated trapeze + (Asymmetric) right-angled trapeze + Asymmetric trapeze w. hollow base and invasive retouch or retouche inverse plate (RIP) (= LBK point) ++ Asymmetric trapeze w. diferent kind of base + Transverse point + + ++ + +++ +++ + Triangular (equilateral) arrowhead with edge retouch ? + Triangular (equilateral) arrowhead with semi-covering (surface) retouch + + ? ++ +++ Triangular (equilateral) arrowhead with covering (surface) retouch + + ? +++ Asymmetric triangle w. concave base and invasive retouch or retouche inverse plate (RIP) (= LBK point) +++ Asymmetric triangle w. diferent kind of base +++ + Triangular (isosceles) arrowhead w. straight base and edge retouch ++++ + ++ + Triangular (isosceles) arrowhead w. straight base and semi-covering (surface) retouch +++ + ++++ ++++ +++ Triangular (isosceles) arrowhead w. straight base and covering (surface) retouch + + + +++ Triangular (isosceles) arrowhead w. rounded base and edge retouch ++ + Triangular (isosceles) arrowhead w. rounded base and semi-covering (surface) retouch + ++ ++++ +++ Triangular (isosceles) arrowhead w. rounded base and covering (surface) retouch + ? + +++ Triangular (isosceles) arrowhead w. concave base and edge retouch + + Triangular (isosceles) arrowhead w. concave base and semi-covering (surface) retouch +++ ++ ++ Triangular (isosceles) arrowhead w. concave base and covering (surface) retouch + +++ Leaf-shaped arrowhead w. edge retouch ++++ Leaf-shaped arrowhead w. semi-covering (surface) retouch ++ ? + Leaf-shaped arrowhead w. covering (surface) retouch ? + Leaf- or droplet-shaped arrowhead w. covering (surface) retouch + ? Tanged arrowhead w. surface retouch +++ Barbed arrowhead w. a longer tang +++ + Barbed arrowhead w. a tang ++ +++ Lozenge-shaped arrowhead w. covering (surface) retouch + Core axe ++ ? Flake axe + + ++ ++ Stone axe w. round cross-section + + + + Stone axe w. rectangular cross-section ? ? + +++ ++ Stone axe w. oval cross-section + + + +++ + + Stone palstave +? Flint axe w. rectangular cross-section + +++ ++ Flint axe w. oval cross-section + + ++++ ++++ +++ +++ +++ ++ Unpolished axe +++ Adze +++ +++ Plano-convex perforated adze + + Perforated shoe-last celt + +++ 101 —

Perforated broad wedge +++ + Knob-buted batle axe ++ + Batle axe types A, B, C, P1, P2, D, E, H, R/S +++ Batle axe types B/A ++ Batle axe types F, G + Faceted batle axe type 1 +++ Faceted batle axe types 2a, 2b ++ Batle axe type Emmen +++ Batle axe type Zuidvelde ++ Batle axe type Epe + Batle axe type K + Heavy duty axe + Disc-shaped perforated macehead + ? (Perforated) macehead / Geröllkeule + Chisel w. oval cross-section (cigar-shaped) ++ ++ Chisel w. rectangular cross-section ++ Chisel on blade +++ Pick +? Notched hoe + +++ Pointed retouched blade +++ ++ + + Grand-Pressigny dagger +++ Pseudo Grand-Pressigny dagger +++ Scandinavian-type fint dagger +++ Sickle-blade (w. gloss) +++ + ? Bifacially retouched sickle ? Sickle type “Schipluiden/Ypenburg” ++ Planoconvex knife / knife w. surface retouch + +++ Quartier d’orange + Flint w. rounded end(s) / strike-a-light + Short blade scraper w. straight scraper edge ++++ +++ Flake scraper > 5 cm +++ Blade scraper > 8 cm +++ Tanged scraper ++ Blade > 8 cm +++ Blade > 6 cm +++

Legend ++++ frequent +++ fairly common ++ present + rare ? unclear

Legend Swiferbant and Funnelbeaker + present absent 102 —

dated between c. 8000-7500 cal. BC, the low become dominant in assemblages until the number of C points contradicts such an appearance of Late Mesolithic trapezoid points atribution. The presence of lancet points in the (c. 6450-6400 cal. BC). There does not seem to Ede-Kernhem 7b and 9 assemblages have been a marked break with the late Early corresponds beter to the assemblages of Mesolithic. In contrast, in the southern and Duvensee 6a (c. 8300 cal. BC BP), Duvensee 6b central Netherlands, the appearance of points (c. 8000 cal. BC) and possibly Duvensee 13 (c. with invasive surface retouch – triangular, but 7650 cal. BC).216 The Duvensee assemblages, mostly feuille de gui (mistletoe-leaf-shaped) and however, also contain trapezoidal and needle leaf points – marks the start of the middle phase (Sauveterre) points, which are absent in the of the Mesolithic, c. 7500-7400 cal. BC. Examples Ede-Kernhem assemblages. The later, come from Roterdam-Beverwaard, Epse-Olthof therefore, partially share characteristics of both and Ede-Kernhem;221 the most northern the Belgian and the Duvensee sites. occurrence is Dronten-N23 Cluster 2, with a date Clusters 3, 9 and 10 from Dronten-N23,217 between c. 6400-6350 cal. BC.222 The contrast dated c. 7650-7550 cal. BC, are slightly younger between the northern and south-central regions than the assemblages from Ede-Kernhem. probably refects diferences in socio-cultural About 50% of the microliths consists of development in these two regions. Whether the triangular points, followed by points without relative importance of points with surface retouched base (30%) and points with a retouch in particular has any signifcance in this retouched base, crescents, lancet and needle respect is not clear. The distinction between the points (for a total of 20%). To some extent, the so-called Sonnisse Heide Group and assemblage composition compares with the Gelderhorsten Group in Belgium is based on the Verrebroek Group, which is, however, older. relative frequency of such points in relation to The Dronten-N23 assemblages also show some dominant quantities of backed bladelets.223 In similarities with the Chinru Group in terms of the Netherlands, again, assemblages do not dating and the importance of triangular points, easily conform to these group descriptions, yet they difer in the low number of points with although one can see similarities. In the case of retouched bases. As yet, the Chinru Group seems Haaksbergen-Hassinkbrink, dated c. 7100 cal. to be regionally connected with eastern France BC, the microlith spectrum is composed of and (south-) western Germany. points with and without retouched base, lancet The clusters from Ede-Kernhem and points, triangular points and leaf points with Dronten-N23 are quite comparable in terms of surface retouch, as well as a trapezoidal point, a microlith spectrum.218 The relative importance of possible needle point, and a Svaerdborg point.224 lancet points at Dronten-N23 Cluster 3 possibly There are few backed bladelets. Interestingly, indicates a stronger northern afnity (i.e. apart from the leaf points, the other points also Duvensee). A second possibility, which does not show rather ‘fat’ retouch. Does this betray a exclude the frst, is that lancet points, as well as ‘southern’ afnity in an otherwise ‘northern’ needle points, belong to the relatively younger assemblage? On the other hand, backed phase. However, trapezoidal points are lacking bladelets are dominant in the assemblages from in the Dutch assemblages, in contrast to Groningen-Meerstad; Scheemderzwaag- Duvensee, where these are nonetheless Scheemda; and Dronten-N23 clusters 2, 3, and relatively sparse. The assemblages of Meerstad 5-9.225 Broadly speaking, backed bladelets seem 216 Holst 2007. (c. 7550 cal. BC) and possibly Ede-Kernhem to be a common element in the Dutch Middle 217 Niekus et al. 2012. Cluster 3 (c. 7400 cal. BC) have yielded relatively Mesolithic, without any geographical distinction. 218 Devriendt 2015; Niekus et al. 2012. 219 Appendix II: PUBID 255; Devriendt 2015. few points, but other than that, they ft the other 220 Lanting & Van der Plicht 1998; Niekus & assemblages well.219 Comparable assemblages Late Mesolithic Smit 2006; Niekus et al. 2012. 221 Appendix II: PUBID 463, 733, 758. were retrieved from Rotsterhaule, Warns and The Late Mesolithic phase is characterised by the 222 Niekus et al. 2012. Recent excavations at Almere-Zwaanpad, with dates between c. 7100- presence of trapezoid points made on regular Soest-Staringlaan (by BAAC) and Kampen-Reevediep (by ADC) have 6850 cal. BC.220 blades. These points have abrupt distal and yielded important assemblages proximal retouch and vary in shape, forming containing points with surface retouch. 223 Crombé 1999. Middle Mesolithic symmetric (oblique distal and proximal ends) 224 Appendix II: PUBID 206. In the northern Netherlands, triangle points and asymmetric (one straight-angled end, 225 Appendix II: PUBID 255, 475; Niekus et al. 2012. (mostly scalene, and sometimes very small) sometimes referred to as straight-angled) 103 —

trapezes, as well as rhombic trapezes. In of distinct traditions in the northern and reference to production direction of the blade, southern Netherlands. Afer c. 7850 cal. BC, the both long and narrow and short and broad picture becomes more blurred: mostly we seem specimens are found, whereby the later are to be dealing with assemblages with ‘mixed’ considered to be the more recent form.226 characteristics. In the north, defning elements Assuming that this chronological shif is ‘real’, for a Middle Mesolithic stage are lacking or it does, however, only represent a general trend uncertain (e.g. lancet and needle points, backed that continues well into the Neolithic. Other bladelets); we note a continuity of development types of microliths continue to be used in in assemblages with triangular points, until the parallel with trapezoid points. As far as can be appearance of trapezoid points. In the southern seen from the few discrete clusters of lithics and central Netherlands, points with invasive excavated, it has been suggested that trapezoid surface retouch defne the Middle Mesolithic points appear c. 6400 cal. BC in the southern and can be linked to a clear shif in tradition. Netherlands,227 and somewhat earlier in the Between c. 7700-6950 cal. BC, backed bladelets northern part of the country.228 Several clusters are well represented in assemblages anywhere at Dronten-N23 indicate the appearance of in the Netherlands. A further shif involved the trapeze points between c. 6450-6400 cal. BC.229 appearance of trapezoid points and punched Trapezoid points dominate the assemblage of blade technology, which characterises the Late Cluster 3, whereas Cluster 2 shows a broader Mesolithic (and Early Neolithic). As a fnal spectrum of microliths. It is assumed that Cluster remark, we note that it is important to bear in 3 is younger than Cluster 2, which has a date mind that the further development of a typo- close to 6400 cal. BC. It is important to note that chronological framework can beneft from a the Late Mesolithic is not simply characterised more systematic study of technology. If by trapezoid points. An important feature is the geographically distinct ‘microlith traditions’ production of regular and relatively long blades existed during (part of) the Mesolithic, by means of indirect percussion (punch ‘technological traditions’ also may have been technique) as blanks for trapezoid points, but diferent. Indeed, such diferences seem to be also Montbani notched blades. This present in the production of bladelets technological emphasis is one that is found all (microblades).232 over Europe, with the exception of the British From the above, it follows that the typo- Isles.230 This tradition of blade production and chronological framework for the Final emphasis on trapezoid points, which explicitly Palaeolithic and Mesolithic still comes with difers from the Middle and Early Mesolithic many problems. The results from development- traditions, continues well into the Neolithic. led projects have certainly shed light on some Indeed, it is widely represented in the geographical and chronological trends, but, Swiferbant contexts.231 nonetheless, there is lack of discrete and well- dated clusters of lithics that can serve as solid Concluding remarks chronological anchoring points. As a result, the In summary, the period between the Late overall picture remains complex. It is, however, Ahrensburgian and c. 8400 cal. BC is poorly clear that the group defnitions of assemblages represented in general. A handful of as developed in Belgium are only partly (incomplete) assemblages and few dates are, applicable in reference to particular phases and 226 Niekus 2009. as yet, all we have. For the Early Mesolithic post- regions. In fact, this is also the case for the 227 Robinson et al. 2013. dating c. 8400 cal. BC, more assemblages and reference to German assemblages. This suggests 228 Verhart & Groenendijk 2005; also see Deeben & Niekus (2016c) for a dates are available from development-led that a nation-wide or broad regionally valid discussion. projects for the period under review. To some typo-chronological framework will be difcult to 229 Niekus et al. 2012. 230 Tolan-Smith 2008. extent, the assemblages are comparable to maintain, and that the defnition of assemblages 231 Peeters et al. 2001; Devriendt 2014; groups defned in Belgium and Germany. with varying characteristics yet partly Peeters & Devriendt 2016. Recent evidence from the eastern Baltic Assemblages from the northern part of the overlapping chronology will be difcult to avoid. suggests the presence of regular blade Netherlands show more afnity with the The question is, however, to what extent such and microblade technologies in the 9th-8th Millennium BC (Hartz et al. 2010; German (Duvensee) groups, whereas those from typological division refects diferences in Sørensen et al. 2013). the southern Netherlands connect beter to the functional and technological traditions. This 232 This aspect is currently being investigated in greater detail at the Belgian groups. This suggests the development question harks back to the discussions about the Groningen Institute of Archaeology. 104 —

interpretation of (dis)similarities in stone tool not represented in our sample of reports. assemblages in the culture-historical and Apparently, the potery-based typo- processual paradigms in the 1970s. In order to chronological subdivision of the Dutch Neolithic obtain beter insight into the socio-cultural does not suit the ‘on the ground’ situation, meaning of spatio-temporal variability in lithic which makes it difcult to be certain about true assemblages, it remains important to keep an absence (or presence) of cultural groups at a eye out for discrete low-density clusters, and to local or (sub-)regional scale. adapt feld methods, project outlines and the Whilst it is clear that ceramic analysis at the involvement of specialists accordingly. site level may provide input for typo- chronological analysis and the study of regional variation, it is equally clear that development- 3.4.2 Chronological and regional trends led research ofen lacks specifed research in potery production and use questions or a research budget to address these issues. Furthermore, re-analysis of assemblages such as those listed in table 3.8 is seriously In the Netherlands, early prehistoric potery hampered by the unavailability of datasets in the assemblages are by defnition confned to the national digital repository DANS-EASY: only Neolithic, following the chronological framework seven relatively small datasets appear to have presented in chapter 1,233 but one could just as well been uploaded (also see section 2.4).235 argue for a ‘ceramic phase’ of the Mesolithic when considering the wider economic and technological Developments in typo-chronology context of, for instance, the early Swiferbant Two projects in Limburg provided ‘Neolithic’ Culture.234 But in our view, whether one considers potery that could be used to build a new regional it a Mesolithic or a Neolithic element is not that chronology, namely, Itervoort and Well-Aijen. important; for this reason we use the terms in a The reports indicate that it is difcult to interpret purely chronological sense (see chapter 1). Of these assemblages using the current typo- greater signifcance here is the fact that ceramics chronological scheme. Atempts to build a new, play a central role as frames of reference for more regional scheme have not been made. Neolithic chronology and for the cultural Moreover, whilst the Itervoort assemblage of 114 assignment of assemblages. Potery has become sherds comes from a feature that also included the structuring material category for the analysis wood charcoal suitable for 14C-analysis, no date is of cultural developments (see fg. 2.11), and as available. Hence, despite its potential to build a such it plays a signifcant – but maybe somewhat new regional typo-chronology, this assemblage underestimated – role in heritage management is, as yet, of no use. The Well-Aijen assemblage decision making. Signifcance assessment of newly originates from test trenches236 and involves a discovered sites does take into account (potential) large feature that yielded 204 fragments of cultural assignment, for instance, to determine potery, alongside stone and fint artefacts. rarity. But of course there is more to ceramics. A 14C-date situates the assemblage around 4300 As yet, relatively litle is known about regional cal. BC.237 The potery sherds bear characteristics variability, fabric and sourcing of compounds (clay, that are reminiscent of Rössen ceramics, notably temper), or function. the presence of perforated lugs (fg. 3.25). There Research topics like these are also echoed in are, however, also diferences, like the absence of the NOaA 1.0; so to what extent has Furchenstich and Doppelstich decoration.238 We development-led work made a contribution to consider it desirable that the fnal report, which increasing our knowledge about ceramics? incorporates results from the test trenches and Table 3.8 lists sites that have yielded early subsequent excavation, contain a full analysis, prehistoric potery remains. The reports make including a wide geographical comparison that 233 Van de Broeke, Fokkens & Van Gijn clear that assemblage size is highly variable, should include contemporary Swiferbant sites, 2005. 234 Peeters et al. in prep. ranging from over 10,000 fragments at especially Schokland-P14. The excavator has 235 Appendix II: PUBID 74, 122, 340, 373, Hazerswoude-Rijndijk to a single fragment at made convincing arguments for a model that 407, 463, 744. 236 The results of the subsequent Emmeloord-Ens. Also, cultural groups that have proposes that Swiferbant potery morphology excavations are not yet available. largely escaped our atention in the twentieth was inspired by the Danubian tradition, but that 237 Appendix II: PUBID 725b. 238 Ten Anscher 2012, 134-142. century – notably Rössen and Michelsberg – are the typical Swiferbant decorative schemes and 105 —

Table 3.8 Overview of sites with early prehistoric potery and number of potery sherds.

PUBID Project Culture N

53 Beek Kerkeveld LBK 3089

194 Elsloo LBK 3100

528 Rijksweg A2 Stein-Heidekampweg LBK 2704

537 Sweikhuizen Geverik LBK 310

651 Beek Bourgognestraat LBK 43

340 Leur Hernense Meer Swiferbant 115

463 Roterdam Beverwaard Tramremise Swiferbant 40

468 Roterdam Groenhagen Tuinhoven De Zwanen Rietpark Swiferbant 415

739 Groenenhagen Tuinenhoven Swiferbant 212

61 Bergerden Geranium Hazendonk 30

338 Nijmegen Oosterhout t Klumke Hazendonk 186

373 Meteren de Plantage Hazendonk 184

732 Schipluiden Hazendonk c. 8000

542 Itervoort Thorn Neolithicum 114

725 Well Aijen werkvak 2 Neolithicum 580

74 Beuningen Vlaardingen-Stein 65

122 Den Haag Steynhof Vlaardingen-Stein 7

219 Sitard Geleen Hof van Limburg Vlaardingen-Stein 528

253 Haren Groenstraat (n BR) Vlaardingen-Stein 219

259 Hanzelijn Hatemerbroek Vlaardingen-Stein ?

272 Hellevoetsluis Ossenhoek Vlaardingen-Stein c. 2260

463 Roterdam Beverwaard Tramremise Vlaardingen-Stein 22

600 Wijchen Oosterweg Vlaardingen-Stein 2200

740 Hazerswoude Rijndijk Vlaardingen-Stein 10658

240 Groningen UMCG TRB 613

253 Haren Groenstraat (n BR) TRB 5

259 Hanzelijn Hatemerbroek TRB 275

260 Hatemerbroek Bedrijventerrein TRB 557

272 Hellevoetsluis Ossenhoek TRB c20

407 Nijverdal de Regge TRB 24

740 Hazerswoude Rijndijk TRB 7

741 Hempens Wâldwei TRB 4

744 Emmeloord Ens TRB 1

750 Groningen Helpermaar TRB 136

752 Groningen Europapark TRB 10

740 Hazerswoude Rijndijk Corded Ware 207 106 —

techniques were excluded from this exchange of covered with clay, resulting in a stratifed site that ideas.239 It is possible that the Well-Aijen yielded ceramics from three cultural groups: assemblage also fts this model. Vlaardingen/Stein, Funnel Beaker and Corded Whilst Itervoort and Well-Aijen provide new Ware. The analysis takes the assignment to these material for potentially developing a regional cultural groups as a given, and concludes that the typo-chronology, a stratifed site may be used for Vlaardingen-style potery was slowly replaced by similar purposes at the local level. The sites under Corded Ware–style potery, an outcome that is review provide one such example: Hazerswoude- interpreted as evidence of a gradual change in 239 Ten Anscher 2012. 240 Appendix II: PUBID 740. Rijndijk.240 This river levee site was gradually behaviour of the inhabitants. Beckerman’s

Fig 3.25 Potery from Well-Aijen: perforated lugs and knobs (lef), decoration (middle), rims (right). The assemblage is of Early Neolithic age (from appendix II: PUBID 725b). 107 —

241 Beckerman 2015. Recent analysis of ancient DNA sampled from human bone remains assocated with Bell Beaker material culture across Europe shows that socio-cultural interaction is complex in terms of expressions of cultural publication on the Corded Ware setlements in chronological ‘frontier’ between the two is a identity and demography (Olalde et al. the province of Noord-Holland provides a new mater of taste; in terms of both morphology and 2017). It warns us against drawing overly simplistic conclusions with regard to the typo-chronological scheme in which the technology, there is no clear break. As Beckerman spread and societal integration of Vlaardingen–Corded Ware transition is proposed notes, this leaves decoration in Corded Ware style material innovations; the understanding of geographical paterns of similarity and to have been one of cultural continuity.241 She as the single distinctive character, and this by variability requires a multidimensional proposes that both groups share a cultural defnition makes it impossible to have an interpretive framework that integrates diﬀerent disciplines. continuum of expressions in which the factual undecorated Corded Ware vessel (fg. 3.26).242 242 Beckerman 2015, 241-243.

Amount of stone grit tempering decreases

Average wall thickness decreases Technology

Amount of grog tempering increases

b c e beaker

Proportion of beakers increases Morphology

ABC DE

Medium to large vessels become more elongated

Perforations under the rim and the wall, knobs decreases

Cord and spatula decoration Decoration Number of All Over Ornamented vessel increases

Early Corded Ware Late Corded (group 1) (group 2) Early Vlaardingen Middle Vlaardingen Late Vlaardingen

Fig 3.26 Chronological developments: Vlaardingen and Corded Ware in the coastal zone (from Beckerman 2015). 108 —

The 14C-dates from Hazerswoude-Rijnsijk do not from the traditional reproduction of potery by permit any distinction either.243 It would have several generations of poters during the been interesting to test this hypothesis using the colonisation phase of the coastal area. The relatively large Hazerswoude-Rijndijk assemblage dissimilarities amongst the three microtraditions and to study the correlation between technology, may then be explained as the result of a morphology and decoration, were it not that the ‘ceramolocal’ marriage system and a household dataset is not available in DANS-EASY. mode of production (fg. 3.27).245

Regional variation The nature and origin of resources used in Development-led projects are by defnition ceramic production focussed on sites. Regional analysis is almost In contrast to the topics above (typo-chronology; absent, due to a combination of, on the one regional variation), the sources of the raw hand, explicitly site-related research questions materials used in ceramic production may easily and, on the other, the absence of datasets in be addressed at the site level. Thanks to the DANS-EASY. A rare example can be found in a ongoing development of research methods (e.g. regional analysis of the Hazendonk potery from XRF, micromorphology), there is great potential Schipluiden,244 thanks to the simple fact that all to generate types of data other than those on of the sites used in the analysis had been studied morphology and decoration, which may be used by one of us (D.C.M. Raemaekers), mostly in an to develop new research questions concerning academic context. The analysis starts from the mode of production (e.g. at the household level, realisation that Hazendonk potery is relatively by a local specialist, imported), inter-group well defned in terms of morphology and contacts (locally produced versus imported), or decorative schemes. Within this framework, transfer of knowledge through generations. regional variation has been observed in terms of Unfortunately, these new approaches to ceramic the tempering agents used and the frequency of research have yet to fully fnd their way into specifc types of decoration, resulting in the Dutch development-led archaeology; whenever identifcation of three microregions along the they are incorporated, it always concerns pilot river Rhine. The patern of regional variation was studies on small numbers of sherds. interpreted as being the outcome of one or Establishing the source of the clay used in more of the following three variables: mobility, potery production allows researchers to exchange and mode of production. Raemaekers determine whether potery was produced locally hypothesises that the similarities in potery in or not. As such, it is an excellent way to address terms of temper and decoration between the a key question about prehistoric local three microregions, or microtraditions, result communities, namely, to what degree they are

^ŝŵŝůĂƌŝƚŝĞƐ ŝƐƐŝŵŝůĂƌŝƚŝĞƐ

DŽďŝůŝƚǇ ŽůŽŶŝƐĂƚŝŽŶ

ĞƌĂŵŽůŽĐĂů ǆĐŚĂŶŐĞŶĞƚǁŽƌŬƐ ŵĂƌƌŝĂŐĞ ƐǇƐƚĞŵ >ŝƚƚůĞĞŵďůĞŵŝĐ DŽĚĞŽĨƉƌŽĚƵĐƚŝŽŶ dƌĂĚŝƚŝŽŶ ǀĂůƵĞ 243 See ﬁg. 7.135 in Fokkens, Steﬀens & Van As 2016, 198. 244 Raemaekers & Rooke 2006; Raemaekers Fig 3.27 Interpretation of similarities and dissimilarities of potery traditions between microregions 2008. 245 Raemaekers 2008. (afer Raemaekers 2008). 109 —

Table 3.9 Correlation between density and average size of stone grit temper from Schipluiden.

<1 mm 1 - 2 mm > 2 mm Totals

Schipluiden < 15% 12,3 21,6 14,7 48,7

N = 2174 15 - 25% 5,2 10,5 10,6 26,3

> 25% 8,8 7,1 9,1 25,0

Totals 26,3 39,3 34,4 100,0

interwoven in regional and supra-regional The potery production process networks. Small numbers of potery sherds from As a rule, the analysed reports provide a good the Hazendonk sites of Schipluiden (n=22) and overview of the potery groups found, and Rijswijk-Ypenburg (n=10) have been analysed for describe technological, morphological and diatom contents. For Schipluiden, the results decorative characteristics. At the same time, the make a strong argument for local production prehistoric poters behind the pots are absent and extraction of coastal/estuarine clay near the from the reports. Questions concerning the surface in the majority of cases.246 However, one production process are not addressed, despite sherd stands out by the exclusive presence of the possibility of querying the databases (had freshwater diatom species, which suggests that these been available in DANS-EASY) generated the pot was produced elsewhere.247 The use of for the report in order to obtain more and new coastal/estuarine clay also shows in the diatom information on this topic. In this section, we content of a sherd from Rijswijk-Ypenburg.248 present one example concerning one stage in Interestingly, some pots there may have been the production process: the tempering of clay. made from the same clay as those from This phase of the process requires a poter to Schipluiden. Another example concerns diatom decide on the material to be used, the size of the analysis of fve Funnel Beaker and Vlaardingen temper particles and the density of the particles. sherds from Hazerswoude-Rijndijk. These All three choices must have been embedded belong to pots that were probably manufactured within a system of knowledge transfer within from local estuarine clay.249 and between generations of poters, in which For Knooppunt Hatemerbroek,250 the normative behaviour is the rule. Had this not analysis was based on geochemical (inductively been the case, we wouldn’t be able to recognise coupled plasma-mass spectrometry [ICP-MS]) prehistoric potery traditions at all. and diatom analysis of a limited numbers of The example presented here illustrates one sherds; as a result the results have limited aspect of this knowledge transfer system: the interpretive power. A single Vlaardingen/Stein size and density of stone grit temper. Table 3.9 sherd – from the Hanzelijn section of this site shows the correlation between the density and complex251 – was analysed. A total of 1.5% (n=4) the average size of stone grit temper particles in of the TRB sherds from this site were analysed. potery from the Hazendonk site of For another location within the site complex, Schipluiden.253 It indicates that the poters namely, the Bedrijventerrein section252, the TRB responsible for this assemblage had no strict sample is 0.9% (n=5). The authors conclude that ideas concerning density and particle size: the all sherds were produced from local clay. top score is 21.6%, some 10.5% above the Regretably, this pilot study did not lead to the expected score, even though all nine cells hold a conclusion that a more substantial study needs similar proportion of the sherds. In itself, this to be carried out, if for no other reason than to table and these fgures have no meaning, but 246 De Wolf & Cleveringa 2006, 294. 247 De Wolf & Cleveringa 2006, 294. verify sample representativeness. One might when one compares the results with those from 248 Cremer 2008, 209. therefore ask whether such a pilot study is of other sites or cultural groups, informative 249 Demiddele 2010, 236-253. 250 Brorsson 2011; Demiddele 2011. any relevance. One might equally ask why such conclusions can be drawn. To this end, small 251 Appendix II: PUBID 259. studies are not conducted more frequently in numbers of TRB sherds from fve megalithic 252 Appendix II: PUBID 260. 253 Data based on Raemaekers & Rooke development-led projects. tombs were studied (table 3.10). For these 2006. 110 —

Table 3.10 TRB sherds from 5 megalithic tombs.

Bronneger <1 mm 1 - 2 mm > 2 mm Totals

N = 28 < 15% 50 25 4 79

15 - 25% 0 4 11 14

> 25% 0 4 4 7

Totals 50 32 18 100

<1 mm 1 - 2 mm > 2 mm Totals

Diever < 15% 39 12 3 55

N = 33 15 - 25% 9 15 12 36

> 25% 0099

Totals 48 27 24 100

<1 mm 1 - 2 mm > 2 mm Totals

Havelte < 15% 55 10 0 66

N = 29 15 - 25% 17 7 7 31

> 25% 0 3 0 3

Totals 72 21 7 100

<1 mm 1 - 2 mm > 2 mm Totals

Odoorn < 15% 48 10 0 58

N = 31 15 - 25% 13 10 6 29

> 25% 0 6 6 13

Totals 61 26 13 100

<1 mm 1 - 2 mm > 2 mm Totals

Spier < 15% 40 13 0 53

N = 30 15 - 25% 3 17 23 43

> 25% 0 3 0 3

Totals 43 33 23 99

tombs, it is clear that most stone grit-tempered exchanged between the poters. Second, the sherds are characterised by a low density and diference in tempering strategies between small average particle size. Notwithstanding the Schipluiden and the TRB assemblages may possibility that the observed diferences relate indicate diferent potery production rules for to more practical aspects of potery production, diferent socio-cultural backgrounds. Hence, two possible interpretations come to mind. data that are generally produced within the First, the various TRB poters may have shared a context of develop-led research potentially notion on this aspect of the production process, provide information on the social aspects of indicating that technological knowledge was potery production. 111 —

Ware analysis Function The central role of potery analysis in New knowledge on the prehistoric past can develop-led research is to provide relative derive from incorporating not only new sites, dating, which aids the chronological anchoring but also new research methods. One of these of sites, features, structures and layers without research topics is that of the function of the the need for 14C-dates. As such, ceramic analysis potery, which can be investigated by scanning is a means, rather than a goal in itself, and electron microscopy (SEM) and chemical therefore, not surprisingly, the chapters on analysis, such as direct temperature-resolved ceramic fnds in the examined reports are mostly mass spectrometry (DTMS). These techniques descriptive in nature. Various aspects of the were applied in combination on seven TRB potery (morphology, temper, colour, thickness, sherds selected for analysis from Knooppunt decoration, etc.) are presented, but one fnds Hatemerbroek.257 The analysis revealed the litle discussion about the human behaviour that presence of both cooked plant remains and resulted in the observed paterning of the animal fats in some pots. Others yielded no material. In our perception, however, variation evidence about their use. The results, however, or, indeed, absence of variation requires have no clear relevance, because they are based explanation in terms of behavioural alternatives. on an extremely small sample (1.3%) of the The potential of this type of study can be available assemblage, they are not compared illustrated using the ceramic variability at the with results from any other site, and they are not Corded Ware site of Keinsmerbrug.254 It concerns incorporated into the synthesis of the project. a setlement site with a few striking A more fruitful application of such analysis characteristics: 86% of the bones derive from is demonstrated by results from Swiferbant-S3, birds – mostly ducks – and several wooden where sherds from 25 unique vessels were structures have a rather fimsy characterThese sampled.258 This analysis was based on the fndings could easily have led to a fnal largest potery fragments available, which interpretation as a specialised campsite, where allowed the researchers to seek correlations processing of fowl took place on several amongst the typological and technological occasions.255 The ceramic evidence, however, characteristics and the results of the SEM and provides crucial information for an alternative DTMS analyses. It became clear that the interpretation of the site.256 Whilst the number of Swiferbant-S3 potery can be divided into two sherds is very small (291 sherds >3 g), analysis subtypes. The frst subtype is relatively thick showed that the poters used seven diferent walled, plant-tempered ware, which appeared tempering materials, in 14 diferent to show no evidence of food containing emmer. combinations. This fnding added a new The second subtype is relatively thin-walled, dimension to the initial, subsistence-oriented stone grit–tempered vessels showing evidence interpretation of the site, and has led to the of food containing emmer (fg. 3.28). This proposition that Keinsmerbrug was a seasonally suggests that particular types of potery were occupied site where groups with diferent used to prepare food that contained emmer, backgrounds gathered to hunt for and feast on which in turn may have socio-cultural or ducks. This example shows that insightful ideological meaning. It is our impression that interpretations do not necessarily require new there is much to gain when this type of analysis types of analysis (it involved the same standard is an integral part of an explicit research design, description as conducted in all development-led instead of an unrelated, small-scale sideline of 254 Beckerman 2015. studies), but, rather, more detailed analysis of interest only to a few specialists. 255 Nobles 2012; Zeiler & Brinkhuizen 2012. the available data and discussion amongst the 256 Beckerman 2012. 257 Kubiak-Martens & Oudemans 2011. specialists involved. 258 Raemaekers et al. 2013. 112 —

Fig 3.28 Ceramic ware and lipid analysis of Swiferbant potery (from Raemaekers et al. 2013).

Concluding remarks no mater how small, there is certainly potential It is clear from the discussion above that to make some progress. The application of development-led projects have made litle scientifc techniques to gain information about contribution to the chronological aspects of clay sources, fabrics and functional aspects of ceramic typology and technology. This is partly potery also have the potential to make a serious due to the restricted size of assemblages: in contribution – a potential that seems to be many cases just a handful of sherds. However, underestimated in development-led research. ‘few sherds’ does not necessarily equal ‘no But it is not just scientifc techniques that need information’. Admitedly, potery fragments are to be used more ofen. Traditional approaches ofen too small to extract morphologically and to fabric analysis, for instance, can provide stylistically meaningful information, but valuable information; however, much of the archaeologists shouldn’t dismiss the potential work in a development context seems to end out of hand. When there is the possibility to with data collection and does not progress to obtain dates and contextualise assemblages, in-depth analysis. 113 4 People and space —

identifable paterning should show in the 4.1 Introduction archaeological record. The same can be expected for what people did at particular locations. Space is an integral dimension of human life. Although broader issues of landscape-scale Individuals move over shorter or longer behaviour in early prehistory may be difcult to distances on a daily, seasonal, annual or other grasp, particularly in the context of cyclic basis, and for a variety of reasons. development-led archaeology, one aspect Through movement, individuals perceive their deserves atention here. As outlined in living space, consciously or unconsciously, from section 3.2, the identifcation of raw materials a range of potential perspectives, and within a – lithics in particular – has received relatively specifc cultural and ontological seting. Hence, much atention. To some extent, work over the space provides meaning to those who move past 15 years has yielded data that permit us to through it; space is composed of zones and draw some further inferences about raw places connected by pathways or routes, material sources and mobility.260 Next, we will inherently referring to multiple scales in space address diferent aspects of the use of space at and time. This scale ranges from single spots the local scale or site level, notably, the where particular actions by single individuals understanding of fnds scaters and the take place at a particular moment in time, to identifcation of built structures. vast landscape areas used by multiple groups in diferent ways over a period of years. These scales represent a complex interplay of 4.2 Lithic resources in the landscape behaviours, which has resulted in the initial and and mobility continuous formation of the archaeological record. And this record has subsequently undergone further, post-depositional Whereas distance to raw material sources transformations. combined with materiality permits At a landscape scale, the cultural dimension archaeologists to develop hypotheses about of space is echoed by the interrelationships social interaction, distance to raw material between meaningful places and zones. Cyclic sources combined with technology permits us to movement and the diversity of locational use by draw inferences about transportation of people create a spatial structure, which, from an materials and mobility. Assemblages exclusively archaeological perspective, is not easily composed of raw materials from locally understood. In contexts of frequent residential available sources – that is, within the home and/or logistical movement of people, range – likely refect diferent strategies of a landscape becomes scatered with debris of resource exploitation and mobility than do which the ‘exact’ age is rarely established. assemblages that (also) contain raw materials Consequently, lack of chronological resolution originating from distant sources. The later induces ‘time collapse’, which makes it almost category of raw materials is commonly labelled impossible to identify systemic contemporaneity ‘exotic’. amongst traces of activity at the local scale.259 In Although raw material sourcing is contexts of limited (or absent) residential and/or conceptually straight forward, in many cases it is logistical movement, the archaeological problematic or even impossible to establish footprint will likely be scaters of debris whether a raw material was acquired locally or refecting the cumulative outcome of activity at at a distant source and transported (see section places in the landscape. Nonetheless, 3.2.1). The Netherlands has a complex geology contemporaneity of spatially separated and should frst of all be considered a sediment phenomena at the local scale will again be basin. Cycles of glacial and interglacial difcult to prove. This is not to say that topics conditions throughout the Quaternary have like ‘setlement systems’ or ‘mobility systems’ resulted in far-reaching processes of erosion and 259 Bailey 2007; Holdaway & Wandsnider cannot be seriously addressed archaeologically. sedimentation that, in turn, have led to 2008; Rossignol & Wandsnider 1992. If people’s spatial behaviour at a landscape scale continuous geomorphological reshaping and 260 Another aspect concerns Mesolithic pit hearths, but this topic will be discussed was repeated over generations, at least some geographical transformation. Major rivers – the in section 5.2.2. 114 —

Fig 4.1 Geographical distributions of primary and secondary ﬂint source areas in and close to the Netherlands that are (potential) sources of raw materials for the production of tools during the Late Palaeolithic through Middle Neolithic. Source areas of Scandinavian ﬂint in northern Germany and southern Scandinavia are not indicated. The background map shows the situation of the source areas relative to the southern North Sea basin, which acted as a large sediment basin and which was accessible during the Late Glacial and Early Holocene (referred to as Doggerland). Flint and other lithic resources that were transported by rivers surrounding the basin accumulated in Doggerland. It is unknown to what extent potential source areas in this basin were exploited (map reproduced with permission of National Geographic).

Rhine, Meuse and vanished Eridanos systems – have been transported by the Meuse River over transported vast quantities of rock (e.g. fint, a distance exceeding 250-300 km. quartz, quartzite, sandstone) from their primary When considering the exploitation of lithic sources far inland and deposited thick layers of resources, it is important to bear in mind that coarse sediment in the basin. Glaciers moving in the present-day geography is not representative from the north picked up rocks (e.g. fint, for the prehistoric situation, which saw granite, quartz) in Scandinavia, which were lef continuous change, e.g. in connection with sea in moraine deposits in the northern half of the level fuctuations. The availability of lithic raw Netherlands. All these rocks could be found as materials was very diferent, and the North Sea gravel and boulders in secondary position in the basin may have been important for the alluvial landscape. Primary sources of fint are only accumulation of nodules from various parts of encountered in the south-eastern part of the the surrounding areas (fg. 4.1). Primary outcrops Netherlands, notably in the southern part of could be found along the British, French and Limburg, where fint of the Rijckholt type is northern Danish coasts, as well as in the German found. However, due to erosion, fint nodules Bight. Hence, the primary fint sources in the 115 —

south-eastern part of the Netherlands and frequently, there seems to have been some adjacent parts of Belgium are simply amongst selection as well. High-quality, ofen transparent, the closest accessible for early prehistoric people fint was used for the production of long and living in what today we call the Netherlands. regular blades, which were modifed into a range Whether these sources were actually exploited of tools (points, scrapers, burins, Zinken), e.g. at throughout early prehistory is, however, another Epse-Olthof.265 Sometimes, assemblages contain mater. only end-products and/or large blades made of Below we will take a closer look at some such high-quality fint, in addition to lesser results that can be drawn from development-led quality fint, e.g. at Stroe.266 In other cases there is research. The purpose is not only to also production waste resulting from the demonstrate to what extent the sourcing of raw knapping of high-quality fint, e.g. at Meerveld materials adds to our knowledge about and Epse-Olthof.267 Such diferences suggest that regional-scale spatial behaviour, but also to nodules, cores, half-products and fnished tools identify some major problems. were carried around, and that waste was lef at various spots in the landscape. As yet, it is uncertain where high-quality 4.2.1 Late Palaeolithic long-distance fint was acquired. Flint of Scandinavian origin transport occurs in substantial quantities in moraine deposits in the northern half of the Netherlands, but mostly nodules are rather small and of poor One notable contribution of development-led to mediocre quality.268 Large nodules of high- research to our insight into long-distance quality fint are rare. The Epse-Olthof transport of fint concerns the Late Palaeolithic assemblage also contains high-quality fint Hamburgian. People of the Hamburgian were (fg. 4.2), and waste indicates on-the-spot the frst to colonise the landscape north of the production of blades, as well as maintenance Rhine afer the Last Glacial Maximum (LGM) and use of tools. Amongst the fint categories – which ended about 20,000 years ago. The was red Heligoland. This particular type of fint is Hamburgian is considered to represent a only known to occur on the nowadays heavily technological tradition rooted in the Late eroded island of Heligoland, in the German Magdalenian.261 The sites atributed to the Bight, at a distance of c.250 km as the crow fies Hamburgian on the ice-pushed ridges of the from Epse-Olthof. A small number of blade Veluwe (central Netherlands) represent the fragments and small chips that could almost all southernmost extent of this tradition. Here, be refted into a single unit provide clear both narrow-spectrum and broad-spectrum evidence for on-the-spot knapping. However, assemblages have been found, suggesting most of the nodule seems to have been worked inter-site functional diferentiation.262 During elsewhere, whilst the core was not been rescue excavations near Deventer (Epse-Olthof), abandoned at this location either (at least not at the same latitude but east of the Veluwe, within the extent of the excavated area). This Hamburgian artefacts were found on a sandy means that a nodule of red Heligoland fint was ridge bordering a palaeochannel.263 Although the transported at least 250 km (in reality this will original context of this material appears to have have been much farther, considering the fact been disturbed by Mesolithic and later activity, that people do not travel in straight lines) and the composition of the assemblage suggest the that only some fragments of that nodule were artefacts to represent the remains of one or lef at Epse-Olthof. more short-lived encampments, possibly In consideration of the Hamburgian being a hunting stands.264 The Epse-Olthof site thus tradition within the Late Magdalenian, which is extends the limit of the Hamburgian territory. represented in the southern part of the 261 Barton et al. 2003; Weber 2012. One striking characteristic of many Netherlands, the fact that people travelled over 262 Rensink & De Kort 2012, 2013; Peeters & Rensink in prep. Hamburgian lithic assemblages in the such a distance is in itself not surprising. 263 Appendix II: PUBID 733. Netherlands is the use of high-quality, fne- Comparable distances have been recorded for 264 Hoebe 2015; Peeters & Rensink in prep. 265 Peeters et al. 2015. grained fint of northern (i.e. Scandinavian) the Magdalenian.269 Apparently, groups of the 266 Rensink & De Kort 2012. origin. Despite the fact that nodules of mediocre Magdalenian and Hamburgian people travelled 267 Rensink & De Kort 2013; Hoebe 2015. 268 Beuker 2010. quality – showing frost cracking – and small were over long distances, probably in a cycle of 269 Terberger et al. 2013. 116 —

Heligoland

Damme Dümmer Epse Olthof

050 100 200 km

Fig 4.2 Geographical locations of Epse-Olthof and Damme Dümmer relative to Heligoland.

winter–summer residential mobility. As humble Neolithic, Bronze Age) for the production of as the Epse-Olthof fnd may appear – afer all, bifacial daggers and sickle blades.272 As yet, we 270 Fries & Veil 2014. 271 Van Gijn & Raemaekers 2014. The cliffs we only have a handful of waste – the do not know whether this variety of fint was of red-stained chalk containing the flint conclusions we can draw from it are far- also used by Hamburgian people. However, may, nonetheless, have had some special significance in this otherwise reaching. Apart from an isolated core from the Heligoland is located relatively close (c.100 km) rather flat, undulating landscape. It is, site of Damme-Dümer (Germany; fg. 4.2),270 it is to the moraine outcrops in Schleswig-Holstein however, not clear what the situation was during the time of the Hamburgian the only evidence we have for the use of fint and adjacent parts of Denmark. These moraine habitation; there may not have been coming from Heligoland; we have no indications deposits are rich in voluminous nodules of in any cliffs at all. 272 Beuker 2010; Van Gijn 2010; Beuker & for any particular selection on this type of fint. high-quality Scandinavian fint. From this Drenth 2014. Bearing in mind the sparseness of information perspective, it may very well be possible that the 273 It should be noted that the identification of the red Heligoland flint was a about source areas, this in itself suggests that high-quality fint that occurs in Hamburgian coincidence. The excavator, first, was the remarkable red fint did not have any assemblages in the Netherlands was in fact not aware that he had discovered Hamburgian artefacts, and, second, was particular connotation for the people of the acquired in this region. On the other hand, high- not certain about the identification of Hamburgian, unlike what has been suggested quality nodules could also have been collected the red-coloured flint. On the basis of photographs, the Hamburgian diagnosis for the use of this type of fint in the Neolithic.271 from glacial till in the northern Netherlands, was established by Dr. J.H.M. Peeters. More importantly, it shows that Heligoland since the quality of fint from glacial till is Once the artefacts could be studied, the cultural affiliation was confirmed, whilst represented a source within the territorial range generally beter than that from boulder sand the possibility of the artefacts being where high-quality fint could be acquired. (residual deposit of eroded till). Hence, the made on red Heligoland flint was deemed probable. Confirmation came Heligoland also is a source for a grey–brown handful of ‘remarkable’ waste from Epse-Olthof from flint experts Harm Paulsen type of fint that is known to have been has led to fresh insights and opened new lines of (Schleswig, Germany) and Jaap Beuker (Assen, the Netherlands) at a later stage. exploited in late prehistoric times (Late investigation.273 117 —

4.2.2 The Wommersom distance-decay distribution

A characteristic raw material used by Mesolithic hunter-gatherers in the southern part of the Netherlands and adjacent parts of Belgium, and continuing into northern France, is Wommersom quartzite.274 This relatively fne-grained rock originates from a natural outcrop near the village of Wommersom, close to Tienen, in Belgium.275 Its knapping qualities are comparable to that of good-quality fint, and it is perfectly suitable for the production of regular blades. Indeed, such blades, as well as trapezoid points made out of such blades and feuille de gui surface-retouched points, are ofen knapped from Wommersom quartzite.276 It was frst noted in the 1980s that Wommersom quartzite shows a clear geographical distribution patern, roughly delimited by the Rhine in the north and east, the Scheldt in the west, the present-day French– Belgian border in the south, and the Ardennes in the south-east (fg. 4.3).277 North of the Rhine, sporadic fnds of Wommersom quartzite have Fig 4.3 Geographical distribution of Wommersom been reported.278 In quantitative terms, the quartzite (source location indicated by the yellow dot) overall distribution seems to be best described during the Mesolithic. The extensions outside the by a distance-decay curve, where the frequency dashed polygon are based on the ﬁnds of incidental of occurrence and quantity of material decreases pieces of Wommersom quartzite at the sites of as the distance from the source area increases. Roterdam-Yangtze Harbour (red dot) and Hoge In view of the co-occurrence of feuille de gui Vaart-A27 (blue dot). The light pink shading represents a points and stylistic aspects of trapeze-shaped possible further extension of the distribution of this points, it has been suggested that the resource (from appendix II: PUBID 462a). distribution patern possibly corresponds to a territory, and hence has social signifcance.279 Wommersom quartzite ofers an interesting case study for raw material distribution paterns decay patern is also inherent to ‘average’ in the Netherlands and surrounding areas. This distributions of other lithic raw materials, but type of rock is relatively easy to identify, and which are impossible to map in the absence of only one source is known where it could have ‘unique’ sources and mutually exclusive been collected by Mesolithic hunter-gatherers in properties. Theoretically, it is possible that the quantity. Other lithic raw materials, such as ‘reality’ consists of a continuous spread of Rijckholt fint and various types of Scandinavian overlapping distributions, making Wommersom fint, do not share these advantageous just one of the components. Furthermore, we properties, since these fint varieties can are looking at a palimpsest distribution at a originate from multiple primary and secondary landscape scale – the outcome of several sources, spread over wide areas. Consequently, millennia of deposition and accumulation, 274 Also known as grès quartzite de Wommersom (GQW). these can barely be mapped in a useful way, if at instead of a ‘discrete’ spatio-temporal 275 Cnudde et al. 2013. all. The social interpretation of the geographical phenomenon. 276 Cf. Amkreutz 2013. 277 Gendel 1984, 1989; Van Oorsouw 1993. distribution of Wommersom quartzite on Despite these problems, it is worthwhile to 278 Peeters et al. 2001. archaeological sites is also not without investigate the distribution of Wommersom 279 Gendel 1984, 1989. 280 Peeters 2007, 198; Peeters & Niekus problems.280 It is possible that the distance- quartzite from the perspective of transportation 2005, 226 (footnote 30). 118 —

of raw materials or tools made out of specifc not be investigated with much precision, but the raw materials. If the distribution patern refects fnds are of Early or Middle Mesolithic age based an ‘averaged’ spreading of a systematically used on the site’s geology.283 At Ede-Kernhem, the raw material over the course of the Mesolithic, great majority of the assemblage dates to the then the Wommersom case helps us to gain Early Mesolithic, and only one piece of insight into mechanisms of dispersion. Reports Wommersom quartzite (a point) was found.284 resulting from development-led work make litle One other ‘geographical outlier’ of Wommersom mention of Wommersom quartzite. It is not clear quartzite remains uncertain with regard to its to what extent this is a consequence of expertise identifcation (Roterdam-Groenenhagen), and in identifying lithic raw materials. Most sites another has been rejected (Schoonebeek) by one where it has been identifed are located in the of us (M.J.L.T. Niekus). provinces of Noord-Brabant and Limburg (e.g. From this overview of Wommerson quartz Borgharen Site 8, Susteren-Aardenweg, Deurne- fnds, it can be concluded that results of Groot Botelsche Akker and Casteren). All are development-led excavations have not located well within the geographical extent produced a radically new picture. Instead they indicated above. The proportion of Wommersom confrm the general distribution map. quartzite in assemblages counting more than Nonetheless, it is important to note the presence 100 artefacts is variable, ranging between 0.3% of Wommersom quartzite in the Roterdam (Borgharen) and 20% (Casteren). The site of region.285 During the Early and Middle Mesolithic, Casteren probably represents a short-lived special the stream valleys of the Rhine-Meuse system purpose encampment where Wommersom and the Scheldt joined just west of the Yangtze quartzite was used as a raw material for blades. Harbour (on what is known as the Maasvlakte, Montbani blades are made exclusively on Roterdam), and eventually formed a vast Wommersom quartzite. Several other estuary (fg. 4.4).286 The river Scheldt and its development-led projects (Keersop-Deelgebied tributaries connected inland parts of the II, Tungelroyse Beek, Sevenum-Gelderdijk, Flanders sandy plain as far as the source area of Ekkersrijt, Middegaal, Holtum-Noord II) have Wommersom quartzite, notably via the Rupel- produced stray fnds of Wommersom quartzite. Dijle-Gete system. These rivers or steam valleys Most notable are incidental fnds of possibly served as routes along which people Wommersom quartzite in the assemblages from travelled and carried various raw materials in a Roterdam-Yangtze Harbour, Roterdam- north-westerly direction.287 If water routes were Beverwaard and Ede-Kernhem, which are important to the Mesolithic ‘infrastructure’, then located at the geographical fringe of the the Scheldt system possibly played a key role in distribution. The Roterdam sites represent the relation to Wommersom quartzite, as the great most westerly occurrences known at present. In majority of smaller rivers in this area drain into the case of Roterdam-Beverwaard, three out of the Scheldt. Smaller rivers in the south of the four artefacts on Wommersom quartzite are Netherlands drain into the Meuse; in fact the leaf-shaped points, whilst the fourth is probably watershed separating the Scheldt and Meuse a fragment of such a point. Importantly, three of drainage systems is located just south of the 281 8465 ± 45 BP; 8435 ± 40 BP; 8135 ± these come from three pits that contained Dutch–Belgian border. However, we must bear 45 BP (all on cremated bone; Niekus et cremated human and animal bone (also see in mind that many tributaries of both rivers al. 2016). 282 6770 ± 40 BP (on cremated bone; section 3.3.2), and for which a Middle Mesolithic originate more or less in the same area, which Niekus et al. 2016). age could be established through radiocarbon would have made it easy to cross from one 283 Vos & Cohen 2015. 284 Devriendt 2015. dating.281 Another pit with cremated human bone system to another. Clearly, none of these smaller 285 Note that the Roterdam is one of the also contained a leaf-shaped point, yet made rivers permited waterborne transport in the most westerly areas where Mesolithic and Early Neolithic sites are found close out of fint, and it is dated in the Late source area. Hence, much transport will have to the present-day surface and can, Mesolithic.282 The specifc context of the been overland in the upstream areas, whilst therefore, be investigated. Hence, the site distribution does not represent Roterdam-Yangtze Harbour Wommersom fnds waterborne transport may have been farther prehistoric reality. (a blade fragment and a burnt fake) is not downstream. 286 Vos & Cohen 2015. 287 Peeters et al. 2015. known since the site is underwater and could 119 —

Fig 4.4 Simpliﬁed map of the Rhine-Meuse-Scheldt drainage system in the Netherlands and adjacent parts of Belgium and Germany between 6500-6000 cal. BC. Brown: Rhine system; green: Meuse system; orange: Scheldt system; blue: open sea and main estuaries (based on Kiden 2006; Vos 2015).

varieties of fint show quite a number of 4.2.3 Newly ‘identiﬁed’ and ‘ghost’ raw similarities, as well as diferences. This is material sources particularly true for varieties in the ‘Rijckholt family’ to which Lanaye (i.e. Rijckholt), Banholt, Lousberg and Rullen fint belong. Similarly, the Traditionally, reports resulting from distinction between Rijckholt, Banholt, Spiennes development-led archaeology list a rather (Belgium) and ‘Belgian grey’ can be problematic restricted range of lithic raw materials, for the untrained eye. The variable experience of generalised under such labels as ‘Rijckholt’, lithic analysts in the identifcation of raw ‘Meuse fint’, ‘terrace fint’, ‘bryozoa fint’, materials may, in fact, be the most important ‘Scandinavian fint’ or ‘northern fint’ (fg. 4.5). reason. Few are familiar with the broad range of The lack of more specifc assignments is partly fint varieties, and even fewer are able to actually due to the materials themselves, as many distinguish between them. As a consequence, 120 —

Fig 4.5 Flint varieties encountered in early prehistoric contexts in the Netherlands: (a) Lanaye, (b) Banholt, (c) Rullen, (d) Lousberg, (e) Valkenburg, (f ) Simpelveld, (g) Vetschau, (h) Obourg. 121 —

Fig 4.5 Continued: (i) Orsbach, (j) Lixhe, (k) Hasbaye, (l) Otignies phtanite, (m) Dan flint, (n) Scandinavian bryozoa flint, (o) red Heligoland flint, (p) tabular (grey-brownish) Heligoland flint, (q) Falster flint, (r) speckled Senonian flint, (s, t) Scandinavian Senonian flint (a-l: photos by T. Penders, reference collection of Dr M. de Grooth; m-t: photos by J. Beuker). 122 —

results of development-led ‘specialist analyses’ What about ‘ghost’ sources? At Knooppunt are difcult to compare and do not permit Hatemerbroek, two unidentifed types of fint researchers to draw reliable conclusions with were reported for several contexts, dating to regard to the use of raw material sources during (possibly) the Late Palaeolithic (Context 2.08 and early prehistory. However, we will take a look at 3.03) and Early Mesolithic (Context 6.12); this some specifc aspects regarding the range of dates may point to a chronologically identifcation of lithic raw materials of which the connected exploitation of those varieties.292 The source had not previously been identifed. frst unidentifed variety is described as a fne- In addition, we will consider some ‘ghost’ grained light to dark beige fint with bryozoa and materials: raw materials of which the source has incidental inclusions (fg. 4.7).293 The material can not been identifed, but actually could be be considered to be Scandinavian bryozoa identifed. fint,294 which can be found in glacial moraine One variety that makes a frst appearance in deposits in the northern Netherlands. The some reports reviewed for this study is so-called second unidentifed variety is described as a Cap Blanc Nez fint, a fne-grained, somewhat fne-grained, multicoloured banded fint; colour translucent blackish to dark grey–coloured fint bands are diferent shades of grey, sometimes that is frequently motled with lighter inclusions with a beige to light brown tone.295 The (fg. 4.6).288 Nodules consist of rounded, ovoid pronounced bands vary in width and are pinkish pebbles with a pounded outer surface and bluish to purplish in colour when burnt. Based on the to greyish colour. A pounded outer surface is published photos, we consider it almost certain characteristic of fint nodules exposed to rolling that this material is actually Senonian Falster in the intertidal zone. As the fint bears some fint, a variety that has its primary origin in similarity to material that is eroding out of the Denmark (Falster, Lolland and southern Zealand) coastal clifs at Cap Blanc Nez near Calais but can be found in secondary position on (France) it is now designated to that source, beaches and in glacial moraine deposits.296 despite the fact that no analytical data Small nodules of mediocre to poor quality can (microscopic; chemical) exist to confrm this be found in the northern Netherlands where origin. The pebbles may have been transported Saalian till is encountered near the surface. over large distance by sea currents – perhaps, Whether or not the material from Knooppunt as some have suggested, as far as the Belgian Hatemerbroek comes from sources in the province of Zeeland289 – which makes it Netherlands is impossible to say on the basis of impossible to determine where the material was the published information. Likewise, it is collected. As yet, this type of fint has only been impossible to say that it comes from sources 288 Houkes 2008; Van Gijn & Houkes 2006. reported from the Hazendonk sites of closer to the primary occurrences in Denmark. 289 Van Gijn & Houkes 2006, 133. Schipluiden and Rijswijk-Ypenburg.290 It is, However, in the context of the Late Palaeolithic, 290 Houkes 2008; Van Gijn & Houkes 2006. 291 Van Gijn & Houkes (2006, 133; pers. however, possible that this raw material is it is far from excluded that materials originated comm. P. Crombé) refer to the site of confned to Neolithic coastal sites, as it is also from distant sources, as we have seen with Doel, near Antwerp. 292 Verbaas et al. 2011, 339, 342, 344, 349, known from Belgian near-coast contexts.291 regard to the use and transport of red 357, 365. The question whether it might be represented in Heligoland fint by Hamburgian hunter- 293 Verbaas et al. 2011, 343. 294 Source assignment based on a assemblages from other periods and sites gatherers. If this was also the case for the Falster published photo (Verbaas et al. 2011, 341, cannot be answered, as its identifcation seems fint, it urges us to think diferently about ﬁg. 7.1). 295 The colour description is based on to be connected with the few researchers who regional-scale use of space and mobility. published photos (Verbaas et al. 2011, are familiar with this variety of fint, at least for 341, ﬁg. 7.1). 296 Högberg & Olausson 2007, 96-99. the time being. 123 —

Fig 4.6 Flint varieties identified at Schipluiden. a: rolled pebble; b: Cap Blanc Nez; c: Belgian flint (Hesbaye?); d: motled greyish flint, probably Belgian, with waxy texture; e: relatively coarse-grained black flint; f: grey flint from Spiennes or possibly Rijckholt (from appendix II: PUBID 732). 124 —

Fig 4.7 ‘Unidentified’ flint varieties from Knooppunt Hatemerbroek. The variety in photo a may be Scandinavian bryozoa flint; the variety in photos b-e can be identified as Falster flint (from appendix II: PUBID 259). 125 —

expected to generate. Low-density scaters with 4.3 Place and space more discrete spatial and temporal resolution are rarely identifed in the survey stage of research, and will only show up in excavations that cover a It is evident that archaeological questions about substantial surface area. But as excavations of the use of space are not confned to the vast areas mostly involve the documentation of landscape scale. Because the great majority of late prehistoric and/or early historic setlement feldwork is focussed on ‘sites’, which are or funerary contexts,298 low-density scaters of defned in terms of concentrations of material early prehistoric remains tend to be found by remains of human activity and anthropogenic chance – the ‘by-catch’ we mentioned earlier. soil features, most questions are related to the If such scaters are excavated at all, standard use of space at the site scale.297 Typically, spatial methods are followed, instead of methodologies analysis of scaters of debris and features aims tailored to the study of early prehistoric remains at the identifcation of dwelling structures or and which can be expected to return the other built structures, and to defne activity maximum gain from the potential at hand. areas or special purpose areas. It is about the To sum up, in development-led archaeology, organisation of space, and, in the ultimate early prehistoric sites face two problems where it situation, analysis should enable us to describe comes to research questions aimed at what people – even individuals – at some understanding how people used (and perceived) moment in time did within the spatially space at the local (intra-site) scale. The frst is constrained window of the excavation, perceptual. Palimpsests are believed to have a snapshot of the past. Clearly, nowadays limited information value, due to the spatially and archaeologists are perfectly aware of the fact likely chronologically mixed nature of the that this is not an easy task and, indeed, in the archaeological remains, which is hard to decipher. majority of cases an impossible one. Repeated The second relates to survey design and priority use of locations or particular landscape zones seting. Low-density scaters with potentially has led to the merging of scaters of material discrete spatial and chronological resolution, and remains and features, as well as accumulation hence high information value, are rarely subjected and mixing. The resulting palimpsests require to excavation, due to problems of site detection being unravelled to make them understandable (survey design) and infexible priorities in contexts in terms of spatially and temporally ‘discrete’ where such scaters are found by chance. To what entities, which can subsequently be interpreted extent can these problems be overcome? As in terms of behaviour. Due to the fact that argued by Bailey,299 palimpsests are the result of ‘archaeological time’ difers from ‘ethnographic particular aspects of structured human behaviour, time’, owing to the rather limited chronological and hence represent an important phenomenon, resolution of the archaeological record, this goal which needs to be considered in archaeological remains the unreachable Holy Grail. theory building. Clearly, there are limits to the sort This problem has led many archaeologists to of information that can be extracted from believe that palimpsests cause nothing but palimpsests: it is of no use to search for ‘Activity X’ trouble when it comes to the ‘reconstruction’ of conducted by ‘Individual 1’ at ‘Point-in-time T’ spatial behaviour, and all the more so for periods unless one is working on a single-phase, low- during which people had the habit of building density site. However, palimpsests can be dwelling structures that lef few recognizable understood in terms of underlying structured traces in the archaeological record. Yet, behaviour, which has resulted in the identifcation palimpsests probably form the overwhelming of recurrent paterns within an otherwise majority of contexts excavated in development- intangible mass of material and/or features.300 led archaeology, simply because these have a To make palimpsests archaeologically meaningful beter chance of being found during surveys due requires a change of perspective and of to their relatively high density of artefacts. And it methodology. To second problem is one of is this high fnds density that is ofen used as an atitude. Here, the development of less rigid 297 See Lock & Molyneaux (2006) for discussion on various issues of scale. argument in favour of excavation: the project outlines and the willingness of frms to 298 Such excavations are normally focussed investment in time and money is valued in terms involve specialists in a timely manner are of on machine-aided feature recovery. 299 Bailey 2007. of the number of boxes of fnds that the site is importance. 300 Peeters 2007, 2010. 126 —

Some examples from academic studies 4.3.1 In search of spatial diﬀerentiation Initial eforts to discern archaeologically meaningful paterns within the extended fnds distribution recorded at Hoge Vaart-A27 – a vast So how has development-led research dealt palimpsest – was limited to visual inspection of with issues of spatial behaviour at the intra-site distribution maps, created for a variety of fnds level? In the NOaA 1.0, research questions are categories and features.302 This, however, did not focussed on the impact of depositional and lead to any deep insight into the underlying post-depositional factors on the archaeological formation processes, nor to identifcation of record, the interpretive possibilities, the spatial structures, such as hut or house plans. identifcation of activity areas and phases of Further analysis of stratigraphically defned habitation, the identifcation of built structures, grid-cell units was conducted along two lines: and even the signifcance of empty space (see 2D ‘fuzzy set’ and cluster analysis and 3D spatial appendix I). analysis.303 The 2D analysis built on the spatial Questions like these generally form the paterning discerned within a low-density basis of the design of feldwork strategies and distribution, which had been recorded in relative choice of methodology. Spatial units of fnds isolation at the site (fg. 4.8).304 It served as a collecting and recording are expected to be ‘model’ for the identifcation of structural appropriate to discern archaeologically paterns ‘embedded’ within the vast palimpsest meaningful paterns within an otherwise that made up most of the distribution. This continuous distribution of material remains and approach also permited the potential soil features. Point location collecting and identifcation of ‘deviant’ paterns, which also recording (piece ploting) is considered the most require explanation. Based on this analysis, it precise by many, but it is clearly a time could be concluded that the palimpsest consuming and consequently costly approach. distribution was primarily composed of material Used within academic research projects until the remains that were also found in the low-density 1980s, such ‘high-resolution’ methods are no distribution, and that despite the process of longer applied. Insights into high-resolution mixing and blurring, the spatial characteristics of distribution paterns of material remains have various aspects of behaviour were still led to a new standard of fnds collection and identifable.305 In a way, specifc aspects of recording in stratigraphic and/or standardised activity had lef a spatial ‘fngerprint’. The 3D (yet arbitrarily chosen) spits in 50 × 50 cm grid spatial analysis more specifcally explored the units. Soil features are either treated as single relationships of diferent categories of features entities (polygons) or integrated into a (pit hearths and surface hearths in particular) to stratigraphically defned grid system. Although the distribution of material remains. The the database structure connected with such a independently obtained results permited the grid-oriented methodology is rather straight researchers to underpin their impressions with forward, spatial analysis of the data is not. regard to stratigraphic relationships and Before turning to the results from development- phasing, which were largely founded on feld led projects, we will briefy illustrate this with observations.306 301 Hogestijn & Peeters 2001; Kleijne et al. insightful examples from ‘academic’ analysis of In the case of the three Late Neolithic sites, 2013; Peeters 2007; Smit et al. 2012; Theunissen et al. 2014. extended datasets from the Mesolithic–Neolithic legacy data were used to analyse spatial aspects 302 Peeters & Hogestijn 2001. site of Hoge Vaart-A27 (province of Flevoland) that had thus far only been described on 303 3D analysis conducted by Merlo (2010); 2D analysis conducted by Peeters (2007). and the Late Neolithic Corded Ware Culture sites the basis of visual inspection of excavation 304 Peeters & Hogestijn 2001. of Keinsmerbrug, Mienakker and Zeewijk plans. As in the case of Hoge Vaart-A27, data 305 Peeters 2007. 306 Merlo 2010. (province of Noord-Holland).301 referred to stratigraphically defned grid-cell 127 —

Fig 4.8 Screenshots of the 3D analysis of ﬂint and potery frequencies in features at Hoge Vaart-A27 (from Merlo 2010). Although such data can easily be displayed in a 2D environment, the 3D visualisation makes stratigraphic relationships between surface hearths (large quantities of ﬂint or potery) and pit hearths (quantities close to zero) more clear. 128 —

units. Through a combination of visual data Some examples from development-led work inspection, quantitative analysis and Let us now turn to development-led multivariate visualisation, it was possible to archaeology. First of all, it appears that any sort discern archaeologically meaningful paterns of enquiry of spatial data is focussed on the within these highly complex datasets.307 Not only identifcation of activity areas – areas that, on was it possible to identify particular activity the basis of material remains as being used as zones, it was also possible to link such zones to proxies for particular behaviours, are interpreted structural features, such as house plans.308 as functionally distinct spatial entities. Clusters Although these structural features were partly of such remains, or proxies, are ofen viewed as apparent from the excavation plans (recorded representative for spatially ‘bound’ activity. This, postholes), more detail was brought to light however, is an assumption: “The traditional concept through the analysis of distribution paterns of of ‘activity area’, although perhaps useful in terms of material remains. In the case of Mienakker, for observable activity performance (e.g. in an ethnographic instance, it became clear that a structure initially context), is not necessarily a valid concept in terms of interpreted as a house by the excavator,309 in fact deposition. Simply put, people might well perform is part of a much larger structure associated with ‘activities’ in ‘areas’, but there is no reason to expect a human burial that is interpreted as a ritual them to map those areas with their garbage; material structure.310 This structure shows striking products of activities may ofen be collected in dump similarities in overall layout with a structure locations along with the products of other activities documented at Zeewijk.311 At Keinsmerbrug, the performed in other areas.”315 possible presence of a single, fimsy hut was In other words, the spatial occurrence of a initially reported, but further analysis permited behavioural proxy does not necessarily map the the identifcation of a succession of rebuilt actual space where that particular activity was dwelling structures in connection with a executed. Activities are part of a more or less surprisingly broad range of potery fabrics and complex context of cultural behaviour styles.312 Finally, the analysis of the Zeewijk connected with the handling of ‘waste’ or dataset permited the academic researcher to ‘garbage’. In addition, archaeological spatial identify particular activity zones and to deepen paterns are the result of myriad post- our insight into depositional and post- depositional processes that have transformed depositional formation processes (fg. 4.9).313 the paterns that initially emerged, and which Altogether, the in-depth spatial analysis of these need to be understood in order to interpret the three sites permited the researcher to develop archaeological paterns in terms of human an entirely new insight into Corded Ware Culture spatial behaviour.316 Indeed, to reach such an setlement dynamics and the social meaning of understanding requires in-depth analysis of places.314 spatially referenced data, data which are It is important to note that none of this collected as a standard procedure in 307 Nobles 2012, 2013, 2014. would have been achieved if the analysis had development-led archaeology. Yet, the great 308 For a discussion about criteria for house reconstructions, see Fokkens, Steffens & been restricted to visual inspection of excavation majority of projects have not incorporated Van As 2016, 59-66. plans and maps showing plain distributions of spatial analysis as a specialism within the project 309 Hogestijn & Drenth 2000/2001. 310 Nobles 2013. selected fnds categories. Furthermore, it has setup. As a consequence, interpretation of 311 Nobles 2013. become clear from these studies that there is no recorded spatial paterns is normally restricted 312 Nobles 2012. 313 Nobles 2014. of-the-shelf solution to the analysis of to visual inspection of (computer-generated) 314 Nobles 2016. Fokkens, Steffens & van As intrinsically complex spatial data, partly because distribution plots, but exceptions to the rule do (2016) have expressed their doubts about these interpretations, due to the data are not necessarily mutually comparable, exist, e.g. for Dronten-N23 and Schipluiden.317 problems of stratigraphic complexity even when similar collecting and recording The lack of in-depth spatial analysis is without a and the palimpsest nature of these sites. 315 Rigaud & Simek 1991, 217. methods are used, and partly because doubt (at least partly) a consequence of spatial 316 Schiffer 1999. each archaeological site has unique restrictions on the feldwork: the area being 317 Wansleeben & Laan 2012; Wansleeben & Louwe-Kooijmans 2006. characteristics. developed does not cover the ‘entire’ site; 129 —

Fig 4.9 Example of the spatial analysis of the Late Neolithic site of Zeewijk (from Nobles 2016). A: Kernel density estimates and multivariate visualisation of the ﬂint data (KDE bandwidth = 1.75 m); B: elevation (a), slope (b), the ﬂint multivariate visualisation (c). 130 —

hence, it is partly excavated. Although this does inferences about spatial behaviour relied on not exclude the possibility of conducting spatial visual inspection of distribution plots. Similar analysis per se, it does restrict broader approaches can be noted for Ede-Kernhem and interpretations of analytical results due to the Dronten-N23.321 fact that the unexcavated part of the site The identifcation of remains of built remains a black box. In such cases, the choice to structures, such as storage facilities and put litle efort into the analysis of spatially especially houses, is as problematic as the referenced data is defensible. identifcation of archaeologically meaningful Our analysis uncovered many examples paterning in fnds scaters.322 One possibility is from the past 15 years of development-led to adopt a data-driven approach that aims at the archaeology. Roterdam-Groenenhagen, for a posteriori identifcation of structures from instance, is a site of the Early Neolithic excavation data. Indeed, many claims for Swiferbant Culture, located on an aeolian river structures derive from the inspection of digitised dune, and of which only a 23 × 10 m trench could excavation plans.323 Of course there are some be investigated (fg. 4.10).318 The full spatial drawbacks to this approach, particularly when extent of the site is unknown, as the assessment the density of features (postholes) is high: was restricted to the area to be impacted. The straight or curved lines are easily found, but the trench clearly does not comprise the ‘full’ individual features may in fact have had no distribution of archaeological phenomena, structural relationship in prehistoric reality. making any interpretation of observed clustering In the case of clear plans, reliability is still of particular materials within the distribution acceptable, but in the case of irregular precarious. Another example is Roterdam- confgurations, reliability decreases without the Beverwaard, where Mesolithic and Neolithic availability of supporting data.324 Preferably, remains were found, amongst which remains of built structures should be recognised (exceptional) pits with cremated human bone in the feld, as this can direct choices with regard dating from the Mesolithic.319 The site is located to the treatment (sections, sampling) of the on the same river dune as Roterdam- features comprising the structure. However, this Groenenhagen. Material remains were collected does not obviate the usefulness of desktop in 1 × 1 m squares within a 20 × 15 m excavation analysis of excavation plans, whether or not in pit, but again the distribution of remains combination with fnds scater analysis; the evidently continues beyond the investigated interaction of specialists in cultural phenomena area (fnd distributions stop abruptly at the and specialists in spatial analysis may prove limits of the excavated area). Spatial paterns proftable. are superfcially discussed on the basis of visual It is mainly in the context of large-scale inspection of the distribution plans. development projects that opportunities exist to It is important to note that the lack of anticipate spatial analysis at a level beyond spatial analysis is not restricted to partially ‘eyeballing’. Not only do possibilities exist to excavated sites. Several sites that were assess the potential on the basis of more subjectively defned as ‘delimited’ scaters of extensive survey results, there is also is – ofen – archaeological remains, and which are supposed more money available for post-excavation to have been completely excavated, were not analysis. In the case of the Mesolithic sites of subjected to spatial analysis, despite the Leeuwarden-Hempens and Dronten-N23, and availability of reliable spatially referenced data. the Middle Neolithic site of Schipluiden, for Again, the lack of analysis may relate to limited instance, more efort was put into the budgets, but it is also possible that the potential characterisation of spatial paterns and the is being underestimated or that there is analysis of interrelationships.325 Typically, these 318 Appendix II: PUBID 468. 319 Appendix II: PUBID 463a. insufcient in-house knowledge about analytic large-scale projects also involve academic 320 Verneau-Peeters & Fermin 2007. methods. At Zutphen-Looërenk, for instance, a researchers. Schipluiden was excavated by the 321 Appendix II: PUBID 538, 745. 322 Nobles 2016. ‘complete’ concentration of Mesolithic lithics commercial frm (Archol) of the Faculty of 323 Fokkens & Jansen 2002; Fokkens, was excavated within an area measuring Archaeology at the University of Leiden, Steﬀens & Van As 2016, 59. 324 See Fokkens & Jansen (2002) for 13 × 10 m.320 Although steps were taken to extract whilst the scientifc supervision lay with reliability classes. information about formation processes on the Prof. L.P. Louwe Kooijmans; M. Wansleeben of 325 Noens 2011; Wansleeben & Laan 2012; Wansleeben & Louwe-Kooijmans 2006. basis of the vertical distribution of fnds, that university performed the spatial analysis. 131 —

Fig 4.10 Horizontal density distribution of ﬂint at Roterdam-Groenenhagen. The circles indicate ‘clusters’ within a continuous, yet only partially investigated, area (from appendix II: PUBID 468). 132 —

Dronten-N23 was excavated by a consortium of landscape, the greater the chance for disturbance commercial frms, one of which was Archol; the (e.g. cleaning), of accumulation of remains, and spatial analysis was conducted in close successive mixing and blurring of paterns. cooperation with M. Wansleeben. Finally, the The NOaA 1.0 therefore explicitly mentions the analysis of Leeuwarden-Hempens was importance of sites with high spatial resolution conducted by G. Noens, as part of his PhD work and integrity.329 However, as outlined in at the Archaeological Institute of Ghent section 4.3, we fnd ourselves in the paradox that University (Belgium). This situation is in fact most of the efort is put into the investigation of comparable to that surrounding the analysis of sites with high fnds densities, which ofen Hoge Vaart-A27, referred to above. The represent palimpsest contexts with restricted excavation and basic analysis of this site was spatial resolution and integrity. conducted in a development-led seting on the We will concentrate on sites that were eve of the implementation of the Malta subjected to more extensive spatial analysis to legislation; the report was published in 2001.326 investigate the behavioural context of the fnds, The subsequent in-depth 2D and 3D spatial namely, the Mesolithic sites of Dronten-N23, analysis was, however, conducted in the context Leeuwarden-Hempens, Ede-Kernhem, Epse- of PhD research.327 Olthof and Zutphen-Looërenk. These sites To sum up, the degree to which spatially primarily consist of distributions of knapped fint referred data is analysed relates partly to the and other types of rock. If present, features on expected representativeness of the excavated Mesolithic sites represent pit hearths. With the ‘sample window’ for what is considered to be exception of Leeuwarden-Hempens, vertical ‘the site’. It also partly relates to money and, density distributions were evaluated by the perhaps more importantly, to the involvement excavators in order to assess the possibility of of academic staf or parallel embedding in PhD stratigraphic diferentiation, prior to the analysis projects. At this point, then, we come to the of horizontal spatial paterning. In all cases this question of what development-led archaeology led to the conclusion that horizontal fnds has brought that may enable us to reach a new distributions could be analysed integrally, understanding of spatial behaviour of early without taking any notice of vertical stratigraphy. prehistoric people. Below, we will further Investigation of horizontal paterns typically explore to what extent insights have been consists of the presentation of density gained concerning the research questions listed distribution maps for various fnds categories at the start of this section. (e.g. fint, stone, bone, hazelnut, charcoal) and typological or technological categories (e.g. points, scrapers, blades, cores). The ploted data 4.3.2 Understanding ﬁnds scaters: concern either counts or weight of some activity areas? selected category per recording unit, which in all of the sites mentioned above consist of 50 × 50 cm squares. In the case of Dronten-N23, In a broad sense, an ‘activity area’ is an undefned additional plots are provided for groups of four swath of space within which one or more squares, resulting in a 1 × 1 m grid on the map individuals executed one or more tasks for some (fg. 4.11). Quantities are ploted according to a purpose. The identifcation of such an area is selected assemblage frequency distribution dependent on the archaeological traces recorded method, such as equal count, equal range, or and the spatial interrelationships of those traces. natural break, but it should be noted that none Highly infuenced by the ethno-archaeological of the reports mention which method was work of Binford,328 such identifcation assumes applied. The distribution plots obtained are that low-density fnds scaters, consisting of both normally taken at face value, without any further material remains and structural features, result transformation of data. In the case of from short-lived phases of human presence and Dronten-N23, however, moving average hence ofer the best chance for successful transformation was applied in order to obtain a 326 Hogestijn & Peeters 2001. 327 Merlo 2010; Peeters 2007. identifcation of activity areas. Of course, this is a smoother distribution surface, at least for some 328 Binford 1978. valid assumption; afer all, the longer the time broad categories of fnds (fg. 4.12).330 For a 329 Deeben et al. 2006, 31. 330 Wansleeben & Laan 2012, 107-108. span of human presence at some ‘spot’ in the comparable purpose, kriging interpolation was 133 —

Figure 4.11 Horizontal density distribution of flint at Dronten-N23. Top: density distribution in 1 × 1 m squares; botom: density distribution in 50 × 50 cm squares (from appendix II: PUBID 538). 134 — Fig 4.12 Horizontal density distributions in 50 × cm squares of various find categories (from lef to right: flint, hazelnut shell, charcoal, stone, bone) at Dronten-N23. The maps in the top row show the original frequency distributions; maps in botom row moving average distributions (from appendix II: P UBID 538).continuous, yet only partially investigated, area (from appendix II: PUBID 468). 135 —

Fig 4.13 Simpliﬁed model of patern formation resulting from random accumulations of ‘ﬂint events’ (red) and ‘nut events’ (blue) for Dronten-N23 (from appendix II: PUBID 538).

applied on the data from Leeuwarden- properties. The recognition of self-organised Hempens, albeit that here the technique was patern formation in fnds distributions is applied to a dense set of coring data.331 Kriging important because it provides clues to a beter interpolation was also applied to data from understanding of palimpsests.335 This, however, Knooppunt Hatemerbroek.332 requires in-depth analysis, such as has been It is evident that the application of such performed by means of fuzzy set and techniques as moving average and kriging ‘percolation’ analysis on data from the Early obscures details that are present in the original Neolithic site of Hoge Vaart-A27 (phase 3).336 data, and that these techniques are unsuitable So what about activity areas? Not for spatial analysis aimed at the identifcation of surprisingly, the ‘strict’ identifcation of discrete activity areas.333 In fact, it is not clear at all what activity areas remains problematic within these the added value of these exercises is meant to palimpsest sites. For Leeuwarden-Hempens, the be, other than obtaining ‘cleaner’ distribution inspection of distribution maps of various fnds maps. The actual evaluation of resulting maps categories and functional categories of continues to rely on visual inspection and microwear has not led to the identifcation of subsequent interpretation of observed paterns. isolated, activity-specifc clusters (fg. 4.14). For Dronten-N23, the conclusion is drawn that The overall distribution is difuse and ‘random’, the resulting paterns are the result of a growth which suggests an underlying process of gradual process that involved continuous accumulation accumulation and merging of initial paterns, as and merging of small individual clusters of has been suggested for Dronten-N23. A similar material remains, which themselves are the picture emerges from the distribution paterns at 331 Noens, 2011, 131-142. product of short-lived activities or ‘events’ Epse-Olthof (fg. 4.15), where three merged sub- 332 Laan 2011; Wansleeben et al. 2011. 333 Nobles 2016. (fg. 4.13).334 It has been suggested that such clusters of comparable typological and functional 334 Wansleeben & Laan 2012, 109-112. paterns have fractal properties, but this is not composition were recorded.337 Zutphen-Looërenk 335 Peeters 2007, 167. 336 Peeters 2007, 132-169. supported by any further analysis of spatial consists of a smaller scater within which two 337 Peeters, Verneau & Admiraal 2015. 136 —

Fig 4.14 Horizontal density distributions of ﬂint at Leeuwarden-Hempens (from Noens 2011). A: frequency counts; B: interpolated densities (kriging; interval 1 artefact); C: subdivision into ‘concentrations’; D: distribution of artefacts studies for use-wear analysis; E: distribution of artefacts showing use-wear traces; F: distribution of artefacts showing traces of hide processing (black dots: studied by V. Beugnier; red dots: studied by J. Schreurs; grey dots: studied for use wear). 137 —

Fig 4.15a Horizontal density distribution of ﬂint at Epse-Olthof (from appendix II: PUBID 733). 138 —

Fig 4.15b Horizontal density distribution of ﬂint at Zutphen-Looërenk (from appendix II: PUBID 733). 139 —

sub-clusters are visible, which, again, are not this postulated hearth is evidence for a fint- fundamentally diferent in composition; scrapers knapping location (debitage); cores and tend to be somewhat more spatially rejuvenation fakes were found at several concentrated compared with backed bladelets metres’ distance. Tools are, however, more and microlithic points.338 Again, we seem to be evenly distributed, although points tend to looking at a compound distribution of material cluster close to the workshop. Taken together, corresponding to multiple activity phases. this patern suggests that there is spatial Within the various sections excavated in diferentiation that can be understood in terms detail at Dronten-N23, several ‘concentrations’ of variable activities (e.g. fint knapping, point were delimited within an otherwise continuous production).341 Ede-Kernhem concentration 5 is distribution of fnds.339 Diferences in assemblage smaller in size and shows a diferent picture composition are reported with regard to (fg. 4.17). Three possible hearths and three or microlith types, and are interpreted in four fint-knapping locations have been chronological terms. Broadly speaking, however, postulated. The distribution map of burnt fint two groups of assemblages were distinguished. counts,342 however, shows that potential One is dominated by high points with an positioning of the hearths is disputable. addition of only a few scrapers; smaller ‘satellite Nonetheless, there seems to be a spatial concentrations’ may represent special activity relation, with higher frequencies of debitage areas. The other is characterised by more or less indicating knapping locations. Distributions of equal representation of points, backed bladelets other artefact categories seem to demonstrate and scrapers, and other retouched tools. It is comparable tendencies to Ede-Kernhem possible that such diferences refect functional concentration 3. The lack of clear clustering of variability amongst campsites, at least under the tool types leads to the conclusion that activity assumption that the two assemblage groups areas cannot be discerned. It is suggested that refect behavioural contexts that were separated the hearths did not function simultaneously. in time and space. At the smaller intra- Instead, each hearth and associated knapping concentration scale – the campsite – no further location would represent a single, short-lived functional spatial diferentiation has been period of activity, which was primarily focussed reported, which may indicate the absence of on hunting (maintenance of hunting gear?); diferentiated spatial zonation within such scrapers are lacking in the assemblage.343 Hence, concentrations. the overall distribution represents a small-scale Even in cases where smaller spatial palimpsest of events. Yet other concentrations distributions have been recorded, it appears at Ede-Kernhem (concentrations 7a, 7b) have difcult to distinguish well-defned activity areas been interpreted as base camps due to the at the intra-site level. Ede-Kernhem broad range of tool categories and quantities of concentration 3, for instance, covers almost debitage, which are suggested to refect longer 338 Verneau-Peeters & Fermin 2007. 339 Niekus et al. 2012. 33 m2. The low average density of fnds (fint) time spans of diverse activity. Variable degrees 340 Devriendt 2015, 163-178. suggests rather short-lived activity.340 The of clustering of tool categories may indicate 341 The patern observed for Ede-Kernhem concentration 3 is largely comparable to location of a hearth is vaguely indicated by more or less well-defned activity areas, but the that of the Hoge Vaart-A27 northern clustered burnt fint in association with charred distribution maps provided remain difcult to concentration (Peeters 2007). 342 Devriendt 2015, 182, ﬁg. 9.14. fragments of hazelnut shell (fg. 4.16). Next to read and interpret due to format and layout. 343 Devriendt 2015, 185. 140 —

Fig 4.16 Horizontal density distributions of ﬂint (grey shading) at Ede-Kernhem within the investigated area, and the relative importance of various microlith types (from appendix II: PUBID 745). 141 —

Fig 4.16 Continued. 142 —

Fig 4.17 Horizontal density distributions of ﬂint at Ede-Kernhem for concentrations 5, 7a and 7b. Green squares represent postulated knapping locations; red and orange squares represent primary and secondary hearth locations (from appendix II: PUBID 745). 143 —

Another small site dating to the Mesolithic understanding of what such paterns could has been excavated at Groningen-Meerstad.344 actually mean in terms of behaviour. In the case The fint assemblage is homogenous and seems of the Mesolithic-Neolithic site of Groningen- atributable to a single chronological phase Europapark, categories of fint were reorganised (Middle Mesolithic). It is probable that the in terms of so-called product groups345 in order to overall distribution represents a short-lived obtain insight into technologically diferentiated period of activity. Distribution maps of various space (fg. 4.19).346 Although this did not result in fnds categories (fg. 4.18), however, do not the identifcation of activity areas, it helped to provide evidence for clearly defned activity obtain a beter understanding of the processes areas, in that distributions are rather difuse. It is that led to the formation of this palimpsest fnds not clear whether this refects an absence of distribution. The previously mentioned analysis spatially separated functional zones, or whether of Hoge Vaart-A27 provides another example of the original patern is distorted due to post- how insight into the behaviour-based formation depositional disturbance. of a palimpsest can be approached, and how this From the above examples, it could be can lead to conclusions about long-term activity concluded that the identifcation of activity areas paterns.347 has been only minimally successful, be it in small, The identifcation of activity areas in the single-phase concentrations of fnds or in small narrow sense, whilst not impossible, is only to extended palimpsest distributions. But realistic in contexts that were immediately dismissing such atempts to identify activity covered over with sediment, under calm areas may well be too easy. First, we need to take conditions, and subject to at most minimal post- a critical look at our expectations. The concept of depositional disturbance. A situation like this activity areas is rooted in the premise that may apply to Well-Aijen.348 The majority of activities in ‘ethnographical time’ are bound to a development-led excavations, however, do not restricted space and functionally focussed. This meet these criteria. For this reason it seems to snapshot perspective does not correspond to the be a beter choice in many situations to focus spatial dynamics of short-lived activities, which spatial analysis of fnds scaters on the are related to a particular context of behaviour, identifcation of paterns that can inform us namely, a set of tasks. Tasks may have been about underlying processes of site formation. conducted not only within culturally structured Afer all, if the majority of sites investigated space, thus creating a blueprint of spatial represent palimpsests – even small, ‘single- organisation, but also in open-air setings, where phase’ sites – we have to develop ways to come one can quickly change locations. In outdoor to a beter understanding of how the spatial setings, the material output of related tasks can paterns therein emerged, and what these result in seemingly unstructured or random paterns tell us about human behaviour or post- spatial paterns. Second, it should be noted that depositional processes. From a broad landscape none of the development-led projects have perspective, the emergence of extended included multivariate spatial analysis in their palimpsests due to long-term use of zones for 344 research. As we note above, we are dealing with various activities is signifcant; determining the Arnoldussen et al. 2012. 345 Collins 1975. one-by-one and category-by-category visual band-width of activities and chronological shifs 346 Fens & Mendelts 2015. 347 inspection of distribution maps. It is only can be highly informative. Peeters 2007. 348 Appendix II: PUBID 592, 725. The ﬁnal occasionally that an efort is made to come to an report is not yet available. 144 — Fig 4.18 Horizontal density distributions of various tool categories at Groningen-Meerstad (from appendix II: PUBID 255). 145 —

Fig 4.19 Flint distribution paterns at Groningen-Europapark organised according to the technological product groups as deﬁned by Collins (1975). Top lef: product group IIA; top right: product group IIB; botom lef: product group III; botom right: product group IV (from appendix II: PUBID 752). 146 —

here yielded evidence for their presence, neither 4.3.3 Tents, huts, houses and other built in the spatial distribution of fnds, nor in structures confgurations of anthropogenic features, such as postholes. Clearly, the identifcation of potential structures, such as tents or huts, Expectations about existence of remains of built within fnds scaters requires in-depth spatial structures are generally higher for the younger analysis.350 However, as we have seen in the periods of prehistory than for older periods (say, previous section, this is notoriously hunter-gatherer contexts). Prehistoric hunter- problematic. A late Early Mesolithic low-density gatherers are equated with tents and ephemeral scater at Zutphen-Ooyerhoek provides a rare huts. Sites dated to the Neolithic are all too example of a postulated tent structure ofen equated with permanent setlements and (fg. 4.20).351 Anthropogenic features other than house building, and the search for built hearths have hardly ever been recorded at structures has become an end in itself. hunter-gatherer sites. Features that could Developing an understanding of how space is potentially be interpreted as postholes are ofen used in a social context generally seems to be considered with extreme caution in the feld, overlooked. Of course, the identifcation of as postholes are equated with more permanent house plans is straight forward in some cases, structures, and this doesn’t match the general but ofen there are plenty of difculties. The idea of hunter-gatherer lifeways. However, identifcation of built structures is largely viewed evidence for huts with sunken foors has been as a mater for visual inspection and reference to found at two of Hardinxveld-Giessendam sites, 349 Arnoldussen 2008; Fokkens et al. 2016; Lange et al. 2014; Waterbolk 2009. house plan typologies and preconceived namely, Polderweg and De Bruin (fg. 4.21), 350 Tools such as the rings-and-sectors layout:349 regular confgurations of features are dated to the Late Mesolithic and Early method (Stapert 1992) are available for analysis of point-density data, but the matched to mental templates of what houses Neolithic.352 Likewise, the numerous huts with application to grid-cell data is not should look like in particular time periods. As sunken foors from southern Scandinavia show without problems. 351 Verneau & Peeters 2001, 40. mentioned in section 4.3.1, in those cases where that such structures are no exception. Plans of 352 Hamburg & Louwe Kooijmans 2001; plans of dwelling structures are less obvious and huts with ‘heavy’ posts have been found at Louwe Kooijmans & Nokkert 2001. 353 Karsten & Knarrström 2003; Peeters cannot easily be identifed, problems arise. In several sites in Scandinavia (Tågerup) and the 2007; Van Haaff et al. 1988; Waddington our view, there is no reason to disregard such United Kingdom (Howick, East-Barns, 2007; Waddington & Bonsall 2015; Woodman 1985. ‘unconfrmed’ structures. Certainly in the Broom Hill, Low Hauxley, Mount Sandel), 354 Geerts et al. 2016. context of early prehistory, we know very litle, and one possible example is known from the 355 Three Mesolithic hut plans have been postulated for the site of Westelbeers so every bit helps. Relationships between fnds Netherlands (Baarn-Drie Eiken).353 It is very well (Dijkstra et al. 2016), excavated in the scaters and features are ofen not investigated, possible that the lack of evidence for built 1970s and 1980s by amateur archaeologists under the supervision of even though researchers cannot know a priori structures in many projects is the result of Dr. R.R. Newell (of the Biologisch- what may be extracted from such analysis. prejudice, excavation methods – excavation in Archaeologisch Instituut, University of Groningen, now the Groningen Below we will provide an overview of the small squares (a much used method in the Institute of Archaeology). Two consist contribution made through development-led Netherlands) makes the identifcation of of semi-circular configurations of ‘postholes’; the diameter of the projects. features difcult – and perhaps an abundance of configurations measures c. 5-6 m, caution. Interestingly, two recent which conforms to other north-west European examples. The third one is Late Palaeolithic and Mesolithic development-led projects – Soest and Kampen- more elongated, and measures Despite the fact that in all cases the possibility of Reevediep354 – have yielded convincing evidence c. 4 × 7 m. Although it may certainly involve hut plans, the information built structures was taken in consideration for Mesolithic huts with heavy posts (fg. 4.22); provided about the features makes it – sometimes explicitly – none of the both are closely comparable to Baarn-Drie Eiken difficult to be certain about this interpretation. development-led projects under consideration and the Scandinavian and British examples.355 147 —

Fig 4.20 A Mesolithic scater of ﬂint at Zutphen Ooyerhoek. The average weight distribution of artefacts per 1 × 1 m square suggests a ‘centrifugal’ eﬀect, which can be explained in terms of Binford’s ‘drop-and-toss zone’ model (from Verneau & Peeters 2001). 148 —

Fig 4.21 Mesolithic features at Hardinxveld-Giessendam De Bruin (top) and Hardinxveld-Giessendam Polderweg (botom) which have been interpreted as sunken-ﬂoor dwelling pits (from Louwe Kooijmans & Nokkert 2001; Hamburg & Louwe-Kooijmans 2001).

Fig 4.22 Mesolithic dwelling structure at Soest. The darker grey features are postholes visible in the C-horizon; the brown illuviation zone within the circular arrangement of postholes appears to represent the sunken ﬂoor of the dwelling (photo courtesy of BAAC). 149 —

Linear Bandkeramik However, this seems to be due to the restricted The highly standardised house plans of the spatial extent of the investigated areas and Linear Bandkeramik (LBK), which are easily hence justifable. This is not the case for recognised even when only partially represented Maastricht-Cannerberg, where large-scale in an excavated area, have been reported at development-led feldwork on an LBK Elsloo-Riviusstraat, Stein-Heidekampweg and setlement permited in-depth analysis. Sweikhuizen-Geverik.356 Although the layout of Although the report falls outside the selection the house structures is well known, litle is criteria of our study due to its publication date known about the internal use of space in LBK (2016),359 it is worthwhile to briefy refer to the houses in general, because the habitation obtained insights. First, it appears that the layers (i.e. occupation surfaces) are not orientation of houses at this location partly preserved, due to erosion. Indirect evidence deviates from the NW–SE orientation that was about the activities conducted in and around considered to be the norm for the LBK. At houses comes from pits adjacent to houses. It is Maastricht-Cannerberg, the main orientation is assumed that pits adjacent to a house are related WNW–ESE to W–E (fg. 4.25).360 Apparently, there to the extraction of loam during the building of is more variability than hitherto expected. that house, and that these pits were Second, the application of a 20-phase relative subsequently used for disposal of waste from the typo-chronology of ceramics permited household.357 Hence, data on the contents of pits researchers to analyse spatio-temporal in relationship to house plans can be informative. dynamics within the setlement area.361 The The extent to which this is the case is variable. results suggest relatively contemporaneous Sweikhuizen-Geverik, for instance, involved a occupation for a limited number of houses, as narrow pipeline trench, which can ofer only well as the rebuilding of houses next to their sparse spatial information; even though the data predecessors, leading to the emergence of are provided, no conclusions can be drawn. In clusters of house plans.362 Typically, house plans contrast, Elsloo-Riviusstraat occupies a larger do not overlap, which suggests that the area, which contains several LBK house plans, locations of the predecessors were still visible some of which are associated with pits (fg. 4.23); when their successors were being constructed. information about the contents is provided at This possibly indicates a rather short-lived use of the level of individual features, but houses, not exceeding a single generation of 356 Appendix II: PUBID 194, 528, 537. interpretations remain restricted to generalised occupants. The relatively high chronological 357 Coudart 1998, 73; Stäuble 1997, 23. 358 Appendix II: PUBID 53. terms, such as ‘loam pit’ and ‘silo’. At Beek- resolution furthermore permited to reconsider 359 Van Wijk 2016. It is important to note Kerkeveld, no house plans were found, but a the issue of house yards. In contrast to what has that the extensive analysis and report are the result of the dedication of the complex of ditches was interpreted as part of an long been assumed, the number of pits per yard researchers involved; much more time LBK enclosure (known by the German term appears variable; however, most house yards has been invested than was contractually assigned. Erdwerk; fg. 4.24); it seems probable that the area have silo pits.363 A further insight involves the use 360 Van Wijk 2016, 101-102. investigated concerns the peripheral zone of a of fint varieties, which originate from various 361 This typo-chronology was developed by Van de Velde (2014) on the basis of setlement.358 The enclosure is the frst to be sources in the region. At some moment in time, statistical analysis of LBK potery atributed to the LBK in the Netherlands, corresponding to ceramic phase 14, a local assemblages. In total, 20 phases were distinguished, as opposed to the seven although some reservations must be made variety of Lanaye fint became dominant,364 phases deﬁned by Modderman (1970) concerning its dating, since litle LBK material and pits with abundant blade production that formed the chronological framework for discussion LBK was found in the ditch fll. Comparable structures waste – amongst which many blade cores – developments until recently. Van de are also known from other Neolithic contexts, indicate specialist activity. As tools of this fint Velde’s phasing is of a strictly statistical nature; it provides no basis for such as Michelsberg; but in the absence of variety are lacking in the setlement itself, it is establishing an absolute chronology in 14 Neolithic fnds other than LBK, its atribution to suggested that this production functioned connection with C-dates, and the duration of each phase is therefore the LBK is, in our opinion, defensible. within a regional exchange system.365 Within a unknown. Generally speaking, the few development- wider regional context, variability in the use of 362 Van Wijk 2016, 102-109; Van Wijk & Van de Velde 2016, 130-139. led projects that have reported on LBK fint varieties between and within setlements 363 Van Wijk 2016, 104-125. structures remain somewhat descriptive an points to the socially bound exploitation of raw 364 This refers to the so-called LC1 variety (Knippenberg 2016, 181). ‘impressionistic’ in their interpretation. material sources. 365 Knippenberg 2016, 193-195. 150 —

Fig 4.23 LBK house plans at Elsloo-Riviusstraat (from appendix II: PUBID 194). The maps provide spatial information about the degree of preservation and finds composition of features (red: potery; green: bone; blue: flint; purple: stone). The photos show the preservation classes of postholes, which basically reflect visibility due to bleaching (top: good; middle: mediocre; botom: bad). 151 —

Figure 4.24 Possible LBK enclosure at Beek-Kerkeveld (from appendix II: PUBID 53). The LBK features are indicated in brown; orange-coloured features are of medieval age. 152 — Fig 4.25 LBK structures and zones of habitation at Cannerberg, and a schematic representation of shifing house yards (from Van Wijk 2016). Fig 4.25 LBK structures and zones of habitation at Cannerberg, a schematic representation shifing house yards (from Van 153 —

Funnel Beaker Culture Furthermore, a 5 m long ditch with closely set Most notable are the built structures atributed posts within the enclosure possibly represents to the Funnel Beaker Culture (known by its the remains of an earlier phase; two 14C-dates German abbreviation, TRB) that were discovered confrm an atribution to the TRB period.368 The at Knooppunt Hatemerbroek. At this extended enclosure found at Knooppunt Hatemerbroek is site complex, evidence for a palisade was found the third in the Netherlands. The other two were in section 7, within the excavated trajectory of found at Anloo (province of Drenthe) and the Hanzelijn railway. The palisade consists of a Uddelermeer (province of Gelderland).369 It is closely set curvilinear alignment of posts in a assumed that these enclosures had a defensive narrow ditch (fg. 4.26), and has been interpreted or ritual function. as part of a causewayed enclosure.366 Within the About 1600 m south-west of the enclosure, excavated area, the palisade could be followed excavations south of Knooppunt Hatemerbroek over a distance of 73 m, but it is clear that the yielded evidence for TRB domestic activities, structure must have been far larger, as this including, potentially, the remains of built distance represents only about one ffh of a structures (houses?).370 Within a 40 × 50 m wide circle. Although ‘complete’ enclosures are never zone, 65 postholes were atributed to a entirely circular, this still means the enclosure particular phase of activity on the basis of their would have had a diameter of approximately appearance and soil characteristics. In some 100 m. The atribution of the palisade to the TRB postholes, TRB potery fragments were found, is tentatively based on the appearance and soil including, in one case, a large part of a pot (rim, characteristics of the features themselves, on neck and body) and a part of a baking plate, comparison with features dating to other which could be interpreted as a ‘construction periods (Mesolithic, Late Neolithic, Bronze Age), ofering’ (see section 3.3.1). The excavators and on their stratigraphical context. Despite the identifed two square or rectangular structures limited quantities of potery fragments securely on the basis of a limited number of postholes atributed to the TRB, these form the majority of (fg. 4.27). These possibly represent houses fnds in this section of the excavated area; comparable to Hunte 1 (north-west Germany), material dating to other periods is, however, although the report recognises that this also present. This makes an atribution of the interpretation cannot be convincingly palisade to the TRB a possibility, although there underpinned due to ‘missing’ postholes. is no direct support from the spatial distribution The spatial distribution of the fnds does not of fnds.367 Several postholes within the provide further support either, despite the 366 Knippenberg & Hamburg 2011, 162-164. 367 Wansleeben et al. 2011, 106-111. enclosure are reported to have comparable relatively good preservation of the fnds layer 368 GrA-39676: 4565 ± 30 BP; GrA-39755: characteristics and are potentially part of a that covers the features. 4560 ± 30 BP. 369 Harsema 1982; Waterbolk 1960. house, although this cannot be confrmed. 370 Knippenberg & Hamburg 2011, 115-124. 154 —

Fig 4.26 Part of a TRB enclosure/palisade (bright green) and a section showing a series of closely set posts at Knooppunt Hatemerbroek (from appendix II: PUBID 259). 155 —

Fig 4.27 Zone with postulated (in green) structures dating to the TRB; linear arrangements of stakes and a water pit (in blue) date to the Bronze Age (from appendix II: PUBID 260). 156 —

Although it is clear that TRB occupation was The recent discovery of a TRB site near well represented in the landscape near Dalfsen, only some 15 km from Knooppunt Knooppunt Hatemerbroek, it is also clear from Hatemerbroek, provides an illustrative example the reports that insight into built structures of what can be expected in the context of remains limited. There is no reason to exclude development-led research – in this case an the atribution of the palisade and traces of extended cemetery consisting of 137 inhumation domestic activity to the TRB period, even though graves and several cremation graves (fg. 4.28) the supporting evidence remains scanty. And the that is the biggest TRB cemetery known in NW fnds near Knooppunt Hatemerbroek make it Europe thus far.372 Right next to the cemetery, clear that we have to remain aware that there is a house plan was found that can also be the potential for TRB occupation in this part of atributed to the TRB. The pots from the graves the country. Evidence for TRB habitation is mostly belong to a single phase – Brindley 371 Ten Anscher 2012; Clevis & Verlinde known from the Noordoostpolder, west of phase 5, maybe some phase 6/7373 – which spans 1991. Knooppunt Hatemerbroek, as well as from a restricted period of two centuries (2900-2700 372 Van de Velde & Bouma 2016. 373 Brindley 1986. Zwolle-Itersummerbroek, to the north-east.371 cal BC). This provides a unique opportunity to

Fig 4.28 TRB ﬂat grave cemetery (light grey), monumental structure (dark grey) and house plan (black) at Dalfsen (from Van der Velde & Bouma 2016). 157 —

study potery traditions at the level of what are Group), and Hazerswoude-Rijndijk and perhaps single but successive households, Hellevoetsluis-Ossenhoek (both Vlaardingen assuming that the people buried in the cemetery Group). The cultural afliation and function of originate from the same setlement and the possible structure from Hof van Limburg are represent several generations of families. far from clear (fg. 4.29).374 Here, two alignments The discovery at Dalfsen in particular urges us to of features, interpreted as postholes within a think diferently about the TRB. Until recently, large pit, form a roughly rectangular our picture of the TRB was largely based on confguration measuring 2.7 × 1.5 m. The nature megalithic graves in the north-eastern part of of this possible structure is unknown due to the the Netherlands; clearly, there is a potential that absence of fnds in the features themselves. exceeds all expectations, given the fact that in Even its atribution to the Stein Group is the Netherlands each square centimetre is uncertain: it is based on the appearance and soil assumed to have been disturbed. characteristics of the features and on the absence of Iron Age potery, which is present in Stein Group, Hazendonk Group and other parts of the excavated area. Vlaardingen Group Middle Neolithic built structures have been reported from Hof van Limburg (Stein Group), Rijswijk-Ypenburg and Schipluiden (Hazendonk

Fig 4.29 A possible Middle Neolithic (Stein Group?) structure at Hof van Limburg (from appendix II: PUBID 219). 374 Van Hoof et al. 2013, 83-84. 158 —

The cultural context for Rijswijk-Ypenburg choices with regard to sampling for specialist and Schipluiden, however, is clear: here we are analysis – prevented a more analytical approach. defnitely dealing with Hazendonk setlements.375 And in contrast to Schipluiden, where moving At Rijswijk-Ypenburg, three house plans and average interpolation was applied, at Rijswijk- three other built structures were identifed Ypenburg it was decided not to mitigate for (fg. 4.30).376 The house plans are two-aisled and missing data through interpolated values, in rectangular or oval in outline, measuring c. 9.7- order to prevent the loss of potentially 8 m in length and 4.5-4 m in width. The other meaningful information within the actually reported structures are simple rectangular yet recorded distributions (see section 4.3.2). irregular confgurations of postholes. The It remains uncertain whether house atribution of postholes to some particular structures are also present at Schipluiden. As the structure was based on several variables: excavators state, the only spatial structures that posthole diameter, depth, shape and infll. Visual can be identifed within the distribution of inspection of distribution paterns of various postholes consist of “single rows of relatively large 375 Appendix II: PUBID 732. fnds categories was used to identify functional postholes set at irregular distances relative to one 376 Houkes & Bruning 2008. zones relative to the structures (fg. 4.31). It is another”.377 The search for house plans 377 Hamburg & Louwe Kooijmans 2006, 62. 378 Raemaekers et al. 1997. important to note that the lack of continuous comparable to those of, for instance, 379 Koot et al. 2008. data in fnds distributions – due to diferences in Wateringen,378 Ypenburg,379 Haamstede- 380 Verhart 1992. 381 Glasbergen et al. 1967. fnds collection strategies, and subsequent Brabers380 and Vlaardingen381 (fg. 4.32), remains

Fig 4.30 House plans of the Hazendonk Group from Rijswijk-Ypenburg (from Houkes & Bruning 2008). Black: features which meet the selection criteria and belong to the structure; dark grey: features which do not meet the selection criteria but belong to the structure; light grey: features which meet the selection criteria but do not belong to the structure; white: other features. 159 —

Fig 4.31 Distribution paterns of various find categories in a zone defined as homestead A of the Hazendonk Group at Rijswijk Ypenburg (from Dorenbos 2008). Top lef: charcoal by number per m2; top right: flint by number per m2; botom lef: potery fragments by weight per m2; botom right: bone by weight per m2; red oval: homestead; dashed black ovals: clusters of postholes; solid black shapes: structures. 160 —

Fig 4.32 House plans from Heemstede-Brabers and Wateringen 4. 161 —

unsuccessful. It is very well possible, however, is noted for Hazerswoude-Rijndijk, where that houses were frequently rebuilt and shifed feldwork was restricted to trenches.386 Several to higher parts of the dune as groundwater alignments of postholes have been documented, tables rose.382 It is assumed that a rectangular and these are believed to be part of a house ditch with rounded corners, found on top of the plan. The same area also produced several dune, delimited the fnal house or hut structure. hearths in several successive layers. Although However, there is no supporting evidence from not unequivocal, it is certainly possible that the the distribution of fnds, as the highest part of features are part of a larger built structure that the dune had been subject to erosion. The may have been rebuilt or repaired on one or emphasis on concentrations (read: high more occasions. densities) of fnds comes with a co-notation of From the above discussion, it is clear that importance and can lead to straight forward development-led archaeology has yielded assumptions about spatial association of fnds evidence for built structures on several and features such as postholes. In the case of occasions. Yet, it is also clear that the Schipluiden, four clusters of postholes combined interpretive possibilities are ofen hampered by with concentrations of fnds were interpreted as the restricted scale of feldwork, by the shape of primary activity areas and possible ‘farmyards’.383 trenches, by site erosion, or by the complexity of The next step is that farmyards are equated with the archaeological data itself. In order to reach households;384 in other words, the built structure sound conclusions regarding the presence of, for and its surroundings are given social meaning. instance, house or hut structures, we argue for This is not to say that farmyards don’t have deeper and integrated analysis of spatial data. social meaning, but the problem is that the Typically, structures and features are reported nature of the built structures themselves has not separately from one another, and are barely been established. used to efect an integrated synthetic analysis of In contrast to the problematic identifcation various categories of information (fnds). At the of house plans, Schipluiden has yielded same time, however, we acknowledge that the undisputable remains of multiple fences cases mentioned above show that there is much surrounding the dune. The fences consist of potential, certainly at sites where well preserved regularly spaced sets of posts (fg. 4.33). Another fnds layers (‘cultural layers’) are likely to cover palisade or fence has been reported from structural features. In anticipation of such Hellevoetsluis-Ossenhoek (fg. 4.34).385 Here, the situations, high-resolution spatial recording of structure consists of closely spaced single posts fnds categories from such layers and the or stakes. The alignment runs parallel to ard underlying features is a prerequisite for an marks. It is highly possible that the fence analysis of both types of data that is focussed on delimited a feld and was primarily meant to the identifcation of built structures. In other protect crops. Postholes for heavier posts, some situations, where the identifcation of built of them with wood preserved within them, are structures is unproblematic but cultural layers also present, and may well belong to house are lacking, meticulous analysis of fnds and data 382 Hamburg & Louwe Kooijmans 2006, 62. 383 Wansleeben & Louwe Kooijmans, 2006, structures. However, because the research was by means of ‘traditional’ approaches can actually 72. limited to several test trenches, there is not lead to new insights, as is shown in the 384 Wansleeben & Louwe Kooijmans, 2006, 86. enough spatial information to isolate any Maastricht-Cannerberg example. 385 Van Hoof 2009, 59. structure with certainty. A comparable constraint 386 Diependaele 2010. 162 —

Fig 4.33 Fences are clearly visible around the dune at Schipluiden. Wells are mainly concentrated along the northwestern and western slopes of the dune (from appendix II: PUBID 732). 163 —

Fig 4.34 Palisade or fence at Hellevoetsluis-Ossenhoek consisting of closely set posts (from appendix II: PUBID 272).

165 5 People and environment —

subsequent datasets. Afer all, information 5.1 Introduction about botanic and faunal aspects of the environment largely depends on the preservation of such remains. In view of the The relationship between people and variable geogenetic evolution of the Netherlands environment is a recurrent theme in at the national scale, taphonomic processes archaeology. Although environment is a rather create major bias owing to the large scale coarse concept, the term is generally applied to development of wetland conditions, or the the entire of biotic and a-biotic features in a absence thereof. Furthermore, it should be mostly vaguely defned or undefned stretch of borne in mind that the interpretation of land or space. Hence, the environment is environmental data at the local scale requires a composed of plants, animals, rocks, land good understanding of the interrelationships of surface, water, and according to some scholars, environmental processes at multiple spatial and even scent and sound.387 If we stick to humans, it temporal scales. This combination of factors is what surrounds an individual at some place in makes any synthetic analysis concerning people- the landscape, and can also comprise other environment relationships in the context of human individuals with whom there will be development-led research difcult. some degree of interaction – a socio-cultural It is therefore no surprise that the subject is environment. Generally speaking, however, the treated in a descriptive fashion in the great archaeological use of the concept of majority of reports. Relatively small-scale feld environment is strongly connected with the investigations and feldwork at sites with poor or non-human dimensions. And ofen, if not most limited preservation of organic components of the time, reference is made to the leave litle room for in-depth analysis and environment as a guiding or even deterministic interpretation, and this is clearly the case in most framework for subsistence and dwelling; this is of the development-led projects. Indeed, our also the case in the NOaA 1.0. Of course, the analysis of reports felt much like a survey of behaviour of early prehistoric people cannot be arguments: to what extent are conclusions set apart from the environment, but the substantiated by the actual data? Comparable environment cannot be set apart from human data are not always interpreted in the same behaviour either. Hence, the relationship manner at diferent sites. And many research between people and environment can be results may be relevant at the site level, extremely complex; indeed, it becomes more but barely at a more general level, if at all. appropriate to speak about cultural Hence, a small number of extensive reports landscapes.388 account for most of the noteworthy results and In this chapter we will focus on the novel ideas about the meaning of organic material interaction between humans and environment. categories for early prehistoric people. Most In the NOaA 1.0, the relationship between notably this concerns the large-scale excavations people and environment features in the chapters at Dronten-N23 (Mesolithic), Roterdam-Yangtze ‘Palaeogeography and genesis of the landscape’, Harbour (Mesolithic), Knooppunt Hatemerbroek ‘Archaeobotany’, ‘Archaeozoology and physical (Mesolithic) and Schipluiden (Neolithic).390 anthropology’ and ‘Early Prehistory’.389 These chapters cover a wide range of topics and aspects from various research angles and 5.2 Exploitation and use of organic address broad themes, such as landscape resources evolution, domestication, trade, subsistence and ritual. Such themes require analysis beyond the site level. The research questions are, however, Traditionally, the frst topics that come to the 387 See e.g. Ingold 2000. rather strongly site- and object-focussed, which archaeologist’s mind with regard to 388 Cavallo et al. (2006, 20) note that the distinction between research questions makes a synthetic analysis at multiple spatio- bioarchaeological studies are about food concerning the natural landscape, on temporal scales difcult. This is all the more true resources and the reconstruction of the the one hand, and the cultural landscape, on the other, is in fact when one considers the diversity of taphonomic environment. Of course, these are important artiﬁcial. processes that may have infuenced the topics for our understanding of how people in 389 Brinkkemper et al. 2005; Cavallo et al. 2006; Deeben 2005, 2006. composition and quality of the assemblages and the past exploited the landscape. 390 Appendix II: PUBID 260, 732, 740. 166 —

People’s stomachs needed to be flled, and the material are mostly expected in parts of the environment set the outer limits in terms of country where these are covered in metres of resource potential. Notably, these aspects have sediment. As a result, the potential availability of received, and still do receive, lots of atention. (new) information about exploitation and use of This being said, the study of other aspects of food resources is extremely fragmentary and human use of organic materials has a long biased. We will discuss some central topics and history as well. These other aspects are related problems below in reference to the numerous, and include the production and use results of development-led investigations. of tools (e.g. combs, awls and harpoons), building and construction material (e.g. fsh Animals in the landscape weirs, thatching, posts) and clothing (e.g. Animals played an important role in early leather, plant fbre clothes). In addition, social prehistoric economies. Sites with good and cosmological relationships between people preservation conditions have provided insight and organic material can be expressed through into various aspects of animals as a resource for specifc objects, such as pendants made out of food, hide, fur and bone/antler. Bone remains bone or plant material, or the symbolic or ritual inform us about which animal species were treatment of items, such as antler, bone or plant exploited, how they were butchered and which material (e.g. branches, fowers). Clearly this parts were used for which purpose(s). makes thematic discussions under such Nonetheless, very litle is known about these headings as ‘People and Material’ (chapter 3), aspects for major parts of early prehistory. This ‘People and Space’ (chapter 4) and ‘People and is particularly the case for the Late Palaeolithic to Environment’ (this chapter) artifcial. People, Early Neolithic. Results from development-led material, space and environment are research permit us to discuss some particular interconnected aspects of life. However, in order problems and peculiarities of the faunal record. to avoid one big tangle of topics being A remarkable fnd dating to the Early discussed, we will focus the present section on Mesolithic is an incomplete red deer skeleton food procurement and processing and on the found in the stream valley of the Tungelroyse use of wood for other purposes than the Beek near Mildert (fg. 5.1).391 The Early Mesolithic production of implements (this topic is age of this assemblage was established by presented in section 3.2.2). means of a 14C-date.392 The skeletal elements found represent the crown tines of the antlers, the skull, the mandible and the anterior half of 5.2.1 Subsistence and food processing the rib cage. A second individual is represented by a mandible. Despite the absence of evidence for butchery (cut marks, bone fractures) or traces Topics related to food economy or subsistence of projectile impact, the report suggests that the dominate much of the archaeological literature. remains may well represent a hunter-gatherer In many cases, however, the archaeological kill site. A fint blade found at 6 m distance may record is rather poor in this respect. Even when be related to this event, but in our opinion this is remains of potential food stufs have survived, it far from certain. The assemblage is difcult to is ofen not all that easy to distinguish between interpret, as it is not quite clear whether the food remains and natural background ‘noise’ or bones recovered constitute the entire original accumulation. Also, it is difcult to establish how assemblage, or, instead, represent an representative assemblages are of the original incomplete assemblage as the result of recovery range of food resources. Generally speaking, the by mechanical shovel.393 The antler remains have further one goes back in time, the rarer the fresh fractures, which indicate recent damage. reports that contain data on organic remains The total absence of posterior body parts and that can be connected with the food economy. limbs is nonetheless remarkable, and may well This is partly due to the overrepresentation of be the result of the removal of body parts by 391 Appendix II: PUBID 395. development-led research in parts of the humans. At least, there are no indications that 392 9650 ± 50 BP (lab.nr. not given in source publication). Netherlands where preservation conditions for this may have resulted from recent disturbance. 393 Poluted soil was removed by means of a organic material are poor (see chapter 2). Older But the situation is complex, given a total of 147 mechanical shovel. This soil could not be checked for the presence of bones. sites with the potential for preserved organic animal bones of various species found in a 167 —

Fig 5.1 The Mesolithic red deer remains from the valley of the Tungelroyse Beek. Top: the bones in anatomical position; botom lef: the skeletal remains during excavation; botom right: schematic representation of the bones uncovered (from appendix II: PUBID 395). 168 —

Fig 5.2 Find locations in the valley of the Tungelroyse Beek nearby the ﬁnd location of the red deer skeleton. On the right, two bones with clear cut marks are shown (from appendix II: PUBID 395). 169 —

nearby location (fg. 5.2). Together with other procurement environment. Together with, for artefacts – fint tools, potery – these bones instance, the Early Mesolithic context of point to long-term use of the stream valley from Zutphen-Ooyerhoek,398 the above examples the Mesolithic to Roman times. A chisel made illustrate that archaeological evidence for the out of an aurochs metatarsus, as well as a red exploitation of these environments is probably deer antler crown tine, have been 14C-dated to not as rare as one might expect. The problem is the Late Mesolithic.394 Bones belonging to frst of all that the resolution of phenomena domesticated animals indicate more recent related to activity in such areas is low, involving occupation as well. Several bones of red deer difusely spread fnds in and along former and horse – probably wild – show cut marks, gullies. Archaeological watching briefs in stream atesting to human activity. Despite the valleys are the only way to ensure, to some impossibility of confrming an anthropogenic degree, the discovery of such phenomena. origin for the partial skeleton, the various Equally, we must bear in mind that evidence observations in this zone of the stream valley for the procurement of animal food resources atest to long-term exploitation of animal outside stream valleys, or outside wetland resources. This picture fts observations in contexts in general, is extremely sparse and stream valleys elsewhere in the country, notably basically a consequence of preservation the Tjonger valley in the northern Netherlands, conditions. Not surprisingly, the great majority where Late Mesolithic aurochs remains with of reports make no or only marginal mention of butchery marks have been collected at several animal remains, as most sites investigated locations.395 connect to dryland setings. This does not A fossil stream valley near Emmen provides necessarily mean that no faunal remains were another glimpse at the long-term use of such found. Our survey of reports shows that fnds of landscape zones.396 One location (SCH1500) only small quantities of burnt (calcined) bone are not yielded an isolated fragment of a red deer antler, necessarily rare (table 5.1), but that litle but, interestingly, several wooden stakes atention is paid to them due to an assumed lack returned 14C-dates that indicate Late Mesolithic of (potential) information.399 We argue that this and Middle Neolithic activity.397 Another location negative atitude is unwarranted. (SCH2600) yielded two clusters of stakes. Their As can be concluded from table 5.1, the 14C-dates correspond to the Early Neolithic and possibilities for taxonomic identifcation are the Middle Neolithic. Yet another cluster of restricted, but not absent. Clearly this sets limits stakes at location (SCH1200) returned 14C-dates to the information potential of such corresponding to the Middle to Late Bronze Age. assemblages, but it is important to note that It is, however, not clear what activity the stakes both animals and humans may be represented. may relate to, but it is very well possible that In the particular case of Roterdam-Beverwaard, 394 6570 ± 40 BP (lab.nr. not given in source publication). they are connected with fshing. the identifcation of the human remains was 395 Prummel & Niekus 2011. Hunting and fshing activity is certainly to based on morphological characteristics. 396 Project Emmen-Olieveld Schoonebeek (Appendix II: PUBID 482). be expected in stream valley landscapes. Techniques such as bone histology and mass 397 Appendix II: PUBID 482. Running water atracts many diferent terrestrial spectrometry permit researchers to distinguish 398 Groenewoudt et al. 2001. 399 Lauwerier & Deeben 2011. animals and of course provides habitat for semi between human and animal material when 400 Histological analysis is applicable to and fully aquatic animals. In a way, stream identifcation based on morphology is not calcined bone remains (see e.g. Cuijpers 2006). Mass Spectrometry requires valleys provide a rather predictable food possible.400 collagen and thus unburnt bone. 170 —

Table 5.1 Sites with (exclusively) calcined bone.

Site Characterisation Context

Duiven-Ploen Zuid Calcination degree IV-V (650-800 °C); mostly Mesolithic; infll of a natural (?) feature. fragments of diaphysis, some epiphysis; no clear evidence for human material.

Finsterwolde Small number (29) of fragments; one fsh Mesolithic (?); difuse distribution. bone.

Fluitenberg Small number (15) of fragments; not Mesolithic; infll of pit hearths. analysed.

Hasselo Two fragments; not analysed. Mesolithic; infll of pit hearths.

Roterdam-Beverwaard Calcination degree IV-V (650-800 °C ); Mesolithic; most material from three pits human and animal remains; skeletal (graves); some remains from fnd layer. elements determined when possible; age indication; incidental gender assignment.

Emmen-Olieveld Six small fragments; not analysed. Mesolithic – Neolithic; from fnd layer. Schoonebeek

Dronten-N23 Heavily fragmented; one species assignment Mesolithic; difuse distribution in fnds layer, (Sus), further mostly large – medium size but slight clustering close to fint mammal; no evidence for human material. concentrations.

Roterdam-Yangtze Harbour Large assemblage (> 12,000) of fragments Mesolithic; material sampled from fnds (not only calcined); many mammal, bird, fsh layer. species identifed, as well as reptiles.

Epse-Olthof Some heavily fragmented remains; no Mesolithic (?); possibility of post- species assignment possible. depositional intrusion in pit hearths.

Hoogezand-Sappemeer Two fragments; one possibly rib or shoulder Mesolithic; from fnds layer. Vosholen blade; medium-size mammal. Further mention of very small fragments from charcoal samples from pit hearths: not collected.

Wijchen-Alverna Small number (23) of fragments; not Mesolithic; possibly from a surface hearth, analysed. associated with charred hazelnut shell and burnt fint.

Ede-Kernhem Small number (107) of fragments; not Mesolithic (?); partly from disturbed context, analysed. but maybe some original context.

Leeuwarden-Hempens Limited number (204) of fragments; one Mesolithic (only 4 fragments) and Middle species assignment (Ovis/Capra). Neolithic (TRB); Neolithic assemblage from a surface hearth.

Roterdam-De Zwanen High number (c. 450) fragments; some Early Neolithic (Swiferbant); collected from Rietpark species identifed (Sus; Lutra lutra) in addition fnds layer. to assignment to large, medium, small-sized mammal.

Emmeloord-Ens High number (c. 450) of fragments; some Neolithic; collected from fnd layer. fsh remains (Anguilla; Cyprinidae); Ovis/ Capra (dental elements).

Groningen-Helpermaar Small number (79) of fragments; some Middle Neolithic (TRB); most material species identifed (Sus; Bos; Ovis/Capra) in collected from fnds layer. addition to assignment to large and medium-sized mammal. 171 —

Fig 5.3 Stable isotope data of human remains (lef) in comparison with those of terrestrial mammals remains (right) from the North Sea (from Van der Plicht et al. 2016).

Apart from the distinction between human The information value of assemblages such and animal remains, it is also important to as the one from Roterdam-Yangtze Harbour is consider the potential of calcined bone, even if high when considered in a long-term regional heavily fragmented.401 For instance, fsh context. Hunter-gatherer activity at this site vertebrae, which are important for taxonomic spans a considerable part of the history of the identifcation, continue to be identifable, even submersion of Mesolithic Doggerland, i.e. the when calcined and fragmented; in fact, they southern North Sea. During the Early Mesolithic, ofen remain intact. Good examples are known the site was located far inland, on a high sand from both Mesolithic and Early Neolithic sites, dune in the Rhine-Meuse valley. Gradually, notably from Roterdam-Yangtze Harbour and however, the coastline moved eastward, and the Hoge Vaart-A27.402 Fragments of bones site became located in the Rhine-Meuse estuary, belonging to small mammals and birds also have and eventually on the coast. The faunal a fair chance of being identifed if represented in assemblage does, however, show only sparse an assemblage of calcined bone; due to their evidence of an increased importance of marine small size, they tend to fragment into fewer food resources. The activities came to an end pieces, and their taxonomically distinctive when the sand dune was inundated by the sea morphological characteristics ofen remain by 6300 cal BC. Stable isotope analyses of preserved. In the case of Roterdam-Yangtze Mesolithic human remains from the North Sea Harbour and Hoge Vaart-A27, for instance, also show restricted evidence for a marine calcined remains of weasel, squirrel and small component in the diet of people who inhabited songbirds provide examples of animals that are the gradually submerging land of Doggerland hardly ever found in bone assemblages. The (fg. 5.3).403 There is a clear aquatic signal, yet analysis of the assemblages from these sites has also one of freshwater food resources (mostly revealed a surprising range of species, because fsh, water plants?). The picture fts that of the the material received the atention required. In people who were buried on the sand dunes of the case of Roterdam-Yangtze Harbour, even Hardinxveld-Giessendam Polderweg and though the total number of bone fragments De Bruin, in the Rhine-Meuse estuary.404 identifed to the species level was not that high (17 for mammals, 84 for birds, 316 for fsh), the 401 Lauwerier & Deeben 2011. absolute number of species identifed was high 402 Zeiler & Brinkhuizen 2015. 403 Van der Plicht et al. 2016. (8 for mammals, 19 for birds, 14 for fsh). 404 Smits & Van der Plicht 2009. 172 —

Where it comes to our insights into the the estuaries and rivers, but also the back- exploitation of animals in the Mesolithic to Early barrier freshwater marshes (fg. 5.4).406 At Neolithic landscape, the picture is still highly Schipluiden, no fewer than 45 species of fragmentary, and it is scatered over many mammals, birds and fsh either were on the published sources. This is also the case for the menu or were caught for their fur (e.g. marten, Middle Neolithic, despite the availability of oter and beaver) or their feathers (e.g. white- relatively well-preserved sites, such as tailed eagle).407 The spectrum at Rijswijk- Schipluiden and Rijswijk-Ypenburg.405 What Ypenburg is comparable to that at Schipluiden. becomes clear from these two coastal sites is At both locations, aquatic environments are well that the contribution of animal foodstufs to the represented by fsh and fowl. Animal husbandry diet is frst of all a continuation of hunter- seems to have been more important than 405 Appendix II: PUBID 732, 775. gatherer traditions, but with the addition of hunting wild animals where it comes to meat 406 Koot et al. 2008; Louwe Kooijmans 2006. domesticated animals. The entire environmental procurement. Nonetheless, the hunting of 407 With regard to the possible signiﬁcance range in the sites’ surroundings was exploited aurochs, wild boar, red deer and a range of fur- of eagle feathers, see Amkreutz & Corbey (2008). for animal resources: the beach plain and coast, bearing animals, as well as predators, such as

Fig 5.4 Schematic representation of the economic range of the Schipluiden household in relation to environmental zones (from appendix II: PUBID 732). 173 —

Figure 5.5 Two molars of Lynx from Schipluiden (from appendix II: PUBID 732).

wolf, brown bear and fox, was more than Our reference framework is based on what the incidental. A broadly comparable picture comes archaeological record shows in terms of from the Vlaardingen site of Hellevoetsluis- exploited animals, but not on what other Ossenhoek.408 animals inhabited the landscape. Of course, So, despite the coastal location of these actualistic models of ecosystems and biotopes sites, people did not turn their back on the back- provide some reference, but they do not tell us barrier landscape. However, how this compares what the actual situation was. In this respect, the with exploitation strategies during the Middle Pleistocene record may have an advantage over Neolithic farther inland remains mostly a guess the Holocene record, as palaeontologists have due to the absence of well-preserved dedicated much work to the reconstruction of assemblages. The site of Wijchen-Oosterweg animal communities under varying climatic yielded only highly fragmented bone from a conditions. For the Holocene, such Vlaardingen context, amongst which only catle palaeontological work is basically lacking; one and pig/wild boar could be identifed.409 To what can notice a clear emphasis on the investigation extent connections existed between the coastal of anthropogenic impact on the environment, Hazendonk and the inland Vlaardingen setlers notably on vegetation. Current developments in in terms of, for instance, the exchange of animal the analysis of ancient DNA in sediments ofer resources remains an open question. In this exciting new possibilities of flling (part of) the context it is important to refer to the Hazendonk gap. Where bones are lacking, aDNA that site of Schipluiden, where three molars of at survived in sediments – notably quickly covered least two lynx were found (fg. 5.5). There exists and saturated sediment layers and soils – can no archaeological and historical evidence that provide unexpected insight into animals that lef lynx was indigenous to the Low Countries, but it their traces in the landscape through, for cannot be excluded that these animals did live instance, excrement.411 Development-led here. Today, the closest habitats are in the research in areas where such conditions can Ardennes.410 If this was also the case in early potentially be expected (e.g. deeply buried prehistoric times, it is likely that the Schipluiden palaeosols in the western Netherlands?) could specimen reached the coastal area through foster the development of such new methods, contacts with groups farther inland, for instance, and make a serious contribution to in the upstream areas of the Meuse and Rhine. archaeologically relevant environmental A fnal problem that needs to be addressed reconstruction. The question should be not only 408 Appendix II: PUBID 87. 409 Appendix II: PUBID 600. is the (probably) biased picture of what which animals were exploited by people, but 410 We must bear in mind that the present- constituted the animal landscape in its totality, also which animals were not. day habitats of lynx in north-west Europe are refugia. and how this may have changed over time. 411 Slon et al. 2017; Smith et al. 2015. 174 —

Plant food resources vegetation at various spatial scales, it does not In contrast to early prehistoric animal food inform us directly about which plants were resources, for which no recent overviews are exploited for food. For this, we are largely available, early prehistoric plant subsistence and dependent on macroremains, and, as is the case the reconstruction of the vegetation has seen for animal remains, preservation is highly much work, particularly in connection with the infuenced by local circumstances. Outside Dutch wetlands during the Late Mesolithic to wetland environments, only charred plant Middle Neolithic.412 A central theme in this remains will normally be found, with the research concerns the impact of human activity exception of specifc contexts that favour good on vegetation and the integration of crop preservation, such as wells. In wetland cultivation in the food economy. The later topic environments, both uncharred and charred is one that has received lots of atention for remains may be found, but the interpretation of decades, and has been extensively discussed in uncharred remains in connection with food recent years in connection with academic and procurement is more difcult because these may development-led research at the Late Mesolithic represent natural accumulations. to Early Neolithic sites of Hoge Vaart-A27, The application of high-resolution SEM Hardinxveld-Giessendam Polderweg and techniques to the identifcation of potential Hardinxveld-Giessendam De Bruin, as well as plant food resources, as an adjunct to the the type site locations at Swiferbant.413 The analysis of morphological features, is a step evidence places the introduction of crop plants forward. Much atention is given to species – emmer and naked barley – in these wetland identifcation of charred parenchymatous areas by 4400/4300-4100 cal BC. The evidence remains (the tissue that makes up the cortex and consists of charred cereal grains and chaf, of pith of the stem, the internal layers of leaves, pollen in sampled peat sequences, and of soil and the sof parts of fruits). This has signifcantly features that can securely be interpreted as a broadened the picture, which used to simply feld tilled by means of a hoe (fg. 5.6). If the show charred hazelnut, acorn, wild apple and suggested dates are correct, the introduction of exceptionally a berry. Important results were crop cultivation in the Dutch wetlands took obtained from the Early to Middle Mesolithic site place at least 600 years later than the of Roterdam-Yangtze Harbour (fg. 5.7).415 disappearance of LBK farmers in the south- Amongst the charred remains, edible parts of eastern part of the Netherlands. The results nine species could be identifed: oak (acorns), of development-led research, insofar as these hazel (nuts), hawthorn (berries), lesser celandine had not already been included in the recent (tubers), sedges (rhizomes, seeds), dogwood 412 Out 2009; Schepers 2014. studies mentioned above, do not argue for a (berries, seeds),416 common club-rush (rhizomes), 413 Bakker 2003; Brinkkemper et al. 1999; diferent perspective. But it should be noted yellow water-lily (seeds), and water chestnut Cappers & Raemaekers 2008; Huisman, Jongmans & Raemaekers 2009; Out that sites dating to the Early Neolithic and (nuts). Lesser celandine, sedges, common club- 2009; Peeters 2007. that are not atributed to the LBK are very rush, and yellow water-lily are additions to the 414 Appendix II: PUBID 340, 463, 468. 415 Kubiak-Martens et al. 2015. much the exception (Leur-Hernense Meer, list of charred food plant remains identifed at 416 Fruit-stones of dogwood are rich in Roterdam-Groenenhagen, Roterdam- Mesolithic to Middle Neolithic wetland sites in non-volatile oil that can be used as fuel or for the purpose of impregnation. Beverwaard).414 the Netherlands and Belgium (table 5.2).417 Large numbers of crushed fruit-stones Plant foods are not restricted to cultivated The species represented at Roterdam-Yangtze from the sites of Bökeberg and Tågerup (Sweden) suggest oil extraction in the crops. A wide variety of vegetative (potential) Harbour relate to various habitats: woodland, Mesolithic (Regnell 2012; Regnell et al. food resources was available in many parts of forest margin, scrub, moist woodland, marsh/ 1995). 417 Out 2009, 349, table 9.5. the landscape. Although pollen is informative on water edge, and shallow to deep water. 175 —

Fig 5.6 A: Plane view of the tilled horizon at Swiferbant-S4. The paterning of dark- and light-grey patches is interpreted as tillage marks made with some sort of hand tool. B: photograph showing a more or less regular patern of tillage marks that become increasingly vague towards lef of photograph. The dark layer on top is the lower part of a midden deposit. Square indicates image C. C: Detail of B, with tool marks indicated by triangles (from Huisman & Raemaekers 2014). 176 —

Fig 5.7 Plant food resources identiﬁed at Roterdam-Yangtze Harbour (from appendix II: PUBID 462). a: living yellow water-lily with fruit capsules; b: charred seed of yellow water-lily; c: water chestnuts with spines (present-day); d: water chestnut charred spine; e: living dogwood plant with berries; f: dogwood charred and broken fruit stones; g: lesser celandine SEM cross section showing polygonal parenchyma cells; h: lesser celandine SEM concentration of solid tissue; i: damp undergrowth covered by lesser celandine; j: tubers of lesser celandine; k: hawthorn charred fruit-stone; l: hawthorn living plant with berries. 177 —

Table 5.2 Carbonised macroremains of food plants (from Out 2009, table 9.5).

Region Central river Western Coastal Eem Vecht Other river

Site Polderweg De Bruin Brandwijk-Kerkhof Hazendonk Randstadrail CS Bergschenhoek Ypenburg Schipluiden 4 Wateringen Rijswijk-A4 Hoge Vaart-A27 Urk-E4 Schokland-P14 Swiferbant-S3 Schokkerhaven-E170 Doel-Deurganckdok- sector B

Taxon

Cornus sanguinea -++++- -+------+

Corylus avellana, nut shells +++++- -+++ + ++++ +

Crataegus monogyna -+-+---+-- - +-+- +

Malus sylvestris -+-+---++- + +-+- +

Malus sylvestris, parenchyma + - -+-+-+-- + --+- -

Prunus spinosa ---+--+++------+

Quercus sp. -+-++----- + ---- +

Quercus sp., cupulae - - + ------++-- -

Rosa sp. ------++------

Rosaceae -+-+------+ ---- +

Rubus fruticosus ---+------

Rubus idaeus ------+ ---- -

Rubus sp. ------+-- -

Trapa natans +++++------+-- -

Viburnum opulus ---+------+

+ present - absent

So far, litle is known about how these plant In the case of Dronten-N23, for instance, no food stufs were prepared and consumed. Lesser remains of potential plant food – other than celandine, for instance, is poisonous when eaten charred fragments of hazelnut shell – were found in raw. Amongst the more frequently echoed any of the analysed samples.419 The same is true for potential functions of pit hearths is the preparation Hanzelijn-Drontermeer and Scheemderzwaag.420 of plant food. Serious atempts to gain insight into One Late Mesolithic pit hearth at Knooppunt this aspect date back to the 1990s, through Hatemerbroek yielded a charred root fragment of pioneering work on charred parenchymatous bracken.421 As yet, the evidence for a function of pit remains from Mesolithic pit hearths in the northern hearths in the context of plant food preparation is Netherlands.418 This analysis provided evidence for extremely limited indeed. We have to keep in the presence of a wide range of vegetative organs mind, however, that the chances of fnding direct of nutritional value, notably, high frequencies of evidence may be small by defnition. Afer all, in aquatic tissue, of taxa such as bulrush, club rush general, any cooked plant food was removed from and wild beet, but also buckler fern. If these the pit in prehistory.422 Of course, overheated represent remains of food, the charred wood in (charred) parts may have remained in the pit these pits should most likely be interpreted as fuel. occasionally, but most probably in minor 418 Perry 1997, 2002. In the context of development-led research, quantities; hence, the probability of recovering 419 Kubiak-Martens et al. 2012, 343. explicit atention has been paid to the presence of parenchymatous remains from samples taken 420 Kooistra et al. 2009; Kooistra et al. 2013. 421 Kubiak-Martens 2011, 468. parenchymatous remains of potential plant food. from such pits will be extremely low. 422 Peeters & Niekus 2017. 178 —

The question arises whether the varying Prior to discussing the use of pit hearths, we results relate to diferent preservation should note that the anthropogenic nature of conditions or refect prehistoric reality. There is these features has recently been called into no demonstrable reason to explain the question: the ‘pit hearths’ have been interpreted diferences in terms of preservation conditions. as collapsed, burnt ant nests.424 The main Although they are fragile, even small charred arguments put forward are the absence of parenchymatous remains have a fair chance of evidence for in situ fre; the near absence of being recognised. Another explanation may be artefacts in the infll of pits; the ‘collapse’ of soil more likely: diferences in context. Samples with horizons due to faunal burrowing activity; and no or minimal parenchymatous remains of a geochemical profles that are explained in terms single species from Knooppunt Hatemerbroek of nutrient enrichment due to the accumulation and Dronten-N23 were taken from pit hearths; of plant and faunal debris transported by ants. samples taken from the fnds layer at Without dismissing the possibility that Knooppunt Hatemerbroek contain slightly collapsed, burnt ant nests are present in the more species. The samples from Roterdam- archaeological record, we do not support this Yangtze Harbour were taken from the fnds layer hypothesis in these particular cases, for several on the slope of the sand dune. It is possible that reasons: (a) the infll of many pits indicates remains of plant food are mostly represented in digging activities post-dating the burning event; layers that correspond to the ancient land (b) several pits contain charred tree logs surface and to waste dumps. Typically, higher arranged on the pit foor; (c) pits that show frequencies of charred hazelnut shell seem to be evidence for in situ fre do exist; (d) pits with associated with surface hearths. As systematic straight walls are frequent; (e) more pits contain analysis of surface hearths dating to the Early artefacts than is generally assumed; (f) the Neolithic at Hoge Vaart-A27 has shown, charcoal majority of artefacts from pits are microliths; samples from these contexts also contain a and (g) pedological and geochemical broad range of plant species.423 It is therefore phenomena can be explained in various ways.425 possible that observed diferences relate to Hence, the picture is far more complex than behaviourally distinct contexts. suggested in the ant nest hypothesis. Pit hearths are a common phenomenon on Mesolithic sites north of the Rhine (fg. 5.8), 5.2.2 Mesolithic pit hearths: wood for where they occur as ‘isolated’ features, as well fuel and tar as in concentrations of several tens to hundreds.426 Concentrations of pit hearths are absent south of the Rhine; and isolated pit The only non-food use of organic material that hearths are rather exceptional in this area. In almost always features in the examined reports contrast to surface heaths, which mostly consist concerns the use of wood for fuel, evidently of scaters of charcoal associated with heat- present as incompletely combusted wood (i.e. exposed to heavily burnt lithics, charred wood charcoal). It is tempting to relate charcoal hazelnut shell and bone (provided preservation found at archaeological sites to the deliberate conditions are favourable), pit hearths are of U- fring of wood, but this does not always need to or bowl-shaped features with an infll of charred be the case. Accidental burning as a result of plant remains and sand. About 3500 examples human activity or natural phenomena (i.e. from c.175 sites have been recorded to date, and lightning) also leads to the deposition of about 750 radiocarbon dates make clear that pit charcoal. In addition, wood may have been hearths are almost entirely restricted to the burnt or heated for purposes other than fuel. Mesolithic.427 Both at the site and the regional 423 Visser et al. 2001. 424 Crombé et al. 2015. Tar production is one of the possibilities that level, paterns of spatio-temporal diferentiation 425 Niekus et al. in prep; Peeters & Niekus may have been overlooked in past studies. can be observed that appear to have behavioural 2017. 426 Niekus 2006; Peeters & Niekus 2017. Development-led research has made a meaning. We will discuss these aspects in 427 Peeters & Niekus 2017. signifcant contribution to several of these section 5.3.1. Despite their common presence on 428 Groenendijk 1987; Groenendijk & Smit 1990; Jansen & Peeters 2001; Kooistra issues, in particular in the context of the many Mesolithic sites, we are, however, still 2011, 2012; Kubiak-Martens et al. 2011; discussion about the function of Mesolithic pit poorly informed about the function of these pit Kubiak-Martens et al. 2012; Peeters & Niekus 2017; Perry 1997. hearths. hearths.428 Nonetheless, some development-led 179 —

Fig 5.8 Geographical location of Mesolithic sites with pit hearths (from Niekus 2006; the map does not show the sites that have been discovered since), and pit hearths under excavation at Knooppunt Hatemerbroek (from appendix II: PUBID 259). 180 —

Fig 5.9 Type A pit hearths at Dronten-N23 (from appendix II: PUBID 538).

projects have included targeted investigations occurring in dense clusters, Type A pits hardly and the development of fresh hypotheses, which ever intersect. Type B pits have a circular to we will discuss below. rectangular shape, and ofen have fat foors and Within the broad category of pit hearths, straight walls, but bowl-shaped and funnel- several types can be distinguished on the basis shaped cross sections also occur (fg. 5.10). Their of pit size, shape and infll. At the site of diameter varies from 80 to 160 cm, and their Dronten-N23, no fewer than 772 pit hearths depth from 60 to 120 cm. The infll of these pit were recorded. These were classifed into three hearths is more complex and chaotic, with types429 (A, B and C) that are also described from charcoal occurring in several layers. Small to other sites. Type A, which could be called the large, angular lumps of charcoal are present in ‘classic’ pit hearth, typically has a circular to oval all layers, and charred logs are sometimes outline with a bowl-shaped cross section. encountered on the pit foor. A band of orange- Occasionally, somewhat fater botoms and red burnt sand is frequently observed. Type B straight walls occur (fg. 5.9). The diameter varies pits can occur as intersecting series, which can from 25 to 160 cm, but most measure between make the delimitation of individual pits difcult. 50/60 and 110/120 cm; the depth varies between Finally, Type C pit hearths have a circular shape, 30 and 135 cm, but is mostly between 50 and with a diameter ranging from 40 to 100 cm, and 90 cm. This variability in size and depth is partly a bowl-shaped cross section, with a depth of up related to diferential preservation, due to to 45 cm. The infll consists of a chaotic mixture surface erosion. Erosion efects seem to occur at of sand and charcoal (although it is not charcoal Knooppunt Hatemerbroek, where a slight rich). A thin band of orange-red burnt sand correlation between diameter and depth has surrounding the pit is frequently observed. been observed.430 The infll of Type A pits usually Although it has been questioned whether Type C consists of a charcoal-rich layer in the lower part pits should be interpreted as pit hearths,431 the 429 Opbroek & Hamburg 2012. and a layer of clean sand in the upper part; there bands of burnt sand on their perimeter argue in 430 Knippenberg & Hamburg 2011. 431 Opbroek & Hamburg 2012, 132. is no evidence for burnt sand. Although favour of such an interpretation (fg. 5.11). 181 — Fig 5.10 Type B pit heaths at DrontenN23 (from appendix II: PUBID 538). 182 —

Fig 5.11 Type C pit hearth at Dronten-N23 (from appendix II: PUBID 538).

The many 14C-dates available indicate that The overarching term ‘pit hearth’ suggests Type A pits occur throughout the Mesolithic, that these features are connected with one or whereas Type B pits are of Late Mesolithic age. more functions that require the burning of fuel Type C pits are mostly of Late Mesolithic age; as as a source of heat. Experiments conducted in yet, only one of Middle Mesolithic date is the 1980s showed that temperatures up to 1000 known, from Dronten-N23. The younger, Type B °C can be reached in an open pit.434 The analysis pits at Dronten-N23 are generally wider and of charcoal sampled from pit hearths at the sites deeper than the Type A pits. Also, at Knooppunt of Knooppunt Hatemerbroek, Dronten-N23, Hatemerbroek there seems to be a tendency for Tunnel Drontermeer, Leeuwarden-Hempens, pits (it is not clear whether all are of Type A) to Epse-Olthof, Scheemderzwaag and become wider and deeper in the course of time. Siegerswoude returned mutually comparable As yet, there is, however, no reason to argue for pictures with regard to a dominance of the wood a general trend, since at other sites the younger of pine and oak. Similar observations had been pits (e.g. Scheemderzwaag-Scheemda) are more made earlier, for instance, at sites in the shallow than the older ones (e.g. Groninger Veenkoloniën (Groningen peat Scheemderzwaag-Opdiep).432 Indeed, in the case colonies), Mariënberg and Hoge Vaart-A27.435 It of Leeuwarden-Hempens, pits that were is generally assumed that this wood was used as identifed deeper in the sandy subsoil appeared fuel, because pine was dominant from the to be older; this can be explained by either Preboreal to the Mid-Atlantic. Oak occurs periodic deposition of wind-blown sand, or throughout the Boreal to the Late Atlantic, but is older pits simply having been cut more deeply always accompanied by pine in the earlier phase. into the subsoil when they were constructed.433 In most cases, charcoal of these taxa is found as Nonetheless, we may have to consider the smaller fragments in the lower half of the pits. possibility of chronological diferences in the Occasionally, larger lumps and logs are found in 432 Appendix II: PUBID 472. 433 Noens 2011. occurrence of the various pit hearth types, and, confgurations that clearly show that wood was 434 Groenendijk & Smit 1990. in view of their variable infll, also functional carefully placed in the botom of pits (Type B; 435 Groenendijk 1997; Van Rijn & Kooistra 2001; Verlinde & Newell 2006. diferences. see above). At Dronten-N23, for instance, larger 183 —

lumps and charred logs were particularly present has been found at Knooppunt Hatemerbroek,439 at the outer limits of the pit foor (fg. 5.10). In Dronten-N23,440 Hanzelijn-Drontermeer441 and one case, large lumps and logs appeared to have Scheemderzwaag.442 Typically, ‘glassy mater’ is been stacked against the wall of the pit but not frequently present amongst and/or atached to in the central part. This could indicate a specifc charcoal remains (fg. 5.12). This ‘glassy mater’ mode of fring and heat control, for a purpose is highly refective and rich in graphite, and has a which remains, as yet, unknown. In more porous structure. In several cases it has been general terms, the function of pit hearths is observed that this structure ‘emerges’ from the believed to have been connected with the wood cells and pores, which suggests that this is preparation of food stufs, for instance, the the result of liquefaction. In addition, gas braising of meat, roasting of hazelnuts, and chromatography-mass spectrometry and cooking of starchy roots and tubers (see section infrared spectroscopy have provided evidence 5.2.1).436 It has also been suggested that pit (in the form of phenanthrene) for thermal hearths were used for the heating of cooking degradation of pine wood through distillation, stones (so-called pot boilers) or heat treatment a process called pyrolysis (fg. 5.13). Tar-like of fint (thermo-preparation).437 globules have been shown to result from the An important step towards the investigation heating of wood or bark under relatively high of pit hearth function in the context of temperatures (c. 400 °C) and oxygen-poor development-led projects involves the SEM and conditions. Because tar production requires physico-chemical analysis of charcoal and tar- temperatures below c. 400 °C, the glassy mater like substance collected from pit hearths. should be interpreted as a by-product, namely, Following up on the initial suggestion wood that was overheated, to temperatures concerning the potential use of pit hearths for between 450 and 600 °C. The tar-like globules the production of wood tar at Hoge Vaart-A27,438 can be viewed as ‘lefovers’, and the charcoal as various researchers have made serious atempts the remains of fuel. to investigate this further. Evidence for such use

436 Groenendijk 1987. 437 Groenendijk 2004. However, there is no evidence for thermo-preparation of ﬂint (or any other siliceous rock) in the Dutch Mesolithic, nor in the Neolithic (Peeters 2001; Peeters, Schreurs & Verneau 2001). 438 Jansen & Peeters 2001. 439 Kubiak-Martens et al. 2011. 440 Fig 5.12 Charred fragments of a glassy substance from Dronten-N23 (from appendix II: PUBID 538). Two fragments Kubiak-Martens et al. 2012. 441 Kooistra et al. 2009. are stuck to wood charcoal that shows anatomical features of pine (red boxes). 442 Kooistra et al. 2013. 184 —

Fig 5.13 Gas chromatography spectrum of a sample of glassy mater from a pit hearth at Dronten-N23 (from appendix II: PUBID 538). The spectrum indicates phenanthrene and derivates, which result from thermal degradation of pine wood through distillation. 185 —

The scenario of pit hearths having been Although the obtained results with regard used for the purpose of tar production is to the Mesolithic pit hearths are far from certainly plausible. Wood tar is known from the conclusive, it must be recognised that archaeological record, and has been used for development-led research has made a tool production (e.g. hafing of microliths in the considerable contribution to the discussion Mesolithic) and possibly medicinal purposes.443 about the function of these enigmatic features. The fact that relatively few archaeological The hypothesis about tar production can be examples of, for instance, fint tools with tested further by extending requirements for atached tar remains are known, should in itself scientifc analysis of samples that are to be not be a surprise. Wood tar is vulnerable to collected in the feld. Experimental work is chemical (oxidation) and biological necessary to explore the conditions under which (microorganisms) degradation, and is best particular phenomena occur, in order to be able preserved under oxygen-poor or oxygen-free to (potentially) identify the range of interpretive conditions.444 On the other hand, a word of possibilities. Clearly, this is not feasible in the caution is needed: unintended formation of tar context of development-led projects. can occur under favourable conditions. Slow cooking of large quantities of meat, for instance, requires low temperatures to prevent charring or 5.3 Environmental dynamics and carbonisation. Ethnographic examples of such behavioural diversity slow cooking involve the heating of meat in closed pits.445 Furthermore, we have to be careful not to connect pit hearths with a single function. The relationship between the dynamics of the Tar production is one possibility amongst environment and human behaviour is amongst several, and it may just be the one that is more the themes that have traditionally received a lot easily identifed archaeologically. However, if pit of atention in Dutch archaeology. The role of 443 Aveling & Heron 1998; Baumgartner hearths should indeed be interpreted in the biology and physical geography has been et al. 2012; Bokelmann 1994; context of tar production, an important question prominent in much research ever since the early Larsson 1983. 444 Regert 2004; Weiner 2005. that arises is how all this connects with the days of Dutch archaeology, and continues to be 445 Jansen & Peeters (2001) reported on the chronological occurrence of the three types of so today. Of course, this is not only the case in geochemical analysis of charcoal and sediment samples from Mesolithic pit hearths described above. Does the the Netherlands. Approaches in Britain and hearth pits at Hoge Vaart-A27, and have appearance of Type C and Type B pit hearths southern Scandinavia are very similar, and, not indicated a possible use of these pits for the preparation of animal-derived food. refect technological changes in production? And surprisingly, these countries share many However, no further chemical analysis what is the potential connection with changes in research themes. In this section we will focus on has been executed since; the results must therefore be approached with the dominant wood taxa represented in the two themes to which development-led projects caution. charcoal remains? Did oak merely serve as fuel, have made a contribution, notably, (dis) 446 Bear in mind that our reference to long- term changes in Holocene vegetation is and was pine preferentially selected as the continuity of occupation and behaviour, and still based on over-generalised models resource for tar? Or was pine replaced by birch in human infuence on the environment. of vegetation succession, which do not take in consideration (sub)regional the absence of pine?446 variability. 186 —

Dronten-N23, Knooppunt Hatemerbroek, Epse- 5.3.1 Continuity and discontinuity of Olthof and Leeuwarden-Hempens. These sites occupation and behaviour principally consist of scaters of lithics and clusters of pit hearths. First, the 14C-dates demonstrate vast time-depths (fg. 5.14; fg. 5.15), The use of space at a wider geographical scale is spanning more than 1000 radiocarbon years in refected in the use of particular spots in the each case. Second, the time span of lithic scaters landscape. At this scale, one would preferably does not fully coincide with that of pit hearths, as investigate inter-site variability in terms of the far as this can be established from the dates. We particular use or function of locations in the will take a closer look at these aspects below. landscape and in terms of location choice. The The long time span of use of the sites short-term/regional scale corresponds to the mentioned above suggest some degree of cultural geography of a defned region, whilst a occupation continuity. At this long-term scale, long-term/regional perspective corresponds to however, one should not confuse occupation cultural landscape history or biography (see continuity with permanent presence of groups of section 1.2.4). As outlined in the foregoing hunter-gatherers, or even visits on a yearly basis. sections, this scale can barely be investigated on The obtained 14C-dates are frst of all indicative of the basis of the results of development-led the entire time frame within which activities took projects without falling back onto prevalent place, as represented by the sampled contexts. models of landscape use. Novel information Although many dates from a single site can needs to come from the investigated sites statistically be considered ‘contemporaneous’, themselves. And to obtain this, it is useful to this does not signify contemporaneity in an take a look at how particular locations have absolute sense. The probability distributions of been used over time. What time-depth is the calibrated dates are generally too wide to involved in the use of locations? Have locations permit any such suggestion. The highest always been used for the same purpose, or can resolution will be in the order of one to two we observe changes or alternations? In other decennia at a 1σ confdence level. What the words, what does information about (dis) results do demonstrate, however, is the use of continuity in occupation and behaviour at the specifc locations in a landscape over many site level learn us about what role particular generations. Table 5.3 provides a (over-)simplistic places played in the broader use of landscapes? estimate of the number of generations that may Of particular interest in this respect are have visited the sites for which we have some several Mesolithic sites (some of which were confdence about the maximum time-depth of used into the Neolithic) for which considerable long-term occupation due to the availability of numbers of 14C-dates are available, notably, many 14C-dates.

Table 5.3 Estimate of the number of generations of hunter-gatherers that potentially visited a site. The maximum time-depth for the individual sites is set by the calibrated extreme start and end dates at the 2σ conﬁdence level. One hunter-gatherer generation is set at 20 years.

Site Maximum N generations time-depth (yr)

Dronten-N23 3100 155

Leeuwarden-Hempens 2100 105

Knooppunt-Hatemerbroek 1800 90

Epse-Olthof 1740 87 2[&DOY%URQN5DPVH\ U$WPRVSKHULFGDWDIURP5HLPHUHWDO 2[&DOY%URQN5DPVH\ U$WPRVSKHULFGDWDIURP5HLPHUHWDO

+$.& ,9 *U$ +DDUGNXLO ,9 +$.% *U$ +DDUGNXLO +$.$ +$.& *U$ +DDUGNXLO +$.& *U$ +DDUGNXLO +$.& *U$ +DDUGNXLO +$.% *U$ +DDUGNXLO +$.% ,,, +$.% *U$ +DDUGNXLO +$.% *U$ +DDUGNXLO +$.% *U$ +DDUGNXLO +$.% *U$ +DDUGNXLO +$.% +$.% *U$ +DDUGNXLO +$.% *U$ .XLO +$.% *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.% +$.& *U$ +DDUGNXLO +$.% *U$ +DDUGNXLO ,,, +$.% *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.% +$.% *U$ +DDUGNXLO ./$ *U$ +DDUGNXLO +$.% *U$ +DDUGNXLO +$.& *U$ +DDUGNXLO +$.% *U$ +DDUGNXLO +$.$ +$.% *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO 1227 *U$ +DDUGNXLO 1227 ,, *U$ +DDUGNXLO +$.$ +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO 1227 *U$ +DDUGNXLO 1227 *U$ +DDUGNXLO +$.$ +$.$ *U$ +DDUGNXLO ./% *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ +DDUGNXLO +$.$ *U$ +$.$ *U$ +DDUGNXLO ,, +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ , +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO 1227 *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ +$.$ *U$ +DDUGNXLO +$.& *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ +$.$ *U$ .XLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ *U$ +DDUGNXLO +$.$ +$.$ , +$.$ +$.$ +$.$ &DOLEUDWHGGDWH FDO%& +$.$ +$.$ +$.$ +$.$ +$.$ +$.$ +$.$ +$.$ +$.$ ./% 1227 +$.$ +$.$ +$.$ +$.& 1227 +$.$ 1227 +$.$ +$.$ +$.$ 1227 +$.$ +$.$ +$.$ +$.$ 1227 1227 1227 1227

&DOLEUDWHGGDWH FDO%&

Fig 5.14 Lef: calibrated 14C dates for Dronten-N23. Dates on charred hazelnut shell from the ﬁnd layer are in red; all other dates refer to samples from pit hearths. Breaks within the chronological sequence are indicated with circles (from appendix II: PUBID 538). Right: calibrated 14C dates for Knooppunt Hatemerbroek. All dates refer to samples from pit hearths. Breaks within the chronological sequence are indicated with circles (from appendix II: PUBID 538). 188 —

Fig 5.15 Calibrated 14C dates for Epse-Olthof; all dates refer to samples from pit hearths (from appendix II: PUBID 538). The dates are presented for diﬀerent parts of the investigated area and diﬀerent excavation campaigns (Olfof Noord; Olthof Zuid phase 1; Olthof Zuid phase 2); this subdivision, however, has no archaeological meaning. 189 —

Fig 5.15 Continued. 190 —

It is impossible to tell whether 2 m in the southern North Sea basin caused by representatives of each generation will actually the catastrophic drainage of Lake Agassiz around have visited the sites. The data do not suggest 6300 cal BC,448 had environmental efects that the contrary, although in the case of could have been experienced within one or two Dronten-N23 and Knooppunt Hatemerbroek, generations.449 Interestingly, the timing of this one can observe several breaks in the overlap of event corresponds with one of the breaks in the calibration probabilities. The question is, what graphs of both Dronten-N23 and Knooppunt causes these breaks? Because they occur around Hatemerbroek. Data on the occupation history the same moment in time at both sites, it seems of the Rhine-Meuse estuary also appear to show improbable that this is related to chance, e.g. a hiatus during this window (fg. 5.16).450 due to coincidental lack of samples These important data, which were corresponding to the time gaps. However, it is generated in the context of development-led possible that the breaks are related to research, also permit us to discern some other fuctuations in the calibration curve and are of a paterns at several spatial scales. In the case of statistical nature. It would require in-depth Dronten-N23, it seems that there has been a statistical analysis to confrm or rule out this change in the nature of activities over time. possibility. Yet another scenario is that the During the frst two millennia of its occupation breaks correspond to interruptions in activity, history, activities involved two behavioural or at least the sort of activity that involved the contexts. One was associated with fint creation of contexts that are normally sampled knapping, fring of surface hearths, food by archaeologists for dating. preparation and consumption (hazelnuts, Whatever the exact nature of these breaks in possibly meat); the other with the use of pit the calibration graphs, the astonishing time- hearths. The frst context seems to disappear depth represented at these sites is all the more from the record in the third millennium of the remarkable in the context of environmental site’s occupation history, afer which we only see dynamics. The Early and Middle Holocene have the use of pit hearths. A somewhat comparable seen the infuence of climate change on ongoing picture emerges from the data of Epse-Olthof. sea level rise, vegetation, hydrology and soil Although less well supported by 14C-dates, here conditions. In the long run, environments have we also see fint-knapping activities and use of changed in terms of ecology and physical pit hearths during the earlier phases of the site’s geography, yet particular places in this dynamic occupation history, but only use of pit hearths landscape continued to play a role in the cultural during the last phase. This suggests that geography of Mesolithic hunter-gatherers.447 The generations of hunter-gatherers continued to 447 Amkreutz 2013; Peeters 2007, 2009a, question is to what extent these people actually occupy particular spots in the landscape, but for 2009b. experienced such change at the generation scale. diferent purposes. Such a combination of long- 448 These climatic effects are also known as the 8.2 Ka Event. It cannot be excluded that a sudden drop in term occupation continuity and behavioural 449 Also see Leary 2009. temperature – due to stagnating thermohaline continuity and discontinuity was previously 450 Peeters et al. 2015. 451 Peeters 2007, 2009a, 2009b. circulation – and a sea level jump of an impressive observed at Hoge Vaart-A27.451 191 — Fig 5.16 Chronological coverage of sites and isolated finds from the Roterdam region (from appendix II: PUBID 462). Note the gap between 6400 and 6100 cal. BC, which corresponds to a phase of the gap Fig 5.16 Chronological coverage of sites and isolated finds from the Roterdam region (from appendix II: PUBID 462). Note climate event, Sea basin,increasing sea level rise and the 8.2 Ky Storrega tsunami, which variably affected the coasts of North which are both related to the catastrophic drainage of Lake Agassiz. The an artefact of research intensity is as yet unclear. what extent the chronological gap signals discontinuity in occupation history or represents To also falls within this time window. 192 —

Scaling up the data to a long-term regional vegetation density and dryer conditions.453 The perspective reveals another interesting patern question is whether Mesolithic hunter-gatherers that may relate to changing environmental indeed abandoned the area. Evidence from conditions. A plot of 14C-dates from several development-led research at various locations Mesolithic sites along the prehistoric drainage near Zutphen points to Late Mesolithic and Early system Hunnepe (partly corresponding to the Neolithic activity, but this activity is not present-day river IJssel, in the eastern associated with fint knapping or the use of pit Netherlands) shows diferences in the temporal hearths. The evidence consists of 14C-dated tree occurrence of pit hearths (fg. 5.17). Upstream trunks with chop marks and incidental (Deventer area) we mainly see Boreal and Early fragments of red deer antler with cut marks,454 as Atlantic dates, midstream (Knooppunt well as digging traces that are thought to relate Hatemerbroek; Zwolle area) predominantly to the extraction of plant roots or tubers Early and Middle Atlantic dates and upstream (fg. 5.18).455 Stray fnds of isolated trapezoid (Dronten-N23) Boreal to Middle Atlantic dates. points and occasionally occurring punched Broadly speaking, the Late Mesolithic is almost blades also indicate the presence of Late entirely lacking in the 14C-evidence for the Mesolithic hunter gatherers in both the Berkel upstream region. This picture coincides with the and the Hunnepe drainage system. This shows virtual absence of Late Mesolithic fint that people had not lef the area, but used technology in the same region.452 Also in the various landscape zones in diferent ways. At an Zutphen area, c.15 km south of Epse-Olthof, but even wider scale, shifs seem to be discernible situated along the Berkel drainage system, there between regions. In the north-eastern part of is virtual absence of Late Mesolithic fint the Netherlands 14C-dates from pit hearths technology despite extensive development-led predominantly fall in the late Preboreal and excavations and explicit atention for early Boreal, whereas dates from the Flevoland region prehistoric sites and phenomena. Unless the predominantly fall in the Boreal and Atlantic patern is the result of insufcient data, this (fg. 5.19). possibly indicates variability in landscape use. From this perspective, it is interesting to Palaeovegetation reconstructions for the return to the use of wood in the context of pit Early Mesolithic have led to the suggestion that hearths. The broad picture is one of an increase the disappearance of human activity in the in oak wood charcoal from the Boreal into the Zutphen area was mainly the result of increasing Atlantic, and corresponding decrease in pine

452 Late Mesolithic technology is characterized by regular, punched blades and trapezoid points. 453 Bos et al. 2005. A comparable hypothesis had already been developed by Waterbolk (1985, 1999), who connected the ‘abandonment’ of the sandy areas in the Mesolithic with the expansion of woodland and successive decrease in game (that is, large mammals) in these areas. 454 Groenewoudt et al. 2001. 455 Fermin 2006. Fig 5.17 Plot of 14C dates (n=243) for four clusters of sites with pit hearths along the Hunnepe drainage system. 193 —

Fig 5.18 Traces of Mesolithic digging activity at the botom of a fen (extent indicated in pink) at Zutphen-Looërenk (from Fermin 2006). These traces possibly result from the extraction of plant roots or tubers (notably arrow-grass, which has been atested to by pollen). 194 —

wood charcoal. This basically fts the overarching Flevoland area, for instance, has resulted in new models of Holocene vegetation succession in interpretations, which propose that nutrient- connection with increasing temperature. If the poor soils in large parts of this region permited availability of pine (and/or birch?) was crucial in pine and heather to persist for much longer, well the functional context of these pit hearths, then into the Atlantic.456 Birch also thrives well in the inter- and intra-regional shifs in temporal these nutrient-poor conditions. It is only in the occurrence of these pits may relate to spatial second half of the Early Atlantic that more variability in vegetation communities. A recent demanding tree species can take advantage of evaluation of palaeovegetation data from the improving soil conditions due to altering

Fig 5.19 Sum probability plot (CalPal version 2016) of 756 radiocarbon dated pit hearths from the Netherlands. The dates were calibrated using the calibration curve IntCal13 (Reimer et al. 2013). 195 —

groundwater fows, which, in turn, are the result of further sea level rise. 5.3.2 Human activity and environment The results from development-led work evidently permit researchers to reconsider models of hunter-gatherer landscape use. Prehistoric hunter-gatherers are generally Landscape dynamics need to be understood at supposed to have had litle ‘measurable’ impact diferent spatial and temporal scales in order to on the environment. It was not until intensifed be useful for archaeological purposes. agriculture became established that the ‘natural’ Palaeolandscape data are much needed, but environment began to be afected, principally their interpretation requires contextualisation through the clearing of woodland and the that reaches far beyond the site level in order to grazing of animals. Woodland clearing and be useful.457 The aspects discussed above do intensifed grazing had major impacts on the bring us back to the burnt ant nest hypothesis composition and spatial structure of the mentioned in the previous section. The long vegetation, but also on the development of time-depth of the occurrence of these features soils, on surface stability, and on hydrological has been used as a proxy for wildfre frequency conditions. Clearly, this would also have afected during the Early and Middle Holocene, where the distribution and composition of the animal synchronicity of 14C-dates of pits with the 9.3 Ka community. and 8.2 Ka climate events is explained in terms In a contrast-and-compare of hunter- of widespread and repeated forest fres due to gathering and agricultural lifeways, hunter- desiccation (cold and dry climate).458 In this gatherers are plainly considered to have been scenario, the relative importance of various fully dependent on their environment. Hence, wood types is merely a refection of local tree hunter-gatherers are conceived of as ‘natural’ species. Fire ecology, be it natural, beings that had to survive in the harsh anthropogenic or both, is complicated, and its wilderness.460 This position has led to a study requires multiproxy data (charcoal, 14C, persistent perspective of hunter-gatherers as pollen, microfauna). Wildfres certainly occurred, living in an environment that was merely a and no doubt occurred frequently in pine- quarry of foodstufs and raw materials. Hence, it dominated landscapes, but this does not negate is not surprising that hunter-gatherer research the interpretation of some or even most of these became dominated by questions related to ‘pit hearth’ features as anthropogenic pits. The subsistence and setlement systems. suggestion that ‘there is no need during Of course, this monolithic perspective on excavation to focus on these kind of features hunter-gatherers − particularly so with regard to any longer or at least less resources can be spent the Mesolithic − has not been readily accepted on them than is currently the case’459 is a step too by all researchers. The whole idea of hunter- far. In fact, even if excavators were to fnd gatherers being people who simply functioned themselves confronted with climate-related as passive ‘receptors’ in a giving environment natural phenomena, we would argue that this has explicitly been challenged.461 The central 456 Kooistra & Peeters in prep. would require serious atention from both argument is that the impact of behaviour on the 457 The required degree of palaeoecologists and archaeologists, since these characteristics of a given environment can be contextualisation and synthesis is at the level of PhD research, and therefore far data are crucial to the interpretation of the considerable, without leaving an easily beyond the constraints of time and dynamics of human ecology. However, not all recognizable imprint on the archaeological or money in development-led archaeology. 458 Crombé 2015. pits are similar, and as yet, an anthropogenic even palynological record. In particular the 459 Crombé et al. 2015, 169. origin seems more likely for many, if not most, deliberate and controlled burning of vegetation 460 Zvelebil & Moore 2006. 461 E.g. Mellars 1976; Zvelebil 1994; of them. would have afected the characteristics of the Simmons 1996. 196 —

plant community and, subsequently, the animal us with a considerably smaller, and thus harder community.462 Simmons goes as far as to relate to identify, environmental impact. Manipulation the creation of the moorlands in England and of plants that are part of the ‘natural’ vegetation Wales to Mesolithic intervention, which would is simply difcult to identify. In the context of have resulted in more open and weter Knooppunt Hatemerbroek, it was even stated environments.463 that “only people who permanently setle somewhere, The suggestion that hunter-gatherers and work the soil, leave recognizable traces in pollen managed their environment has not remain profles” (translation ours).468 Whilst this is, in our undisputed, despite historically documented view, overstated, it rightly points out the evidence for deliberate and controlled burning problem – the key word probably being of vegetation by hunter-gatherers to be a recognizable. Also, it must be mentioned that common practice in various parts of the world, detecting an anthropogenic ‘signal’ is not e.g. North America, South America, and necessarily the same as detecting anthropogenic Tasmania.464 Of course, one can argue that the ‘impact’.469 behaviour of (near-)recent hunter-gatherers The infuence of human activity on the provides no evidence for similar practices in the environment is explicitly discussed in the context past. However, there is mounting evidence from of Groningen-Meerstad, Almere-Overgooi, microfossil and archaeological analysis that is Knooppunt Hatemerbroek, Dronten-N23 and strongly suggestive of deliberate clearance of nearby sections in the Hanzelijn railway vegetation through fre by Mesolithic and trajectory.470 At Knooppunt Hatemerbroek, possibly Late Glacial hunter-gatherers. Most of geological indications are mentioned, notably the evidence currently comes from Britain,465 the presence of wind-blown sand deposits, whereas the number of claims for clearance which are believed to testify to soil degradation practices by hunter-gatherers from the continent as a result of human activities. In the Mesolithic, was quite sparse until recent years.466 these disturbances were very local and of limited Development-led projects in the Netherlands, intensity.471 However, the implications of this however, have contributed considerably to this observation are downplayed by the palynological topic. results, which show no defnite sign of human Palaeoecological analyses are classically impact.472 What can and cannot be considered an considered to potentially provide insight into anthropological indicator in pollen diagrams is a anthropogenic impact on the environment. The classic debate in palynology. Frames of reference chance of actually detecting human impact is are strongly linked to large-scale interventions, dependent on a variety of factors. First, as was such as systematic forest clearance and food pointed out some 20 years ago by Bos and production as an elementary mode of Janssen,467 the resolution and distribution of subsistence. More subtle interventions, such as sampling play a key role in the potential to localised, small-scale disturbances or clearances detect human impact. Hence, sampling of vegetation, are not easily identifable. strategies need to be adjusted to specifcally Typically, these types of interventions, whether answer this question, as illustrated by the study deliberate or accidental, will have been more by Bos and Janssen; however, this is hardly ever common in early prehistoric contexts. the case. Obviously, the feasibility of such an Direct indicators for fre come from charcoal approach is highly dependent on the constraints particles in pollen slides or burnt macroremains that every project faces in terms of research in natural deposits. Indirect indicators consist of 462 Mellars 1976. potential (in this case, palaeoenvironmental), remains of organisms that are considered to be 463 Simmons 1996. 464 Mellars 1976; Scherjon et al. 2015. practical approach and budget. fre indicators, such as Pteridium and 465 Innes & Blackford 2003; Mellars & Dark A second factor that infuences the chance Gelasinospora. Coprophilous fungi are becoming 1998; Simmons 1996. 466 Bos & Janssen 1996; Bos & Urz 2003; Bos of detecting human impact involves age. In increasingly important as an additional et al. 2005. general, one could state that the further one argument for human presence and impact. 467 Bos & Janssen 1996. 468 Van Haaster 2011, 461. goes back in time, the smaller the chances are of In addition to these, specifc plant taxa are 469 Behre 1981. detecting human impact. This is caused by both considered indicators of human presence 470 Appendix II: PUBID 747; De Moor et al. 2009; Opbroek & Lohof 2012; Woldring practical restrictions, such as preservation, and because of their present-day ruderal growth et al. 2012. the reality of the past. Various reports point out location, such as Plantago spp., Chenopodiaceae, 471 Van Zijverden 2011, 67. 472 Van Haaster 2011. that pre-agricultural societies are likely to leave and Artemisia. As is apparent from table 5.4, the 197 —

locations for which indications for fring have chosen. Human involvement in environmental been suggested produced variable evidence. dynamics is one of spatial and temporal and of With respect to the various arguments qualitative (i.e. species composition) and summarised in table 5.4, it must be kept in mind quantitative (i.e. species importance) scale: that some of the authors list all taxa leading to • Human presence: localised signals ‘isolated’ in their interpretation, whereas others make a time (single event) without qualitative or selection. Moreover, the categories listed in the quantitative efects on the environment tables are not strict: various indicators of open (charcoal); vegetation may also be eutrophication or fre • Human infuence: localised and recurrent indicators. signals (sequential, interrupted events) with More important is the distinction between limited quantitative and no qualitative efects human presence, as it is tentatively termed for on the environment (charcoal, pollen); Roterdam Yangtze Harbour,473 and human • Human infuence: widespread, single events impact. In English, the term human impact is and/or sequential, interrupted events with commonly used in this context, whereas the limited quantitative and qualitative term commonly used in Dutch is 'menselijke perturbations (charcoal, coprophilous fungi, invloed' which arguably is beter translated as pollen); ‘human infuence’. Without wanting to get • Human impact: localised or widespread bogged down in a semantic discussion, we feel sequence of perturbation with measurable that the degree of human involvement is not qualitative and quantitative efects (charcoal, always considered when various terms are coprophilous fungi, pollen).

Table 5.4 Various anthropogenic indicators mentioned in the selection of reports. Milheeze Zutphen-Ooyerhoek Harbour Roterdam-Yangtze Groningen-Meerstad Almere-Overgooi Hanzelijn Nieuwe Land

Clearance Tree drop pine

Coprophilous funghi x tentatively x x

open ground vegetation x

Euthrophic Several ruderals x

Chenopodiaceae x

Urtica x x tentatively

Artemisia

Fire Epilobium x

Charcoal x x tentatively x particle size x x

Melampyrum pratense x

Gelasinospora retispora xx

Sphagnum x

Pteridium x

contrast samples x

Plantago xx

Extensive list of anthropogenic indicators x 473 Kubiak-Martens et al. 2015, 271. 198 —

Fig 5.20 Regional vegetation maps for the Younger Dryas (from Bouman et al. 2013). DdTp: park landscape with birch, pine, dwarf shrubs and many herbaceous; DdTb: park landscape with birch, dwarf shrubs and many herbs; DdKk: open vegetation rich in herbs, with heather and dwarf shrubs; DvKk: open vegetation rich in herbs, with heather and dwarf shrubs; DnOo: open plain with dwarf shrubs, birch and pine along the edge. Triangles indicate Late Palaeolithic sites. 199 —

In this scheme, presence, infuence and impact It is therefore important to consider designate an increasing intensity of human methodological approaches that permit involvement, which, we feel, is useful to archaeologists to increase the interpretive characterise datasets in terms of interpretation. possibilities in a beter way. Notably, the It is evident that in the context of inclusion and re-interpretation of existing pollen development-led work, the evidence found will data at a regional scale may provide valuable primarily be relevant at the local scale. Judging support for the interpretation of new data. from the reports, we are inclined to speak in An example concerns a vegetation map of the terms of human presence or infuence at most. Younger Dryas in Twente (eastern Netherlands), As mentioned in the previous section, the which shows a more diferentiated picture than interpretation of palaeoenvironmental data in a generalised model would (fg. 5.20).474 terms of humans’ infuence on their Combined with the distribution of Late surroundings requires a higher level of synthetic Palaeolithic sites, the map gives enough insight analysis. The ecological interconnectedness of into (potential) setlement locations and plants, animals and humans is too complex to vegetation zones to permit development of new be understood on the basis of data that are research questions and approaches that can also interpreted at the local (i.e. site) scale. The be integrated in development-led work. research questions are, however, highly relevant. 474 Bouman et al. 2013; Van Beek et al. 2017.

201 6 Research agenda, ﬁeldwork and — reports

dimensions, which, in combination, form 6.1 Introduction diferent knowledge domains. This permits us to clarify in what way the research agenda, feldwork practices and reporting conventions Research agendas should primarily be concerned aid or obstruct archaeology as a discipline with the investigation of unknowns. In primarily aimed at an understanding of, and the archaeology, which is the discipline that studies creation of narratives about, human behaviour human behaviour in the past, the unknowns are in the past. The outcome of this critical analysis about what people did and experienced under will help us to (a) identify strengths and variable environmental circumstances and in weaknesses of the NOaA 1.0 as a tool for equally variable socio-cultural contexts. Of course development-led archaeology and (b) make there are other unknowns and problems to suggestions for improvement. address, for instance, in connection with archaeological patern formation. Understanding post-depositional efects on the initial 6.2 The research agenda: knowledge archaeological record is crucial for the domains and research perspectives interpretation of paterns as they are documented during feldwork. But this is not a goal in itself: the goal is to interpret the 6.2.1 People and Material archaeological record in terms of human behaviour, and for this it is necessary to decipher this record in order understand how it came to be. Chapter 3 focussed on the relationship between The investigation of aspects echoed in the people and materials. As this theme NOaA 1.0 is complex, and the previous chapters encompasses topics that are traditionally show that the degree of success is difcult to dominant in archaeological research, it is not determine. What is the scientifc gain of a decade surprising that a broader range of aspects could of development-led investigations in the be addressed. We used three headings, namely, Netherlands? The preceding three chapters craf activities, socio-cultural meaning and primarily focussed on the synthetic analysis of frames of reference, to discuss several topics to feldwork results produced in the context of which development-led research has made a development-led archaeology. From this it contribution, and in some cases has resulted in becomes clear that the gain of knowledge new insights. In many cases, however, we come relative to the NOaA 1.0 research questions, to the conclusion that insights are somewhat as well as several national reference incidental, and this may leave the impression frameworks,475 is highly variable. In many cases that this analysis consists of a haphazard we cannot take the narrative further than the collection of examples. When we categorize the anecdotal level; in some cases new insights could results into the knowledge domains defned in be gained due to the availability of abundant and section 1.2.5, it becomes clear that the gains of clearly published data (table 6.1). The purpose of our exercise are strongly determined by the this chapter is to critically evaluate the presentation of research results in reports. possibilities for synthetic analysis of the research When considering the spatial scale of results, not only in direct connection with the investigation, we were confronted with NOaA 1.0 itself, but also against the background problems of representativeness. Mostly, results of feldwork approaches and methodologies, as from development-led research can only be well as the practices of reporting. So why can assessed at the local scale. The sites that have particular aspects be synthesised to a reasonable been subject to development-led excavation degree, and others cannot? We will show that over the past 10-15 years were selected on the this is not simply a mater of critical mass. basis of what one might call ‘anticipated In order to maintain a clear connection with successes’. Basically, this involves the selection the research framework, and as we outlined in of sites for which, in consideration of survey 475 Presented in the edited volume chapter 2, we will develop our arguments under data, chances are high of fnding the expected. De steentijd van Nederland (Deeben et al. the umbrella of the overarching research themes Hence, the results from projects that were 2005) and in a chapter in The Prehistory of the Netherlands (Van den Broeke, Fokkens (chapters 3, 4 and 5) and spatio-temporal concerned with the investigation of ‘familiar’ & Van Gijn 2005). 202 —

Table 6.1 Contribution of reports to the themes and topics discussed in the chapters 3, 4 and 5. Reports which have not delivered any useful information are not listed. + minor contribution; ++ medium contribution; +++ major contribution.

People & Material People & Space People & Environment PUBID Production, use & discard of lithic tools Organic resources for artefacts Materiality Death & ritual fnal Typochronology and Palaeolithic Mesolithic Chronological & regional trends potery Lithic resources & Mobility Find scaters & arctivity areas Dwellings & other structures Subsistence & food processing Mesolithic pit hearths Occupation (dis)continuity Human impact 1 ++ 7++++ 15 + + 43 + ++ 49 + ++ 53 ++ ++ + ++ ++ 61 + 65 ++ 74 + 104 ++ 105 + ++ 122 + 126 + 142 ++ 143 ++ 147 ++ 161 ++ +++ 163 + + 170 ++ 174 ++ 194 + ++ 210 + + 211 +++ 216 + 219 + + + + 240 + 253 ++ + + ++ 255 ++ ++ +++ ++ 258 + ++ +++ 259 +++ ++ +++ ++ ++ ++ + +++ +++ +++ 260 +++ ++ ++ ++ + ++ +++ 265 ++ 272 +++ ++ ++ ++ ++ ++ + 308 ++ + 309 ++ + ++ 330 ++ 338 ++ + + + 340 + ++ 351 + 365 + ++ 373 + 392 +++ 395 + + + 407 + + ++ 462 +++ + +++ ++ +++ +++ ++ 463 +++ +++ +++ ++ ++ ++ +++ +++ 203 —

People & Material People & Space People & Environment PUBID Production, use & discard of lithic tools Organic resources for artefacts Materiality Death & ritual fnal Typochronology and Palaeolithic Mesolithic Chronological & regional trends potery Lithic resources & Mobility Find scaters & arctivity areas Dwellings & other structures Subsistence & food processing Mesolithic pit hearths Occupation (dis)continuity Human impact 467 + 468 ++ ++ ++ + + 475 ++ ++ 482 +++ 483 + ++ + + ++ 485 ++ ++ 492 + 504 + + 515 516 + 524 ++ 528 ++ ++ + ++ ++ + 537 + 538 +++ +++ +++ +++ +++ ++ +++ +++ +++ 542 + + 554 + 561 ++ 562 + + ++ 567 ++ 599 ++ 600 + + 617 ++ 651 + 725 + + 731 ++ + 732 +++ +++ +++ +++ +++ +++ +++ +++ +++ +++ 733 ++ ++ ++ + +++ +++ 737 ++ + + 738 + 739 ++ + + + 740 ++ ++ + ++ 741 + + + + +++ + +++ ++ 742 ++ + + + +++ 744 + + 745 ++ +++ + +++ + + 747 +++ 750 + ++ + + +++ + 752 ++ + + +++ 754 + 755 ++ 758 + 760 ++ 762 + 204 —

contexts have produced insights that mostly ft The technological modes of production that or confrm the existing picture. Situations in seem to be connected with these two groups which the outcome is uncertain are preferably – and perhaps other too – have received litle avoided, which leaves litle scope for the atention. Yet, socio-cultural transmission of evaluation of expectations or predictions. In technological knowledge and practical expertise other words, the outcome is representative of must underlie the persistent existence of these the known, but there is no way to tell how technological traditions, which themselves will representative it is for the unknown or the have been subject to variation and change. It is poorly known. This makes any inter-site or more than worthwhile to consider just these regional investigation difcult. Aspects of aspects in future analyses. Many projects have geographical diferentiation in, for instance, the yielded suitable material, but for this type of use of fint tools at a short-term scale remain analysis, it will be necessary to return to the difcult to investigate due to the lack of enough fnds since the reports do not provide the assemblages analysed in sufcient detail that information needed. date roughly to the same period. There are too Along the same lines, there is still much to many gaps. Nonetheless, the use-wear studies gain regarding ceramic technological traditions conducted to date, although they are as yet (section 3.4.2). Even more so than lithic studies, relatively few in number (section 3.2.1), have ceramic studies seem to be imprisoned in demonstrated considerable variability in tool automatism, leading to untransparent use (fint and other rock) and have also provided classifcations, which only seem to serve relative evidence for the working of materials that dating purposes. It will require a major shif for themselves were not previously known from the archaeologists to step beyond this superfcial archaeological record (shell and jet in the approach to ceramics, and move towards an Mesolithic). In addition, these use-wear analyses understanding of socio-cultural meaning of give archaeologists much to think about potery production and use. Research questions concerning the relationship between tool focussed on ceramics need to be more explicit typology and tool function; drawing inferences regarding modes of production and use, the about site function based solely on tool typology study of which can rely on various methods and is not straight forward. techniques that are readily available. Because However, despite these problems of such data are as yet barely available, use of representativeness, there have been gains in these techniques will allow archaeologists to knowledge at the regional and long-term scales. gain knowledge not only in the local and short- With regard to long-term regional trends in term domains, but also in the regional and long- Mesolithic tool types – microlithic components term domains. in particular – development-led projects have The possibility of increasing our insight into generated enough critical mass to deepen the socio-cultural meaning of material insights into typo-chronological developments (materiality) is not restricted to lithics and (section 3.4.1).476 Partly, this has resulted in the ceramics in terms of utilitarian objects. Several confrmation of the established phasing of the development-led projects have provided Mesolithic. At the same time, it is also evident information with regard to ritual or cultic that there is considerable geographical dimensions of objects and burial practices variability, with technological developments in (section 3.3). Generally speaking, the the northern Netherlands following a diferent manifestation of belief systems through the trajectory than appears to be the case in the treatment of objects and of the deceased are central and southern Netherlands, at least disconnected in archaeological studies, and during the Early and Middle Mesolithic. This may development-led projects are no exception mirror diferent technological traditions, herein; in many cases, fnds categories are a hypothesis that connects to the distinction decontextualised. The interpretation of made between the Northwest Group and the particular fnds contexts as ritual depositions Rhine-Meuse-Scheldt Group on the basis of may appear less straight forward than the 476 The publication of the reports on Well- Aijen and Kampen-Reevediep will typology many decades ago.477 The dataset identifcation of burials, but this is certainly not probably lead to further detail being available today provides a more solid basis for (always) the case. In the example of Roterdam- available. 477 Gob 1985. such a distinction, although it still shows gaps. Beverwaard, the Mesolithic cremation graves 205 —

were only identifed as such thanks to the other sources along the coast of northern France analysis of the small number of calcined bone and southern Scandinavia. Although the data are fragments. The presence of a perforated cobble scanty and observations are incidental, they mace head alone would not have led to the show that identifcation of potential raw same conclusion. However, it makes one wonder material sources provides new insights into the about comparable fnds contexts where bone geographical sphere of interaction in various remains are lacking. Hence, there is a need to time periods. Importantly, such insight draws pay particular atention to fnds context, even our atention to the degree of mobility of groups when objects could just as easy be explained as and to the social interaction between groups. In having been lef in a secular seting. The line the case of Heligoland fint in a Hamburgian between the secular and the sacred is probably context, for instance, it begs the question about not as hard and fast as we tend to assume. the origin of the particularly high-quality fint Indeed, depending of the cosmological context, that is found in many Hamburgian assemblages. there may be no line at all. New research will be needed to get a beter grip on these maters. However, it also demonstrates the importance of expertise. 6.2.2 People and Space Under the umbrella of the theme people and space, most of the acquired results relate to the local use of space, that is, the intra-site level. The relationship between people and space was Spatial information about fnds distributions addressed in chapter 4. In contrast to the and features is provided in all of the reports situation with people and materials, the included in this study, but analysis of data and situation here is somewhat ambivalent subsequent interpretations of paterns remain regarding the gain of knowledge. The regional superfcial. Standard research questions dimension of space appears to be one which can regarding the nature of activity areas and the barely be addressed. We cannot reasonably presence of built structures appear to be asses knowledge gain in relation to research addressed in only a basic manner. Eforts to questions related to variability in landscape use increase our understanding of site-formation at the short-term and long-term scales, no processes and palimpsests are an exception, mater how valid these questions are. As despite the fact that the overwhelming majority indicated in section 1.2.2 and section 4.2, of sites excavated manifestly result from long potential new insights into such topics as occupation histories and complex depositional setlement location choice and setlement and post-depositional processes. systems can only be evaluated when hitherto The main reasons for the present situation underrepresented landscape zones and are not related to the research question geographical regions (including the North Sea themselves, nor to how data were collected basin) are investigated. However, because survey during excavation; in fact, many sites have strategies largely build on the known produced datasets that are potentially suitable archaeological record through predictive models for rigorous analysis (table 6.2). Instead, the (section 4.2.1), and sites selected for excavation reasons are as follows: frst, the majority of sites ft those models, underrepresented zones and investigated are palimpsests, which complicate regions are simply not investigated. In addition, the deciphering of the archaeological record at deeply buried sites are rarely excavated, and the site level; second, there is a lack of thus underrepresented. Hence, we are faced application of suitable spatial methods and with a systematically biased record. techniques. The two reasons are related to some Nonetheless, to some extent, we can draw extent: palimpsests are difcult to unravel, and inferences with regard to regional aspects of there is no clear-cut analytical approach. In most landscape use in connection with mobility. cases, spatial paterns within fnds distributions Development-led projects have permited are mostly only subject to visual inspection of archaeologists to gain insight into long-distance maps that present various atributes of fnds transport of raw materials, notably red categories. Statistical approaches are almost Heligoland fint (Late Palaeolithic), Wommersom lacking, as are multiscalar techniques, which quartzite (Mesolithic), and possibly fint from permit archaeologists to investigate spatial 206 —

Table 6.2 Methodological approaches to spatial aspects in reports and suitability of datasets for future analysis.

PUBID Focus of Spatial Answers the Applies Excavation Excavation Suitability Future analysis Remarks questions questions statistical/ method units possible spatial methods 15 Flint concentrations in Superfcally Statistical testing Corings and grid 1 m squares Unsuitable Limited due to Focus given to relation to borehole and for varience excavation (0.5 m segments) the methodology the central grid excavation borehole 253 Artefacts concentrations Superfcally Interpolation of Grid excavation 1 m squares Unsuitable No due to in relation to possible artefact (0.5 m segments) sampling structures distributions strategy 259 Artefact concentrations Superfcally interpolation Sampling every 5m. 1 m squares Unsuitable Yes Interpolation in relations to activities (kriging – with Block 7 continuous applied on a standard area excavation continuous deviation) dataset 260 Artfeact concentrations Yes Interpolation Grid excavation 1 m squares Unsuitable Limited due to in relations to activities (sampled) sampling 732 Artefact concentrations Yes Moving Average Grid excavation 1 m squares Suitable Yes The use of MAW in relations to activities Windows keeps interpretations on a general level 752 Chaine operatoire of Yes Interpolation Grid excavation 0.5 m squares Unsuitable Yes Interpolation lithic production, site applied on formation processes, a continuous palimpsests dataset 745 Artefacts concentrations Superfcally Interpolation Grid excavation 1 m squares Unsuitable Limited No incorporation in relation to possible (sampled) (0.5 m segments) of evaluatation structures data 468, 738, Artefacts concentrations Yes, but only Undetermined Boorholes, test 1 m squares Undetermined Undetermined Applied mehods 739 in relation to possible because the trench and grid not described structures questions are excavation limited 538 Finds distribution and Partially Nearest Neighbour Grid excavation 1 m squares Unsuitable Limited Misuse of assessment sampling and Moving (sampled) (0.5 m segments) sampling and strategy Average Windows interpolation 543 Artefacts concentrations Not possible No Grid excavation 1 m squares Not available No in relation to possible due to the (sampled) (0.5 m segments) structures methodology 463 Artefact concentrations Yes, but only Visual inspection Grid excavation 1 m squares Not available Yes in relations to activities because the questions are limited

paterning amongst diferent fnds categories. but then again, this new output requires As a result, in the majority of cases, spatial interpretation. Without interpretation, this data interpretations are built on plain descriptions of is not necessarily less ‘impressionistic’. Although clusters of materials. Connecting fnds visual inspection of distribution maps is a useful distributions to features and the identifcation of frst step towards archaeologically meaningful built structures equally sufer from a lack of interpretations of spatial paterns, only in-depth thorough analysis. Here, too, we see the use of analysis of reliable excavation data will help us visual inspection of distribution maps (see to develop a deeper understanding of the section 6.3.4). interrelationships of data and can potentially This is not to say that more elaborate lead to the discovery of other meaningful spatial methods and techniques will lead to new paterns that would otherwise remain hidden in insights by defnition. The nature of spatial the data. The approaches are complementary, paterns in the archaeological record is highly for which reason we favour the application of complex, and the application of elaborate multiple techniques – visual, statistical, analytical techniques is equally beset by multiscalar – to potentially increase insights into problems of interpretation. Statistics-based and short- and long-term dimensions of the use of multiscalar approaches produce new output locations. Hence, the research questions in the through the transformation of input data, NOaA 1.0 remain topical. 207 —

specialists in the feld leads to exciting results. 6.2.3 People and Environment Such apparently dull categories as charcoal and other charred plant remains can provide information that was not easily accessible until The third and fnal theme, addressed in chapter recent years. Although the number of 5, concerns the relationship between people and assemblages is still small, insight into the environment. Again, this theme is covered by exploitation of a range of plants for food has several research questions and topics in the increased considerably, most notably with NOaA 1.0. Evidently, the chances of regard to wetland species, the remains of which archaeologists acquiring new insights are have been found in charred condition. Research directly dependent on the availability of regarding the use of Mesolithic pit hearths has environmental proxies and materials, which for resulted in the development of a new uncharred organic mater is determined by hypothesis, notably their potential function in preservation conditions. Clearly, there will be a the process of wood-tar production. bias towards wetlands, as far as these have been Notwithstanding the fact that several subject to excavation at all. The number of data potential interpretations of the pit hearth data points is restricted, and this hampers any efort are possible, the importance of the reported to progress our knowledge. Archaeobotanical results carries further than just the possible use remains have been subject to synthetic analysis of the enigmatic Mesolithic pit hearths. Because in recent years,478 whilst few archaeozoological these features are prominently present on many assemblages have become available during the sites north of the Rhine throughout the period under review. Nonetheless, some aspects Mesolithic, the question arises how diferences have been shown to deserve atention, notably in time and space can be explained, for instance, with respect to the use of animal and plant in connection with climate-driven models about resources (section 5.2.1). vegetation succession. Other related questions Bone, tooth, and antler, mostly calcined and refer to the time-depth of locational use – this mostly in small numbers, appear to be found can span more than 2000 years – in connection more ofen at Mesolithic and Neolithic sites than with such pits. Analysis of this type of feature at one tends to think. Although the information the site level thus signifcantly adds to the that can be extracted from these small investigation of questions at various spatio- assemblages is relatively restricted, the evidence temporal scales. adds useful information to the overall still poor The same holds, to some extent, for record dating to early prehistoric times. research into human impact on the Furthermore, animal bones whose environment. Traditional models about the anthropogenic origin is questionable can still restricted infuence of early prehistoric people, provide information about the composition of mostly hunter-gatherers, on their environment the ‘animal landscape’ in terms of species are now being questioned. Development-led diversity. This is a much needed source of research again has made a considerable information, as reference frameworks for contribution to this debate, mainly because of subsistence strategies are largely founded on an open-minded involvement of specialists. actualistic models of animal behaviour and The methods applied are traditional, but the communities. Yet other phenomena, such as frameworks of reference have changed under arrangements of wooden stakes in stream beds, the infuence of an international debate. In fact, provide information about food procurement development-led work in the Netherlands is at techniques, notably fshing techniques. 14C-dates the forefront of these new developments. of such stakes point to surprisingly signifcant Therefore the fascinating results should not time depths (see section 5.2.1). remain hidden in grey literature reports. The Development-led research has made a publication of such results in scientifc papers by considerable contribution to our knowledge professionals who work in development-led about hunter-gatherer plant (food and non- archaeology is, however, frustrated by the food) resources. Here we see that the dictates of commercial continuity. application of modern techniques of analysis by 478 Out 2009. 208 —

It is important to note that recent years can contribute to the answering of research have seen the development of innovative questions in the NOaA 1.0. This potential can still techniques that use data sources that also are be tapped, but only by returning to the fnds and (or can be) found in the Netherlands. Analysis of the primary data, which is far beyond the scope ancient DNA is potentially possible; both the of the present study. Hence, for the most part, extraction and the analysis of aDNA from fossil it is not so much the research questions bones have witnessed great technological themselves that are the problem, but, rather, improvements. A recent study of skeletal their embedding in project outlines and remains from Bell Beaker contexts in Europe strategies. In this section, we identify the most – including the Netherlands – demonstrates the pressing issues that, in our opinion, need to be potential for genetic studies.479 Equally, the addressed if the next decade of research is to potential presence of aDNA in buried provide new insights. We will do so by following palaeosols – sediment aDNA – may shed light on the various steps of the research process animal, plant and human populations in ‘extinct (fg. 6.1) to identify where problems exist. landscapes’, which are so far only partly Each section includes a brief description of the represented in the fossil physical record.480 standard procedure, a description of the Another source of information concerns stable problem, and a proposal to solve the problem. isotopes – notably 15N and 13C (the later of which Ofen the solution needs to be found in the is normally measured for the purpose of frst step of the research process, namely, the 14C-dating). Although it has been known for a project outline. Within an archaeological ‘market’ considerable time that stable isotope analysis in which competition is primarily expressed in can yield important information about human fnancial terms, commercial frms are evidently and animal diet and habitat,481 such analysis is hesitant to do more than is required. The quality not conducted in the context of development- of the project outline therefore determines the led research. Ofen not even the 15N and 13C quality of the research project to a large extent.483 values are provided. A recent study of isotope We feel that it is the shared responsibility of the data from Mesolithic human remains has shown archaeological community to determine that even contextless fnds, in this case from the acceptable qualitative baselines for North Sea, can provide meaningful information development-led research. As we express in our about human diet.482 introduction (section 1.1), the purpose of archaeology is to develop insights about human behaviour in the past from a unique but 6.3 Connecting to development-led vulnerable source of information. It is not about research in practice solving a ‘problem’ for economic development today. Hence, the quality of the project’s project outline should be the frst concern, and this 479 Olalde et al. 2017. 480 Recent examples involve the question of The preceding sections make clear that a decade requires the adequate involvement of expert the introduction of cereals in Britain of development-led archaeological research knowledge right from the beginning. (Smith et al. 2015), the interplay between foragers and Neolithic farmers concerning the early prehistory of the (Bollingino et al. 2013), and the co- Netherlands has yielded limited gains and that existence and/or presence of ^ƚĞƉ WƌŽĚƵĐƚ Neanderthals and Denisovans on sites, we ofen remain stuck at the anecdotal level. e.g. in the absence of fossil bone Returning to the title of this report, it appears remains (Slon et al. 2017). Step1.Researchplan Predictionofarchaeologicalvalues 481 For an example from the Netherlands, that paradise is, indeed, lost. The hopeful see Smits & Van der Plicht (2009). Other expectations around the NOaA 1.0 with regard to Step2.Prospection Presenceofarchaeologicalsite examples concerning Mesolithic populations in Europe have been safeguarding archaeological heritage and Step3.Excavation Datacharacteristics published since the beginning of the making considerable contributions to our twenty-ﬁrst century (e.g. Schulting 2005, 2010; Schulting & Richards 2001, knowledge about the early prehistoric past of Step4.Analysis Interpretation 2002). the Netherlands have not come true. 482 Van der Plicht et al. 2016. 483 More than a decade ago, Bazelmans et al. Nonetheless, there is much potential, which Step5.Report Registration 2005 (2005; see also Lauwerier et al. means that part of paradise is still waiting to be 2006) published a report on the quality of the PvE. His conclusion was that discovered. Especially reports on the more Figure 6.1 Schematic representation of the research project outlines were chary in their extensive projects make clear that in-depth process within the context of development-led research ambition. This still appears to be the case. study of fnds categories and excavation data archaeology in the Netherlands. 209 —

It is important to note that the organisation of cultural space at a landscape archaeobotanical research did not raise any scale. We do not have a confrmed serious problems that we feel to be necessary to representative record – a quantitatively and address here, neither in terms of applied qualitatively sound sample – of spatial methodologies and techniques nor in terms of behavioural systems (i.e. the range of the quality of analysis itself. It is our impression behaviours) in early prehistory. This is not only that archaeobotanical contributions to the case in the Netherlands, but also elsewhere, development-led research do not sufer from indeed, anywhere in the world; it is inherent to the faws that contributions relating to other the nature of the archaeological record.484 In the groups of materials sufer. The high quality is Netherlands, however, predictive models of site atributed to the fact most contributions are location provide important input for the design produced by a relatively small community of of research strategies in the context of highly qualifed specialists, who tend to easily development-led archaeology. Straight forward share ideas and expertise. This community may translation of model expectations easily leads to serve as an example for other specialists, who, self-fulflling prophesies, since models are like, archaeobotanists, are required to meet the mostly built on statistical correlations between formal standard of training and experience. archaeological observations (a biased dataset) and derived palaeolandscape characteristics (geomorphology, soil characteristics and 6.3.1 Predictive models hydrology). As a consequence, new data points will primarily confrm the model. A site containing, among other things, fint microliths One characteristic of development-led on a relatively pronounced sand ridge next to a archaeological research in the Netherlands is the low-lying area that is assumed to have carried use of predictive models, which are based on the water will end up in the report as ‘a Mesolithic idea that people did not behave randomly in the hunting camp’, full stop. landscape. The rationale is that resources were The problem of correlative or inductive exploited using various strategies and modelling for heritage management purposes technologies based on (expected) availability, has been discussed extensively over the past experience and tradition. Many landscapes 15 years or so.485 It has been argued that such ofered a rich reservoir of potential resources, models generally lack the ‘cultural factor’ but this does not mean that the entire range needed to account for spatial and contextual actually was exploited. The landscape as it was variability that is only partially trapped in the known to one or more groups of people was archaeological record. Hence, deductive models, fundamentally diferent from the reconstructed based on hypotheses of human spatial ‘landscape’ into which we project our early behaviour, should be given priority.486 So the prehistoric people. Diferences, which we cannot basic idea is to ‘predict’ where archaeological even defne, concern aspects of geographical remains of some particular nature are scale as well as ‘contents’. Which parts of the potentially present, in consideration of a geographical space that we have defned as landscape model and some aspect of researchers were actually known by groups of behaviour,487 for instance, the hunting of large people? And what can we actually say about mammals. However, both approaches bear the variability in resource availability in time and problem of inherent uncertainty, an issue that space? What drove people to exploit particular hitherto has received litle atention.488 This 484 Brouwer Burg, Peeters & Lovis 2016; resources, or to not exploit them? Cosmological uncertainty is partly related to the biased nature Lovis 2016; Whitley 2016. and ideological considerations, for instance, may of the archaeological record. However, other 485 Several volumes have been published on this topic, by Van Leusen & have directed the choices that were made. sources of uncertainty relate to systemic Kamermans (2005), Kamermans et al. Hence, with regard to the archaeological interrelationships of variables, parameter (2009), and Verhagen (2007). For an in-depth review, see Verhagen & knowledge about how people behaved in the setings and the theoretical baseline for a model. Whitley 2012. landscape, we face many uncertainties. Any theory (read: hypothesis) about how people 486 Verhagen 2007; Verhagen & Whitley 2012. In addition, the archaeological record is like behaved in any sort of landscape seting is built 487 Peeters 2007; Verhagen 2007; a Swiss cheese, full of holes, and extremely on prior archaeological knowledge (which is Whitley 2000. 488 See papers in Brouwer Burg, Peeters & biased where it comes to our knowledge about biased) or makes assumptions about what Lovis 2016. 210 —

people would decide to do under certain normally has implications for the overall circumstances. Hence, for the purpose of planning and budget. These factors in particular heritage management, deductive models are not frequently lead to decisions that turn early necessarily beter than inductive models.489 prehistoric remains into ‘by-catch’, which is From the above, it could be concluded that reported on to only a minimal extent. Yet it is the use of models – be it formal (computational) precisely this by-catch which may represent low- or informal (as narrative) – for the development resolution phenomena (low-density fnds of research strategies in development-led scaters, isolated pit hearths, isolated projectile archaeology should be avoided. However, this points) which are otherwise underrepresented in does not need to be the case. Predictive models the archaeological record. can be useful as a source for archaeological theory building if feldwork is actually considered to be a form of model validation. In 6.3.2 Prospection: augering depth and this perspective, research should not only focus sampling grid on acquiring information that fts or confrms the model, but should also consider ‘negative’ outcomes. In other words, absence of evidence In many projects, and in many parts of the also calls for explanation; it may very well be Netherlands, preliminary stages of research that our underlying assumptions are wrong. include surveying by means of augering. Going back to the ‘inevitable’ hunting camp, for Augering is ofen manual, unless potentially instance, the model may predict that a location interesting layers are known to be present at to be excavated has a good chance of bearing several metres below the present-day surface. remains of such a campsite, and ofen this will In theory, the augering depth is limited to the prove correct. However, if such remains are not depth of the proposed spatial development, that found, contrary to what was predicted, this is to say, the anticipated depth of disturbance.491 requires explanation. At the individual site level, As a result, to minimise costs, augering will not we cannot expect to fnd such an explanation, reach much deeper than that. In addition, within other than, for instance, post-depositional the strict context of the development plan, it is efects on site preservation. Reporting on not considered relevant to know if ‘negative outcomes’ in reference to expectations archaeological remains are present at a deeper is, nonetheless, valuable for the purpose of level. Furthermore, surveying by means of model evaluation. Models quickly become too augering is largely focussed on deposits dating generic, due to generalised ideas about how from the Holocene, and sampling will only early prehistoric people operated in the include the top of underlying deposits from the landscape, and what drove them to act. As a Pleistocene, which, in the Netherlands, are result, variability is barely considered. However, mostly coversand. the sheer number of potential behaviours As a consequence of this standard underlying the initial formation of the procedure, one specifc group of archaeological archaeological record is almost endless,490 as is remains systematically receives too litle the range of post-depositional processes, atention: Middle and Late Palaeolithic remains including research methods and strategies. covered by younger Pleistocene sediments. Sites 489 It should be noted that formal Hence, if archaeologists would like to use dating to these early periods of human activity modelling (i.e. computational modelling) for the purpose of results from development-led research in are perceived as too difcult to detect. This, exploratory research oﬀers a powerful connection with research questions that refer to however, does not need to be the case when tool with which to investigate sources of paterning in the archaeological a landscape scale of people’s behaviour, it will experts on these maters are consulted. Two record (Peeters & Romeijn 2016). be necessary to make expectations more projects (Den Bosch-Vonk en Vlam, 490 Van der Leeuw 2016. 491 Disturbance is not restricted to physical explicit. Next, we will need to establish to what Amstenrade-Buitenring Parkstad Limburg) disturbance due to (machine) digging. extent results ft these explicit expectations. demonstrate that Palaeolithic sites can be found Lowering of the groundwater table, for instance, also involves disturbance, Diverging results can involve absence of the during development-led archaeology.492 As yet, notably of the hydrological and expected or presence of the unexpected. The anticipated atention for Palaeolithic sites geochemical conditions. 492 Both fall outside the time range that later readily causes problems if identifed remains exceptional, and as a consequence, was to be analysed within the present during feldwork, as a change of strategy and fnds fall systematically in the category ‘by- project; also, the reports are not yet available. methodology is ofen needed, which in turn catch’. Many development plans do, however, 211 —

result in the disturbance of Pleistocene deposits, probabilities of sites characterised by fnds for instance, due to piling for foundations of scaters have demonstrated that only high- buildings. Any Middle and Late Palaeolithic sites density scaters of at least a medium-sized that are present are thus disturbed and horizontal extent – that is, palimpsests covering inaccessible for archaeological research. more than 1000 m2 – can reasonably be detected In order to improve the situation with in a 20 × 25 m sampling grid (table 6.3). Scaters regard to the Palaeolithic, any project outline with lower densities of material and of smaller related to development that is expected to size require denser sampling grids (4 × 5 m). involve disturbance (piling, trenching, dredging) Detection of scaters <50 m by means of auger of Pleistocene deposits should explicitly include sampling is almost impossible. But the reality prospection for Middle or Late Palaeolithic sites, shows that even the detection of larger sites is depending of the geological situation at hand. difcult, particularly because materials are Hence, the project outline should make explicit distributed unevenly over a surface; average or that surveys, whether by means of augering or random distributions do not exist.494 Therefore by means of trenching, should include auger sampling as part of a standard Pleistocene sediments to a depth that can prospection strategy is a disadvantageous reasonably be expected contain archaeological starting position for the discovery of the kinds of material and be disturbed. A good starting point smaller, high-resolution sites (e.g. low-density may form the mapping of palaeosols, such as ‘single event’ scaters) that we would like to those encountered in loess deposits and Late have more information on in order to answer Glacial coversands. many research questions in the NOaA 1.0. The Similarly, the detection of early prehistoric only way to increase the chances of locating such sites of Mesolithic and Neolithic age by means of sites is to anticipate their potential presence auger surveys faces many problems.493 Various (e.g. based on palaeolandscape information) in statistical studies that have considered discovery situations where terrestrial deposits dating to

Table 6.3 Recommended survey strategies for Stone Age sites (adapted from Verhagen et al. 2013).

Type Core sampling grid Auger diameter Observation technique Very small (<50 m2) Low fnd density - - - (40-80 per m2) Very low fnd density - - - (<40 per m2) Small (50-200 m2) Low fnd density 4 x 5 m 15 cm 3 mm sieving mesh (40-80 per m2) Very low fnd density 4 x 5 m - - (<40 per m2) + test pits Medium-sized (200-1000 m2) Medium-high fnd 13 x 15 m 12 cm 3 mm sieving mesh density (>80 per m2) Low fnd density 8 x 10 m 15 cm 3 mm sieving mesh (40-80 per m2) Very low fnd density 4 x 5 m - - (<40 per m2) + test pits Large (>1000 m2) Medium-high fnd 20 x 25 m 12 cm 3 mm sieving mesh density (>80 per m2) 493 Tol et al. 2004; Verhagen 2005; Verhagen Low fnd density 13 x 15 m 12 cm 3 mm sieving mesh et al. 2011, 2013. 494 Verhagen et al. 2013; also see Katenberg 2 (40-80 per m ) et al. 2008; Peeters 2007. 212 —

the Pleistocene or Early-Middle Holocene are such as sediment layers, are continuous three- reached. In practical terms, the use of dimensional phenomena, artefacts as composite mechanical shovels – which are almost always elements of sediment layers are not. used in the Netherlands – should be controlled Archaeological fnd-scaters are most ofen in such a way that encounters with such sites do approached as two-dimensional paterns, whilst not lead to instant destruction. Furthermore, in reality they are three-dimensional. Indeed, consultation of specialists remains important. each and every artefact can only take up as much space in a sample volume as the volume of the artefact itself. Hence, if one wants to take a 6.3.3 Excavation: phased strategies purely quantitative approach, the fnds density of a sample unit is beter expressed in terms of volume (whereby weight can be used as a Stone Age sites are characterised by the proxy). Also, it is important to note that Stone presence of a scater of fnds, in contrast to Age sites that have been extensively excavated younger sites, where features are encountered (e.g. Hoge Vaart-A27, Dronten-N23, Ede- more ofen and are the dominant focus of Kernhem and Schipluiden) demonstrate feldwork. By defnition, fnds scaters comprise substantial quantitative variation regarding the zones with higher and lower densities of number of fnds within high- (or low-)density materials. Within development-led projects, the zones, even at short distances. This implies focus is ofen on excavating the high-density intrinsic uncertainty with regard to quantitative zones, and not ‘wasting’ time and money on interpolation of archaeological sample data – relatively empty zones. To maximise the uncertainty that is not easily accounted for. And outcome of the excavation, apparently defned because, by defnition, relatively empty zones in terms of quantity of fnds, a phased are lef unexcavated, there is no check to excavation strategy has become common in determine whether our selection was correct. Dutch archaeology. In this strategy, excavation This problem was made explicit for starts with the excavation of a small grid cells Dronten-N23, where the fnds densities with a relatively large distance between them; predicted on the basis of the interpolation were clear examples are Dronten-N23 and compared with the fnds densities from the Ede-Kernhem (fg. 6.2). This sampling strategy subsequent excavation (fg. 6.3). Although the has replaced sampling by means of a large (12 or diferences between expected and observed 15 cm) auger, a practice which had become part grid-cell values roughly average out, the of Dutch feld methodology.495 Diferences in distribution paterns are diferent. fnds numbers are used to delineate fnds- This implies that quantitative variability density clusters. Low-density zones are mostly within sites can be signifcant and also difcult excluded from further research. This appears to to predict.497 be an efcient strategy if maximisation is a The second concern is of a very diferent function of numbers of fnds relative to the nature, but does build upon the frst. The budget available. Unfortunately, research results quantity-driven selective approach to excavation are not by defnition optimised when price per results in patchy and discontinuous excavation fnd is minimised: research quality is not based plans (see fg. 6.3), in which clusters or on the number of fnds but, rather, on the quality concentrations of fnds are ofen only partially and information value of the data. In this excavated. This leaves litle potential for the 495 There is no fundamental difference respect, two concerns need to be addressed. investigation of spatial aspects, even though between sampling in small grid-cell units and auger sampling in terms of The frst concern is that the use of grid-cell these are explicitly part of several research information value regarding find densities to predict variation in the density of questions in the NOaA 1.0. Therefore, it is density, as long as similar sampling depths are used. the fnds scater is methodologically important to carefully consider the 496 Isaaks & Srivastava 1989; Chilès & problematic. We must remember that fnds consequences of choices made with regard to Delfiner 2012. 497 In the recent case of Kampen-Reevediep scaters are of a fundamentally diferent nature the collection of spatial data: which avenues are (Mesolithic), a different approach was than distributions of soil or geological units, for kept open, and which are cut of. Many chosen. Instead of selecting delimited concentrations for excavation, several which point observations can be interpolated techniques for spatial analysis are available, and transects were investigated, which also relatively securely by means of geostatistical could be used, but their applicability requires included ‘empty’ zones. The preliminary results are very promising. techniques.496 Whilst soils and geological units, continuous areas of fully excavated grid cells, 213 —

Figure 6.2 Lef: sampling grid and extrapolation of the quantity of ﬂint for Dronten-N23 (from appendix II: PUBID 538). Right: sampling grid and extrapolation (inset top right) of the quantity of ﬂint for Ede-Kernhem, and the clusters recorded during subsequent excavation (from appendix II: PUBID 745). 214 —

preferably including empty zones to minimise grid cells within the constraints of time and edge efects. The meagre gain from money, instead of a calculation based on development-led projects regarding questions potential information value. about the spatial organisation of early Whilst a phased excavation strategy prehistoric sites may partly be a consequence of remains smart, it is proposed here that the this ‘fragmentation’ of spatial data. above concerns may be addressed by changing Furthermore, the tight delimitation of zones that two aspects of the strategy. First, in order to are selected for excavation seems to involve a account for localised variability in fnds simple calculation of the number of allocated densities, the frst phase should be intensifed so 215 —

Figure 6.3 Diﬀerences between expected and recorded quantities of ﬂint for Dronten-N23 (from appendix II: PUBID 538).

as to increase the predictive value of interpolation models. The intrinsic uncertainty 6.3.4 Analysis is related to uneven distribution of fnds within isolated grid cells – fnds that also occupy variable volumes within a unit of sediment. The phase following the excavation involves the Second, within large projects, such as analysis of the collected data and fnds Dronten-N23, it may be wise to focus the categories. Our analysis of results from excavation on some fnds clusters and excavate development-led projects has identifed several them ‘fully’ instead of excavating more clusters, issues that need further atention in this post- but only ‘partially’. However, making decisions excavation phase. about which clusters to excavate and how should frst of all be a concern about potential Standards in material culture analysis information value: Which research questions The description and analysis of cultural material is have the best chance of receiving an answer not standardised, which is a serious hindrance with data expected to be obtained from the area when one intends to use the results of an analysis to excavate? And where do we potentially have for new research. For a part this is the result from the opportunity to obtain new insight into the fact that the categorisation of cultural questions that are less easily answered? remains is subject to variability due to specialists With respect to questions requiring spatial data, who may use diferent classifcation frameworks 498 Peeters 2007; Spikins et al. 1995. a well-thought-out strategy for the collection of and defnitions. As a result, defnitions of cultural 499 Initiatives to overcome this situation have been undertaken recently by the those data in consideration of methodological groups may change through time. This situation foundation Specialisten Archeologisch prerequisites remains important. The use of is complicated further by the absence of a Materiaal Platform (SAMPL), a platform for specialists in archaeological 50 × 50 cm grid cells as a basic unit for fnds national forum to discuss standards in the materials. Its objectives are the collection potentially enables high resolution description and analysis of ceramics, fints and improvement of the quality of material culture studies; knowledge analysis.498 Coarser grid-cell units, such as 1 × 1 m other stone fnds.499 These problems lead to the dissemination (towards professionals squares, do not permit fne-grained conclusions that (a) the reports under study are and the public); and protecting the interests of specialists. Several interpretations of spatial paterns, e.g. in terms strongly infuenced by the author’s personal guidelines have recently been published of activity areas and post-depositional choice in terminology and (b) reports from with regard to inorganic materials (htps://www.sikb.nl/archeologie/ kna- processes. diferent authors are difcult to compare. leidraden). 216 —

Some examples may help to understand the the time the pilot results come in, litle time (and efect of these problems. A frst example is that money) remains to expand the study without of the categorisation of stone tools. Whilst some hindering the two-year term for fnalising the authors prefer to talk about querns, others use report. the term grinding stones. There is no way to Whilst this situation is understandable from know whether they describe the same type of a managerial and a fnancial perspective, object and are, in fact, synonymous. Because it means that in most projects relating to earlier most of these tools are not depicted in the prehistory, site function is determined on the reports, the only way to use the reports for basis of lithic tool typologies, without further study would be to extract the fnds from questioning the correlation between tool the repositories. A second example concerns the typology and tool function. As a result, terminology related to Lanaye fint. In some typological classifcations still dominate our publications, the author distinguishes between understanding of the past. Functional analysis of the Rijckholt and Banholt subtypes, whereas in potery and stone tools is even rarer. Whilst others is the author only uses the term Rijckholt. many projects consider site function an Is this then used as synonym for Lanaye, or is the important research topic, potential information author of the opinion that the Banholt subtype sources are largely lef understudied. is absent? A third example relates to the use of To improve the situation and make it the terms Vlaardingen, Stein and Vlaardingen/ possible to tap the enormous potential, it is Stein, and the discussion about the application important to design projects that include more of ‘culture’ or ‘group’ as ordering principle. All specialist research on fnds. Clearly, this needs to three feature in the reports studied, without any be realised within the constraints of time and reference that makes clear why one particular money, and will require archaeologists to make term is used, or what the arguments are for a diferent choices, in favour of post-excavation preference for a particular option.500 analysis. The drawback is that less money is then The above are merely examples – the lack available for feldwork, which is not desirable, of any standards concerning typology, since excavation can only be done once. Ideally technology and raw materials makes one would like to see strategies where the comparison and synthetic analysis of the results analytical potential of the diferent fnds very difcult. categories is assessed in the early stages of the Stimulating the experts to develop project, e.g. the frst excavation phase of spatial standards will help to overcome many problems. sampling, so as to provide a qualitative basis for Diferences in opinion will, however, always decisions to be taken. During excavation, the exist to some extent. These standards should concern should not be to collect as much include all aspects (typology, technology, raw material as possible, but to collect enough material) and lead to consistent and mutually material accompanied by good contextual data comparable datasets for analysis, and provide to enable meaningful analysis. Pilot studies on a basis for standardised data presentation in the function of ceramic, fint and stone tools reports. In addition, expert meetings provide need to become full-fedged. Now that they a venue for exchanging ideas and discussing have shown sufcient success, they should be analytical aspects; eventually this may lead to expanded to add relevant and new insights. shared approaches and best practices. Spatial analysis Functional analysis Research questions concerning intra-site Functional analysis of ceramic, fint and stone paterns are standard on many Stone Age tools is almost absent in develop-led research projects. As a consequence, spatial archaeology in the Netherlands. Pilot studies are information is presents in most reports, in one sometimes undertaken as part of the larger form or another. The general lack of dedicated projects, but it appears that this is only an chapters or sections for spatial analysis appears appetiser, and never develops into a main to suggest that spatial analysis is, in general, course. Possibly this is the result of the fact that underrepresented and undervalued in Dutch 500 Compare Beckerman & Raemaekers 2009. the results of such studies are uncertain, and by development-led archaeology. 217 —

Occasionally, spatial analysis is conducted examples where considerable contributions by material specialists, but in fact this mostly could have been made with a relatively small concerns no more than visual inspection of plain investment. distribution maps, instead of the application of First of all, it has to be acknowledged that spatial techniques. However, many of the spatial there is litle expertise on the Middle Palaeolithic questions from the NOaA 1.0 are more difcult within the Dutch companies. Few archaeologists to answer than one would think, requiring in- in the Netherlands are trained in recognising depth analysis and assessment of interpretive artefacts dating to this period, and even fewer possibilities. Much of the observed work is are capable of assessing the potential superfcial, and merely leads to suggestions or information value in a broader geographical and indications of paterns, without further chronological context. As an example, we refer discussion of how the paterns can be to the project Oentsjerk, for which two Middle understood in terms of human behaviour and Palaeolithic artefacts were reported, and post-depositional formation processes. Indeed, registered as such in the national database clustering of material may indicate activity areas, Archis.501 Both fnds are, however, of natural but how does this relate to everyday life, to the origin and represent pseudo-artefacts. social use of space? What is the relationship Consultation with a specialist would not have between artefact distributions and features been time consuming, and would have (individual features, feature groups, structures, prevented an erroneous dot on the map. setlement)? These are questions that are, on The second example comes from the whole, lacking in the reports under study. section 5.2.1, where Mesolithic faunal remains When so many research questions relate to are discussed. The overview makes clear that on the use of space, opportunities to improve many occasions these remains have not been current practices are easily found. First of all, it is analysed, whilst they ofen provide the only best to incorporate a dedicated chapter or evidence of animal food and/or natural habitat. section on spatial analysis in reports, allowing all Animal remains dating to the Mesolithic – and spatial data to be analysed in an integrated Palaeolithic – are extremely rare, and every manner. Furthermore, the integration of a opportunity to extract information from even specialist in spatial analysis, next to specialists the scantiest remains should be taken if we want on material culture studies, provides a sound to obtain new insight into hunter-gatherer basis for interaction and development of subsistence and exploitation. It appears that the interdisciplinary interpretive frameworks. scientifc relevance of this somewhat unimpressive-looking fnds category is Expertise on early prehistory underestimated. Development-led projects strive for the best The third example comes from section 5.2.2, research results within a framework defned in where the use of pit hearths is discussed. This terms of time and budget. Where early landmark feature of the Dutch Mesolithic does prehistory is explicitly the subject of their receive ample atention in the context of the research, it is evident that researchers involved bigger projects, such as Dronten-N23 and have ample experience with the investigation of Knooppunt Hatemerbroek, where specialists in sites dating to this period. Choices are of course early prehistory are involved. Typically, the made within the constraints of time and money, interaction amongst specialists on these projects but they are nonetheless based as much as demonstrably leads to sound sample assessment possible on scientifc considerations. In contrast, and selection procedures, in particular for the projects in which early prehistoric remains are purpose of 14C-dating. Such opportunities are found as a by-catch, it is in a way logical that mostly missed on projects where early prehistory litle efort is spent in making the best out of is less the focus of research; samples for these remains, as the expertise of researchers 14C-dating are taken without prior assessment of involved is ofen limited or non-existent. It is no their suitability. However, relatively small steps in surprise, then, that in the context of our study, the organisation of the research could lead to a we feel compelled to make critical remarks on major contribution to our knowledge of the missed opportunities. We will present three period. For example, project managers should 501 Appendix II: PUBID 771. 218 —

consider involving archaeobotanists, who are able to determine wood species in an easy way, 6.3.5 Reporting at litle cost. Not only will this lead to more reliable sample selection, because 14C-dates remain a high-cost item on project budgets, but Above, we proposed that the analysis of material it will also add to studies on vegetation and culture could be improved by developing wood selection. standards. This improvement should also be refected in the reports, which are currently highly Regional comparison diverse in terms of data presentation. Standards Several questions in the project outline entail are ofen considered as undesirable straitjackets, comparison of the fnds and features under but they should only be seen as a way to share study with those from other sites. For research information in a structured fashion, so as to allow questions on regional occupation, this is self- for comparative analysis. The use of standards evident, but it may even be argued that such and conventions in reports should be considered comparisons should be made to determine site a basic precondition for research, and is no function. No mater the specifc aim of any diferent from the requirements for publication of inter-site comparative analysis, the availability research results in scientifc journals. This should of digital datasets is important. The Dutch not be restricted to the reporting on material regulations demand deposition of datasets in culture remains, but extend to all aspects. the national repository DANS-EASY, which Particular atention is required for reporting the implies that relevant data are, in theory, results of spatial analysis, which ofen lack explicit available to the community. information about the use of methods and Unfortunately regional analysis is almost seting of parameters, and hence leave the critical absent in the reports under study due to the reader with an unreadable black box. incompatibility of many datasets (the result of a lack of standards; see above). Moreover, comparative studies cannot be executed for the 6.4 Conclusions simple fact that datasets are ofen not available in the digital repository, in contravention of the regulations. This is clearly a great faw in Dutch This section provided several avenues for development-led archaeology. One way or improving the quality of development-led another, guarantees are needed that all data will archaeology in the Netherlands concerning early be digitally preserved for future reference. We prehistory. At a more abstract level, these propose two options. The frst is to include recommendations relate to three areas for verifcation of this obligatory data archiving in improvement. First, many recommendations the work of the Dutch heritage inspection need to be addressed in the earliest research agency (Erfgoedinspectie). The second option is phase, when the project outline is drafed. This is to create an explicit link between DANS-EASY the best guarantee that specifc improvements and Archis and force the uploading of datasets will be carried out, rather than being lost sight of – each of which should receive a unique ID due to budget constraints. This recommendation linked to the Archis registration code – as part of is related to augering depths, phased excavation the registration process; this is the only strategies, functional analysis, spatial analysis and guarantee that specialist datasets are digitally the use of expertise on early prehistory. Second, available when the Archis registration is changes need to be made so that the availability fnalised. of datasets through digital repositories is guaranteed. Third, the development of standards for the analysis and publication of the various categories of cultural material and other aspects of research needs to be stimulated. 219 7 The national research agenda and — future perspectives

In this study we have tried to demonstrate to However, the decision of what to investigate, what extent development-led archaeology, in and what not, is not straight forward. We feel the period 2005-2014, has made a contribution that the assessment of research potential in to answering research questions regarding the terms of scientifc gain could use a more solid early prehistory of the Netherlands. These basis. A research agenda forms a good starting questions derive from the NOaA 1.0, and were point, but that’s not where preparations should considered to be of importance in the context of stop. As for all research, in order to be able to Dutch archaeology when it was drafed more a answer a question, information or data are decade ago. It has recently been stated that required, and it must be clear by which means NOaA 1.0 resembles a regional research these can be collected and analysed. Hence, the framework much like those developed in assessment of the research potential of Britain.502 Indeed, there is a strong tendency to archaeological sites should not only involve the address issues at a regional scale. In the establishment of, for instance, the variability in Netherlands, we have witnessed the increase of features and fnds categories, and probable research agendas at a smaller-than-national dating, but also an explicit evaluation of which scale: provincial (n=3), regional (n=2) and information sources connect to research municipal (n=20). On the one hand, we can look questions, and what possibilities exist to obtain at this as a positive development, as it makes the data that is actually needed. more explicit what the archaeological potential Let us frst return to the research agendas. is to decision makers and the wider public. In broad terms, the younger we get in time, the On the other hand, we can look at this as a smaller the geographical scale on which negative development, because it bears the risk archaeological phenomena can be interpreted of knowledge fragmentation, since research within a socio-cultural (socio-economical, socio- agendas can potentially difer between two political) context. Local communities during neighbouring regions or municipalities. Since early medieval times operated in a smaller research agendas are strategic instruments, geographical space than did hunter-gatherers choices could be made from very diferent during the Mesolithic. And Mesolithic hunter- perspectives, and potentially lead to missed gatherers probably operated in a smaller opportunities to make a signifcant contribution geographical space than Late Palaeolithic to archaeological research questions because hunter-gatherers. And within the Late decisions are made by local authorities. This Palaeolithic, Hamburgian and Ahrensburgian confict of interests could represent a major hunter-gatherers may have operated in a wider threat to the archaeological heritage. geographical space than those of the Federmesser Gruppen. However, one thing is for certain: early prehistoric communities actively 7.1 Significance assessment and organised their life on a spatial scale that scientific knowledge gain exceeds that of modern-day administrative units in the Netherlands. From this perspective, research questions concerning the early So what does this mean when we focus on the prehistory of the Netherlands need to be defned early prehistory of the Netherlands? Let us frst on a national scale, and in various respects even and foremost state that choices have to be on an international scale. In contrast, in a made. We cannot salvage everything, simply due context of ‘community archaeology’, it can be to the fact that resources are not infnite. But we argued that “local research, whether guided by local never did investigate ‘everything’. Choices have agendas or not, provides detailed, but local always been made, and will always be made. information, for the focus is generally on the local. And we should be honest: not everything needs The sum of in-depth ‘scientifc particularism’ does not to be investigated. Judging from the reports automatically contribute signifcantly to eforts to 502 Groenewoudt et al. 2017, 183. selected for our study, a large number of projects address supralocal questions, or generate supralocal 503 Compare studies on the Late Neolithic have not contributed anything other than a dot theories.”504 The second part of this quote we can and the Bronze Age (Fokkens, Steffens & Van As 2016) and the occupation history on the map, labelled ‘early prehistory’ (this is not endorse; our study provides a good illustration of the western part of the province of specifc to early prehistory; it applies to all other of this situation. The frst part is more Noord-Brabant (Ball & Van Heeringen 2016). periods of Dutch prehistory and early history503). problematic where it comes to early prehistory. 504 Groenewoudt et al. 2017, 192. 220 —

Inhabitants of any town in the Netherlands context of community archaeology, but can also – take Zeewolde in the Flevoland polders, for serve educational purposes.505 Where instance – are easily surprised when scraps of a development-led projects are required to follow Mesolithic hunting camp are found in their strict protocols506 and have to work within tight backyard, so to speak. But from a scientifc constraints of time and money, there is litle to no perspective, this may not be exciting at all, and room for the involvement of non-professionals or certainly not when preservation is limited. This for the training of, for instance, students in creates a dilemma: in such cases, should one archaeology. Although regulations already make decide to excavate if destruction cannot be it possible to leave the investigation of deselected prevented, or leave the site to be destroyed? sites to amateur archaeologists, we have the Afer all, excavation involves money coming impression – which may, of course, be wrong – from developers and eventually taxpayers. that this does not ofen happen. The question is to what extent scientifc knowledge gain should weigh in decision making. Indeed, because academics in 7.2 The National Archaeological archaeology are paid by the state, one could Research Agenda and research argue that scientifc knowledge gain concerns questions issues of national signifcance. If the assessment of the information potential of archaeological sites is connected with a national research While we were conducting our study, the agenda, and this assessment does play a role in National Heritage Agency launched the National decision making, then scientifc knowledge gain Archaeological Research Agenda 2.0 (NOaA 2.0). is given weight. In that case, we want to argue It difers from the NOaA 1.0 in its ambition: the for expert-based, more rigorous choices with NOaA 2.0 intends to include, as Groenewoudt regard to the question of whether or not a site et al. put it, “only the most important national needs to be excavated. The argument that (supraregional) archaeological questions, focusing archaeological heritage will be lost when specifcally on development-led (Malta) excavation or, indeed, in situ preservation is not archaeology”.507 Groenewoudt et al. also point out possible does not sufce. Both excavation and in that the compilation of this new agenda did not situ preservation have consequences in terms of involve current academic research ambitions, in spatial planning (e.g. development of industrial order to enable a balanced set of research zones or urban districts) and resources. It has to questions that cover a broad range of themes, be made clear, insofar as this is possible, what and to “serve the management of the archaeological the expectations are about potential knowledge heritage as a source of knowledge of our collective past gain. This should be done through a research now and, above all, in the longer term”.508 At the strategy that outlines which data that are same time it is recognised that a research required for the answering of specifed research agenda intends to enhance scientifc quality, questions need to be collected, and by which which requires that it be used, and that quality methodological and analytical means this can be and costs are a concern of those who have to achieved. make choices. Somehow, all this leaves us with Of course there is always a factor of the impression that knowledge gain in a uncertainty. Archaeological reality is ofen development-led context difers from that in an diferent than expected, and it would be a academic context. We cannot see, however, that mistake to leave the unexpected aside because it one can be disconnected from the other. was not part of the signifcance assessment. Development-led work needs to be anchored in The unexpected always leads to knowledge gain, scientifc approaches in the same way as 505 We are aware of the fact that this brings and should always receive atention in any academic work does, and there is no reason to with it other problems, but that is not a context of development-led feldwork. Another assume that academic research ambitions – no reason to avoid the dialogue. 506 Namely, the archaeology quality issue involves the fate of sites that are expected mater whether these are vulnerable to change standard (PvE) and archaeology quality to have too litle potential for knowledge gain or not – do not contribute to knowledge of our standard (KNA; Willems & Brandt 2004). 507 Groenewoudt et al. 2017, 185. from a scientifc perspective, and should collective past. Hence, it is not an easy task to 508 Groenewoudt et al. 2017, 185. therefore be deselected for excavation. Such please all parties in archaeology, and further 509 This is also expressed by Groenewoudt et al. 2017, 192. sites can possibly play an important role in the dialogue will be necessary.509 221 —

The defnition of the NOaA 2.0 research feldwork). Technological aspects of lithic tool questions involved a number of basic restrictive production and potery production are given too principles (A-F below), the most important of litle atention, but could be analysed by means which, in the context of our study, we list of targeted morpho-technological studies and below,510 and briefy comment upon within the scientifc methods (clay source analysis). Insight context of our study and in reference to the into the social use of space is still restricted, but NOaA 1.0. could possibly be enhanced through the application of more rigorous spatial analytical A) Questions concern essential knowledge gaps techniques. The identifcation of Mesolithic and promising directions for research from an dwelling structures needs particular atention (inter)national perspective. during feldwork, through continuous mapping Our study demonstrates that the knowledge of fnds densities and awareness about the gain with regard to NOaA 1.0 questions is limited potential presence of (even quite heavy) post and that many gaps remain. Typo-chronologies features underneath fnds scaters. Structural of lithics have received considerable atention, features also need to be analysed from a more but there are various problems to tackle: it is integral perspective, which includes distributions crucial to pay atention to ‘clean’ (single-phase) of material remains. The extent of human assemblages associated with independently infuence on the nature/character of prehistoric datable material (botanical macroremains from cultural landscapes in hunter-gatherer and early surface hearths). Promising directions for farmer contexts has received limited atention, research involve the study of Mesolithic pit but could be addressed by means of multiproxy hearth function by means of scientifc methods approaches (despite analytical and interpretive (SEM, micromorphology and physico-chemical difculties). analysis), and based on targeted sampling strategies in the feld. Reliable sourcing of lithics C) Questions reflect scientific ambitions, (use of good reference collections, thin-section as well as aspects of field practice, notably the analysis, physico-chemical analysis) from all rare focus on early prehistory. socio-cultural contexts in the Palaeolithic, Some NOaA 1.0 questions of higher scientifc Mesolithic and Neolithic is important to enable ambition refer to the archaeological signifcance research into geographical scales of landscape of ‘empty’ space and the understanding of exploitation and socio-cultural palimpsests. Both aspects – which are, in fact, interconnectedness. strongly related – have only sporadically been addressed in development-led work, despite B) Only archaeological (cultural heritage) their scientifc pertinence. The focus on high- questions are included, with the exception of density concentrations of materials (clusters) historical cultural landscapes (biotic and restricts any possibility to take our knowledge a abiotic) for which anthropogenic activity is step further, whilst the high-density considered to have significantly impacted upon concentrations themselves are poorly the landscape. understood in terms of formation processes. Many questions in the NOaA 1.0 concerned Strategies to investigate ‘empty’ zones (i.e. those material culture, notably, the production, use, with meagre fnds) in efective ways need to be discard and socio-cultural meaning of tools and developed. Despite methodological complexities, other objects (e.g. ornaments/jewellery), but enhanced spatial analytical approaches are also dwelling structures, the use of pit hearths, needed to develop further understanding of the and the exploitation and sourcing of various raw spatial structure that now remains ‘hidden’ in materials. Our study shows that limited progress fnds distributions. Archaeologists need to do has been made, but that there is much potential. this not only to increase our understanding of the Insight into the functional aspect of lithic tools is socio-cultural meaning of space, but also to be improving, but altogether there is restricted able to develop feld practices that can be knowledge about tool type–related functional adapted to the variable conditions (constraints of variability and assemblage variability because of time and money, logistical and technical insufcient sample size. Larger samples should constraints) under which development-led work be analysed in an early stage of research (during is conducted. 510 From Groenewoudt et al. 2017, 187-188. 222 —

D) Questions are as independent as possible of approaches, which are proven to result in theoretical views or principles. knowledge gain. Although many research questions in the NOaA 1.0 appear to be theoretically ‘neutral’, F) ‘Why’ questions (cause-and-effect; processes theoretical views or principles cannot be of change) are excluded, as these require a level eliminated from archaeological research. of synthesising analysis that reaches beyond Apparently straight forward questions about the purpose of site-focussed development-led tool type and tool function can instantly be research. placed in the perspective of a processual More complex question about the origin and theoretical debate about style and function. spread of cultural phenomena and the A question about the socio-cultural (indigenous, long-term dynamics of socio-cultural behaviour emic) ‘meaning’ of objects can be cast in the and environment featured in the NOaA 1.0. context of processual/post-processual/ Such questions, not surprisingly, appeared to be phenomenological debate. As yet, there is still beyond the possibilities of development-led much to learn with regard to many of these research, and are not addressed in the reports. aspects, notably, burial practices and the socio- Questions in the ‘why’ category are in the cultural signifcance of objects. In order for domain of academic research. However, the archaeologists to have the opportunity to building blocks – in other words, data – do have identify aspects of past human behaviour to come from development-led work, at least beyond, for instance, the strict function of tool for a big part, since universities only have types, there is pertinent need for rigorous feld restricted means and resources to collect data documentation of fnds contexts and awareness themselves. It is therefore important that there amongst feld crew about the potential presence is a stronger involvement of academia in of archaeological phenomena that may tell us development-led research. From our study it has more about how people were connected with become clear that those development-led materials and the environment. projects that were more closely connected with the academic environment generated most of E) ‘Specialist’ questions (e.g. bioarchaeology, the knowledge gain. physical geography, absolute dating) are Despite the principles followed for the excluded, as these are considered to be related development of the NOaA 2.0, much remains to to questions on cultural heritage (specialist be done. The research questions are connected input defines fieldwork quality). with ‘research felds’,511 which are basically The NOaA 1.0 addresses many ‘specialist’ themes or focal points. Consultation of the questions, which focus on particular aspects of online NOaA 2.0, which functions as a search research themes. We have seen that, in many engine,512 shows that the research questions respects, the gain from development-led work is concerning the archaeology of the Stone Age of an incidental nature. The most progress has number 39, of which only 32 are relevant for been made in the feld of archaeobotanical early prehistory (table 7.1). The focus of these research on charred parenchymatous remains questions is highly unbalanced in terms of (plant foods), and the functional interpretation themes, and is inconsistent with regard to the of pit hearths based on SEM charcoal analysis principles of question defnition. The list in and physico-chemical analysis of organic table 7.1 includes many questions that do not substances. More substantial work along these meet the restrictive principles, the most lines is needed to test the hypotheses that were important one being the exclusion of ‘why’ generated from this pioneering research. questions that involve cause-and-efect, long- Specialist input for the development of a project term perspectives, and analysis on a regional, 511 Groenewoudt et al. 2017, 188. 512 www.noaa.nl outline is a conditio sine qua non to ascertain such or even national/international, scale. 223 —

Table 7.1 NOaA 2.0 research ﬁelds and questions (translated from Dutch) queried for ‘Stone Age’. Several questions appeared irrelevant for early prehistory (e.g. the evolution of ship construction; the colonisation and layout of unembanked tidal ﬂats and peat marshes), and have been omited.

Areas where litle is known of the archaeology When and in which way were areas that are now submerged used by humans, and how does this compare to our knowledge about the terrestrial record? The dynamics of the Netherlands’ cultural landscape Did humans before to the Neolithic have infuence on the landscape and/or vegetation? If so: where, in what way, for which purpose(s) and on what scale? Occupation and adaptation in the rivers area and along the coast Which connections exist between physical-geographical change, land use and occupation paterns along the coast? Which connections exist between physical-geographical change, land use and occupation paterns in the rivers area, including the IJssel? The archaeology of ritual How are ‘sacrifcial landscapes’ situated and structured, and what is the time-depth? What are the nature, context and meaning of intentional (ritual?) depositions in and around house and yard? Funerary practice and grave monuments What are the characteristics and context of mortuary practices in Early Prehistory? What are the nature and context of variation and change in burial practices? What are the archaeological and landscape context of isolated graves or difuse groups of graves? How are graves and cemeteries situated relative to rural setlements, and how does this relationship change? To what extent and for what purpose were pre- and early historic cemeteries and grave monuments reused? What are the context and meaning of stray human skeletal elements in and outside setlements? The earliest occupation in the Netherlands What are the origin, mobility and territory size of Neanderthal groups, and connections with other areas outside the Netherlands? Where, in which phases of the Pleistocene and under which climate conditions did early hominids stay in the Netherlands? Late Palaeolithic – Early Mesolithic transition In which way were (which species) of plant and animal resources used for subsistence, and raw materials for objects and items of exchange and trade? When, where and under which climate conditions appear Late Palaeolithic cultures and cultural traditions? What is their geographical and chronological extent, and how are they interrelated? Neolithisation process How dis subsistence change during the Late Mesolithic through middle Bronse Age? How do buildings and setlements show before the middle Bronze Age B (1500 BC)? Which landscape zones were used for setlement, hunting, crop cultivation and animal husbandry in the Late Mesolithic and Early Neolithic? Which changes and diversifcation occurred in the late phase of the LBK? To what extent does (dis)continuity in the Neolithic occupation occur afer the LBK? Consolidation of farming Which changes take place in methods, intensity and location of food storage? How were houses internally structured and functionally partitioned? The role of natural food resources since the introduction of farming How has fshing evolved in terms of technology and economy? Which role did the exploitation of natural resources (including hunting and fshery) have since the introduction of farming? Dynamics of land use What indications exist for seasonality and specialisation of setlements? What economic functions had periferal areas (‘marginal’ landscape zones), and how does this this show archaeologically? Which infuence had farming (crop cultivation and animal husbandry) on vegetation and fauna? When, where and to what extent did aeolian and slope erosion occur, and to what extent is this (directly or indirectly) related to (which?) human activity? To what extent does (inter)regional and diachronic variability exist in frequency and the distance over which setlements, felds and pastureland were relocated? Human-material culture relationships What is the nature and meaning of utilitarian objects of organic material in material culture? How was waste dealt with? 224 —

Although the NOaA 2.0 is intended to highlight the most important themes and to 7.3 What’s next? foster operationalisation in the feld by means of practical suggestions, many questions implicitly have a high degree of complexity. Take, for Of course it is possible to draf a new set of instance, the following question: “Did humans research questions based on our analysis of before to the Neolithic have infuence on the reports resulting from development-led work. landscape and/or vegetation? If so: where, in Afer all, we have shown in the preceding what way, for which purpose(s) and on what chapters what development-led research has scale?” (translation ours). To determine whether contributed to the answering of research or not humans had infuence on the landscape/ question from the NOaA 1.0. Since this vegetation, rigorous but methodologically contribution is rather limited, we could satisfy complicated analysis of multiproxy data is ourselves by stating that all questions remain required. How human activity infuenced the relevant. In a way they do, and some more landscape/vegetation involved the question of specifc questions could be added – but this agency. And to establish for what purpose would be in confict with one of the selective humans infuenced the landscape/vegetation, principles given above. But it is equally clear that one has to rely on theory-based interpretive several NOaA 1.0 question surpass the scope of frameworks. The fnal aspect, scale, can only the development-led context of research, satisfactorily be investigated by means of notably, questions in the ‘why’ category. Indeed, sampling well beyond the site level to which we feel that such questions should be omited, development-led work is usually bound. at least as questions which should be the direct Furthermore, empirical aspects are concern of development-led research. interwoven with interpretation, which We are, however, of the opinion that this is complicates their day-to-day use in the practice not the right time or place to propose a new set of development-led work. To take another of questions for the NOaA 2.0. The issues example: “What are the nature, context and addressed above lead to a more fundamental meaning of intentional (ritual?) depositions in question: What should be the role of and around house and yard?” (translation ours). development-led and academic research in the The aspects ‘nature’ (what is the phenomenon?) broader context of archaeological feldwork and ‘context’ (association, location, sedimentary (which should be aimed at the collection of data embedding?) involve empirical evidence, that is that can stand at the basis of question-driven to say, archaeological feld observation. The knowledge gain), as well as analysis and aspect ‘meaning of (intentional?) depositions’ interpretation? The answer to this question involves interpretation, since it is impossible to should set the baseline for the improvement of excavate ‘meaning’. Whether or not depositions the NOaA 2.0, which certainly should and can are intentional is also a mater of interpretation. serve a purpose in quality control. We are And what is meant by the term depositions? convinced that a more balanced and targeted set In Dutch archaeology, the term implicitly refers of research questions can come from the to single objects or groups of objects that have dialogue between the development-led been deliberately placed at a location. Such community and the academic community, and placement is intentional by defnition. Of course for which the present study may well form the one cannot exclude that an object or group of starting point. To this end, two example research objects may have been accidentally lost, but this topics are discussed. is not a deliberate action. The characterisation of a context as deposition or loss is a mater of interpretation. 225 —

Mesolithic pit hearths • Do newly discovered pit hearths ft the existing Pit hearths are a characteristic feature of the typochronology? In order to get to an answer it Mesolithic of the Netherlands north of the river is necessary to clearly describe morphological Rhine. As these features are known in large aspects of pits and to obtain 14C dates for all numbers, the NOaA 1.0 addresses their function morphological types and identifed types of (albeit with some caution): “are there surface wood. This involves the collection of basic hearths, pit hearths and other fre places and information for all pit hearths (or a substantial what are their characteristics?” Section 5.2.2 and representative sample of the pit hearths made clear that signifcant knowledge gain has when found in large numbers). been realised with regard to the function of pit • What is the correlation between pit hearths. Perhaps this gain has led to the absence morphology and wood species, and to what of any research questions on pit hearths in extent might the observed diferences result NOaA 2.0, although this type of research may from technological changes in function (notably have been perceived as being too specialist (see tar production)? As it requires a quantitatively restrictive principle E in section 7.2) and not of sound set of data, this question can be directly national relevance (see restrictive principle A in addressed on sites with large numbers of pit section 7.2). hearths; data from sites with few pit hearths The knowledge gain derives from detailed add to synthetic research at the regional scale. work conducted at some of the larger • What is the (variation in) function of pit development-led projects in the period under hearths? SEM and physico-chemical analysis study. The analysis has gained a lot from the fact point to tar production, but it is equally clear that a small number of dedicated researchers that further analysis is needed to test this were involved. The following is a summary of hypothesis, while the evidence does not the knowledge gain as described in Section 5.2.2. necessarily ft all pit hearths. The continuation There is now a morphological typology of this type of research, in combination with comprising three types (A, B and C) that allows soil micromorphology, is needed to gain more for inter-site comparison of available data. insight into the function of pit hearths. The large number of 14C dates suggests that these types may have a chronological meaning The end of the Linear Bandkeramik as well. The combination of 14C dates and One of the research questions in the NOaA 1.0 is charcoal analysis shows that wood from pine related to the seemingly abrupt end of the and oak was burned. Pine was dominant from Neolithic occupation in the Dutch loess area at the Preboreal until the Mid-Atlantic, while oak the end of the LBK: ”is there a continuity or occurs throughout the Boreal and into the Late discontinuity in the occupation afer the LBK? Atlantic. SEM and physico-chemical analysis If there is a discontinuity, what is the duration of gave more insight into the function of the pits. the hiatus in occupation?” In the NOaA 2.0, this There are strong indications for tar production, question is maintained although rephrased as but of course other – yet unknown – functions “to what extent is there (dis)continuity in need to be considered as well. Neolithic occupation afer the LBK?” Apparently, While much progress has been made, the there is a lack of knowledge gain with regard to interpretation of these features as anthropogenic this aspect on the basis of development-led has been challenged recently.513 It is proposed that archaeology in the Netherlands, while at the the pit hearths should be interpreted as collapsed same time it is considered of national relevance burnt ant nests. This dichotomy in interpretations (see restrictive principle A in section 7.2). makes perfectly clear that more research is To address this question, a small excursion needed to understand this phenomenon. How needs to be made to discuss the concept of can development-led archaeology contribute to discontinuity. The implicit use of the term above its study? Importantly, the issue needs to be encompasses two aspects of discontinuity. First addressed as a regional/national research of all it embodies discontinuity in spatial terms. ambition with a set research methodology to In general, continuity in occupation may be ensure a consistent approach to the answering of based on discontinuity of setlements: with relevant questions. The following subquestions shifing occupation within a small region we are proposed: might still uphold occupation continuity. A clear 513 Crombé et al. 2015. 226 —

example of this perspective may be found in the sources). To answer this subquestion absolute occupation history of the river dunes in the dating needs to be seen as essential: dating on Rhine-Meuse basin: while individual sites were the basis of cultural material would lead to a occupied in an on-and-of patern, a continuous circular argument. occupation can be seen for much of the ffh • What are the technological, typological and millennium cal. BC on regional scale. With the morphological characteristics of the restricted spatial extent of any archaeological absolutely-dated material culture? In our project as a given, the spatial dimension of opinion it is – at this stage of research – of discontinuity can easily be taken into account restricted value to invest time (money) in the when addressing this research question. description and analysis of cultural material The second aspect of discontinuity is more that is not dated. It will not help us answer the difcult to deal with: the temporal dimension. research question. As argued above it is of In this specifc case it deals with the absolute essence that in the analysis and publication of date of the last stage of the LBK in the Dutch the material culture a standard system is used loess area in comparison to the oldest absolute to allow intersite comparison. date of the post-LBK occupation in the area. • What subsistence evidence is available? While the Dutch LBK is characterised by a The analysis of archaeobotanical and detailed ceramic typochronology,514 the limited archaeozoological material should again be understanding of post-LBK occupation (hence based on the available absolutely dated the research question) frustrates dating on the subassemblages. Evidence of agricultural basis of established (potery) chronologies. felds is very limited for the Neolithic due to The need for absolute dates is therefore evident: the difculty to recognise them in area with a series of absolute dates on cultural material no or litle sedimentation. The best option to would allow for a much needed cultural fnd agricultural felds is to include soil chronology. While the importance of absolute micromorphological research as a standard dating cannot be underestimated to address this research tool when there is a chance that the research question, we should take into account Neolithic surface is preserved. that all dating methods have a limited • What evidence is available for human resolution: the interpretation of a series of 14C impact beyond the setlement? While dates into occupation continuity requires development-led archaeology is ofen auxiliary arguments to be found in the focussed on setlements because this site type archaeological material itself. is more easily found during surveys, an This being said, what can development-led important subquestion pertains to human archaeology contribute to the study of this infuence beyond the setlement proper. research question? As for the preceding topic, Already during the desktop phase of the issue at hand should be considered a development-led projects, locations with the regional/national research ambition to be potential for obtaining environmental data approached with a coherent set of (e.g. on vegetation, sedimentation/erosion) methodologies. The following subquestions are should receive extra atention: while these proposed: locations would probably not yield large • What is the absolute date of (sub-)assemblages number of boxes with material culture they for which temporal integrity can reasonably be are the anchor points to understand the assumed (e.g. features)? As stated in NOaA 2.0 development of the environment in the this subquestion should also be taken into period under study.515 These locations are 514 Van de Velde 2014. account for presumed Late Mesolithic sites in defned by sedimentation and therefore also 515 Cf. Peeters 2007, 271. For examples see the region because of the lack of knowledge on hold great promise for the study of Bouman et al. 2013 and Kooistra & Peeters in prep. post-LBK fint (typology, technology and fint agricultural felds. 227 —

archaeological research needs to be beter 7.4 Conclusion aligned with the potential and practice of what can reasonably be achieved within such a context. Importantly, our study also leaves us with the The aim of this study was to evaluate the gains in impression that there is a lot of potential to be knowledge that are is achieved through tapped from all the work that has already been development-led investigations with regard to done. To tap this potential, all data need to NOaA 1.0 research questions concerning the early become available, but even once that has been prehistory of the Netherlands. Our synthetic achieved, in many cases researchers will probably analysis has made clear that basically all NOaA 1.0 need to return to the fnds material itself. So research themes and questions remain either paradise is not lost, at least not entirely. Yet, it is unanswered or partially answered. Several necessary to develop further practices which lead questions, particularly those in the ‘why’ category to useful and accessible research results from or which require synthetic work at the regional development-led projects, in order to ensure that scale, are well beyond the possibilities of paradise doesn’t crumble much further. This is the development-led projects. Any future research responsibility of the archaeological community as agenda that is intended to inspire development- a whole. led work and is concerned with the quality of the

229 Bibliography —

Aalbersberg, G., J.B. Hielkema Arts, N., & M. Hoogland 1987: Baumgartner, A., M. Sampol- Beuker, J. 2010: Vuurstenen & D. Mol 2013: Mammoetresten A Mesolithic setlement area Lopez, E. Cemeli, T.E. Schmidt, werktuigen: technologie op het bij het Oranjekanaal: with a human cremation grave A.A. Evans, R.E. Donahue & scherp van de snede, Leiden. aardgastransportleidingtracé at Oirschot V, municipality of D. Anderson 2012: Scheemda-Ommen (A-661), Best, the Netherlands, Genotoxicity assessment of Beuker, J.R., & E. Drenth 2014: KR-077, catalogusnummer 35, Helinium 27, 172-189. birch-bark tar – a most Prehistoric artefacts of red gemeente Midden-Drenthe, versatile prehistoric adhesive, Heligoland fint from Lower archeologisch onderzoek: Aveling, E.M., & C. Heron Advances in Anthropology 2(2), Saxony (Germany) and the begeleiding, Weesp (RAAP- 1998: Identifcation of birch 49-56. Netherlands – an outline, rapport 2320). bark tar at the Mesolithic site Siedlungs- und Küstenforschung im of Star Carr, Ancient Biomolecules Bazelmans, J.G.A., südlichen Nordseegebiet 37, Amkreutz, L.W.S.W. 2013: 2(1), 69-80. O. Brinkkemper, 113-124. Persistent traditions: a long-term J.H.C. Deeben, J. van perspective on communities in the Bailey, G. 2007: Time Doesburg, R.C.G.M. Lauwerier Binford, L.R. 1978: Nunamiut process of neolithisation in the perspectives, palimpsests and & P.A.M. Zoetbrood 2005: ethnoarchaeology, New York. lower Rhine area (5500-2500 cal the archaeology of time, Journal Mag het ietsje meer zijn? Een BC), Leiden (PhD dissertation of Anthropological Archaeology 26, onderzoek naar de door bedrijven Binford, S.R., & L.R. Binford Leiden University). 198-223. opgestelde Programma’s van Eisen (eds.) 1968: New perspectives in voor archeologisch onderzoek uit de archaeology, Chicago. Amkreutz, L., & R.H.A. Corbey Bakker, R. 2003: The emergence periode 2003-2004, Amersfoort 2008: An eagle-eyed of agriculture on the Drenthe (Rapportage Archeologische Bloo, S.B.C. 2010: Aardewerk, in: perspective: Haliaeetus albicilla Plateau: a palaeobotanical study Monumentenzorg 120). D.E.A. Schiltmans (ed.), in the Mesolithic and Neolithic supported by high-resolution Roterdam Groenenhagen- of the lower Rhine area, in: 14C-dating, Bonn Beckerman, S.M. 2012: Tuinenhoven De Zwanen-Rietpark: H. Fokkens, B.J. Coles, (Archäologische Berichte 16). Dutch beaker chronology een opgraving van een vindplaats uit A.L. van Gijn, J.P. Kleijne, re-examined, Palaeohistoria het neolithicum op een rivierduin op H.H. Ponjee & C.G. Slappendel Ball, E.A.G., & R.M. van 53/54, 25-64. IJsselmonde (vindplaats 13-78), (eds.), Between foraging and Heeringen (eds.) 2016: Roterdam (BOORrapporten farming: an extended broad Westelijk Noord-Brabant in het Beckerman, S.M. 2015: 491), 43-54. spectrum of papers presented to Malta-tijdperk: synthetiserend Corded Ware coastal communities: Leendert Louwe Kooijmans, Leiden onderzoek naar de bewonings- using ceramic analysis to Bokelmann, K. 1971: (Analecta Praehistorica geschiedenis van het westelijk deel reconstruct third millennium BC Duvensee, ein Wohnplatz des Leidensia 40), 167-180. van het Brabants zandgebied, societies in the Netherlands, Mesolithikums in Schleswig- Amersfoort (Nederlandse Groningen (PhD dissertation Holstein und die Arnoldussen, S. 2008: A living Archeologische Rapporten 51). University of Groningen). Duvenseegruppe, Ofa 28, 5-26. landscape: Bronze Age setlement sites in the Dutch river area Barton, R., R. Jacobi, Beckerman, S.M., & Bokelmann, K. 1991: (c. 2000-800 BC), Leiden (PhD D. Stapert & M.J. Street 2003: D.C.M. Raemaekers 2009: Duvensee Wohnplatz 9. dissertation Leiden University). The Late glacial reoccupation Vormvariatie van Vlaardingen- Ein präborealzeitlicher of the British Isles and the aardewerk: een nieuwe Lagerplatz in Schleswig- Arnoldussen, S., J.P. Mendelts, Creswellian, Journal of typochronologie van het Holstein, Ofa 48, 75-114. R.L. Fens & J.H.M. Peeters Quaternary Science 18, 631-643. aardewerk van de 2012: Een mesolithisch kampement Vlaardingengroep Bokelmann, K. 1994: te Meerstad – vindplaats 2a, Bats, M., P. Crombé, (ca. 3400-2500 v. C.), Frühboreale Mikrolithen mit Groningen (Grondsporen 12). I. Devriendt, R. Langohr, Archeologie 13, 63-82. Schäfungspeck aus dem J.H. Mikkelsen, C. Ryssaert & Heidmoor im Kreis Segeberg, Arts, N. 1988: Archaeology, A. van de Water 2010: Een Behre, K.-E. 1981: Ofa 51, 37-47. environment and the social vroegmesolithische vindplaats te The interpretation of evolution of later band Haelen-Broekweg (gem. Leudal, anthropogenic indicators in societies in a lowland area, in: provincie Limburg), Amersfoort pollen diagrams, Pollen et C. Bonsall (ed.), The Mesolithic in (Rapportage Archeologische Spores 23, 225-245. Europe, Edinburgh, 291-312. Monumentenzorg 190). 230 —

Bollingino, R., O. Nehlich, Brindley, A.L. 1986: The Brounen, F.T.S., & Clevis, H., & A.D. Verlinde M.P. Richards, J. Orschiedt, typochronology of TRB West J.H.M. Peeters 2009: (eds.) 1991: Bronstijdboeren in M.G. Thomas, C. Sell, Group potery, Palaeohistoria 28, Halfabrikaten van vroeg- Itersumerbroek: opgraving van een Z. Fajkošová, A. Powell & 93-132. neolithische dissels: enkele bronstijdnederzeting in Zwolle- J. Burger 2013: 2000 years of ‘nieuwe’ vondsten uit het Itersumerbroek, Kampen. parallel societies in Stone Age Brinkkemper, O. 2007: Limburgse dekzandgebied, de central Europe, Science 342, Emplacement Centraal Station lösszone en het Maasdal, Cnudde, V., J. Dewanckele, 479-481. Roterdam: archeobotanisch Archeologie 13, 5-16. T. de Kock, M. Boone, onderzoek van een vindplaats J.M. Baele, P. Crombé & Bordes, F. 1968: The Old Stone uit het laat-mesolithicum Brouwer Burg, M., H. Peeters E. Robinson 2013: Preliminary Age, New York. gelegen op en bij een rivierduin, & W.A. Lovis (eds.) 2016: structural and chemical study of in: A.J. Guiran & O. Brinkkemper Uncertainty and sensitivity analysis two quartzite varieties from the Bordes, F. & D. de Sonneville- (eds.), Roterdam-Randstadrail: in archaeological computational same geological formation: a Bordes 1970: The signifcance archeologisch onderzoek, 1: modelling, Cham. frst step in the sourcing of of variability in Palaeolithic emplacement Centraal Station: quartzites utilized during the assemblages, World Archaeology onderzoek van vindplaats 05-42 uit Bulten, E.E.B, Mesolithic in northwest Europe, 2(1), 61-73. het laat-mesolithicum gelegen op en F.J.G. van der Heijden & Geological Belgica 16(1/2), 27-34. bij een rivierduin door middel van T. Hamburg 2002: Emmeloord, Bos, J.A.A. & C.R. Janssen 1996: mechanische boringen, Roterdam prehistorische visweren en fuiken, Collins, M.B. 1975: Lithic Local impact of Palaeolithic (BOORrapporten 318), 24-60. Amersfoort (ADC report 140). technology as a means of man on the environment during processual inference, in: the end of the last glacial in the Brinkkemper, O., Busschers, F.S. 2008: E. Swanson (ed.), Lithic Netherlands, Journal of W.-J. Hogestijn, H. Peeters, Unravelling the Rhine: response of a technology: making and using Archaeological Science 23, 731-739. D. Visser & C. Whiton 1999: fuvial system to climate change, stone tools, the Hague, 15-34. The Early Neolithic site at the sea-level oscillation and glaciation, Bos, J.A.A. & R. Urz 2003: Hoge Vaart, Almere, the Amsterdam (PhD dissertation Conneller, C. 2011: Late Glacial and Early Holocene Netherlands, with particular Free University of Amsterdam). An archaeology of materials: environment in the middle reference to non-difusion of substantial transformations in early Lahn River valley (Hessen, crop plants, and the signifcance Cappers, R.T.J., & prehistoric Europe, London. central-west Germany) and the of site function and sample D.C.M. Raemaekers 2008: local impact of Early Mesolithic location, Vegetation History and Cereal cultivation at Coudart, A. 1998: Architecture et man – pollen and macrofossil Archaeobotany 8, 79-86. Swiferbant? Neolithic wetland société néolithique – l’unité et la evidence, Vegetation History and farming on the north European variance de la maison danubienne, Archaeobotany 12, 19-36. Brinkkemper, O., L.I. Kooistra, Plain, Current Anthropology 49, Paris. H. van Haaster, L. van Beurden 385-402. Bos, J.A.A., B. van Geel, & F. Bunnik 2005: 9: Cremer, H. 2008: B.J. Groenewoudt & Archeobotanie, Amersfoort Cavallo, C., K. Esser, Diatomeeënonderzoek van R.C.G.M. Lauwerier 2005: (Nationale Onderzoeksagenda R. Lauwerier, W. Prummel, aardewerkscherven, in: H. Koot, Early Holocene environmental Archeologie 1.0, htps:// L. Smits & J.T. Zeiler 2006: 10: L. Bruning & R.A. Houkes (eds.), change, the presence and archeologieinnederland.nl/ Archeozoölogie en fysische Ypenburg-locatie 4: een nederzeting disappearance of Early sites/default/fles/9%20 antropologie, Amersfoort met grafveld uit het midden- Mesolithic habitation near Botanie.pdf). (Nationale Onderzoeksagenda neolithicum in het West-Nederlandse Zutphen (the Netherlands), Archeologie 1.0, htps:// kustgebied, Leiden, 203-212. Vegetation History and Brorsson, T. 2011: Slijpplaat- en archeologieinnederland.nl/ Archaeobotany 15, 27-43. ICP-analyse van aardewerk uit sites/default/fles/10%20 Crombé, P. 1999: Vers une TRB-context te Hatemerbroek, Zoologie%20en%20fys_ nouvelle chronologie absolue Bouman, M.T.I.J., J.A.A. Bos & in: E. Lohof, T. Hamburg & antrop.pdf). pour le mésolithique en R. van Beek 2013: Van wildernis J. Flamman (eds.), Steentijd Belgique, in: A. Thevenin (ed.), naar cultuurlandschap: een opgespoord: archeologisch Chilès, J.P., & P. Delfner 2012: L’Europe des derniers chasseurs: reconstructie van de regionale onderzoek in het tracé van de Geostatistics: modeling spatial epipaléolithique et mésolithique, vegetatie ontwikkeling van Twente Hanzelijn-Oude Land, Amersfoort/ uncertainty (2nd edition), Paris (Actes du 5e Colloque in het holoceen, Amersfoort (ADC Leiden (ADC report 2576/Archol Hoboken. Internationale UISPP, report 3413). report 138), 609-614. Commission 12), 189-199. 231 —

Crombé, P. 2015: Forest fre Crombé, P., E. Robinson, Deeben, J., D.P. Hallewas & (eds.), Vuursteen verzameld: over dynamics during the Early and M. Van Strydonck & T.J. Maarleveld 2002: Predictive het zoeken en onderzoeken van Middle Holocene along the M. Boudin 2013: Radiocarbon modelling in archaeological steentijd vondsten en -vindplaatsen, southern North Sea basin as dating of Mesolithic open-air heritage management of the Amersfoort (Nederlandse shown by charcoal evidence sites in the coversand area of Netherlands: the indicative map Archeologische Rapporten 50), from burnt ant nests, the north-west European Plain: of archaeological values (2nd 112-121. Vegetation History and problems and prospects, generation), Berichten van de Archaeobotany 25(4), 311-321. Archaeometry 3, 545-562. Rijksdienst voor het Oudheidkundig Deeben, J., & M.J.L.Th. Niekus Bodemonderzoek 45, 9-56. 2016c: Mesolithicum, in: Crombé, P., J. Deeben & Crombé, P., J. Sergant, L. Amkreutz, F. Brounen, M. van Strydonck 2014: E. Robinson & J. De Reu 2011: Deeben, J., D.P. Hallewas, J. Deeben, R. Machiels, Hunting in a changing Hunter-gatherer responses to P.C. Vos & W. van Zijverden M.-F. van Oorsouw & B. Smit environment: the transition environmental change during 2005: 8: Paleogeografe en (eds.), Vuursteen verzameld: over from the Younger Dryas to the the Pleistocene-Holocene landschapsgenese, Amersfoort het zoeken en onderzoeken van (Pre-)Boreal in Belgium and the transition in the southern (Nationale Onderzoeksagenda steentijd vondsten en -vindplaatsen, southern Netherlands, in: North Sea basin: Final Archeologie 1.0, htps:// Amersfoort (Nederlandse J. Jaubert, N. Fourment & Palaeolithic-Final Mesolithic archeologieinnederland.nl/ Archeologische Rapporten 50), P. Depaepe (eds.), Transitions, land use in northwest Belgium, sites/default/fles/8%20 123-135. ruptures et continuité en préhistoire Journal of Anthropological DEF%20Deeben%20 / Transitions, rupture and continuity Archaeology 30, 454-471. Paleogeografe.pdf). Deeben, J., H. Peeters, in prehistory, 2: paléolithique et D. Raemaekers, E. Rensink & mésolithique, Paris (Actes XXVIIe Crombé, P., M. Van Strydonck Deeben, J., H. Hiddink, L. Verhart 2006: Vroege congrès préhistorique de & M. Boudin 2009: Towards a H. Huisman, A. Muller, prehistorie, Amersfoort France, Bordeaux-Les Eyzies, refnement of the absolute J. Schokker & J. Wallinga 2010: (Nationale Onderzoeksagenda 31 mai-5 juin 2010), 583-604. (typo)chronology for the Early Middle Palaeolithic artefact Archeologie 1.0, htps:// Mesolithic in the coversand migration due to periglacial archeologieinnederland.nl/ Crombé, P., H. Groenendijk & area of northern Belgium and processes: a geological sites/default/fles/11%20 M. Van Strydonck 1999: the southern Netherlands, in: investigation into near-surface Vroege%20Prehistorie%20 Dating the Mesolithic of the P. Crombé, M. Van Strydonck, occurence of Palaeolithic +%20links.pdf). Low Countries: some J. Sergant, M. Boudin & M. Bats artefacts (Limburg-eastern methodological considerations, (eds.), Chronology and evolution Brabant coversand region, the De Grooth, M.E.T. 1987: in: J. Oberlin, C. Daugas, within the Mesolithic of north-west Netherlands), Netherlands The organisation of fint tool J.P. Salles & J.F. Evin (eds.), 14C et Europe: proceedings of an Journal of Geosciences 89, 35-50. manufacture in the Dutch Archéologie / 14C and archaeology, international meeting, Bandkeramik, Analecta Paris (Actes du 3e Congres Cambridge, 95-112. Deeben, J., & M.J.L.Th. Niekus Praehistorica Leidensia 20, 27-52. International, Lyon, 6-10 Avril 2016a: De Federmesser-traditie, 1998) (Mémoires de la Société Cuijpers, A.G.F.M. 2006: in: L. Amkreutz, F. Brounen, Demiddele, H. 2010: Préhistorique Française 26/ Histological identifcation of J. Deeben, R. Machiels, Diatomeeënonderzoek aan supplement 1999 to Revue bone fragments in M.-F. van Oorsouw & B. Smit 7 aardewerkscherven en een d’Archéométrie), 57-63. archaeology: telling humans (eds.), Vuursteen verzameld: over het bodemmonster, in: apart from horses and catle, zoeken en onderzoeken van steentijd- S. Diependaele & E. Drenth Crombé, P., R. Langohr & International Journal of vondsten en -vindplaatsen, (eds.), Archeologisch onderzoek G. Louwagie 2015: Mesolithic Osteoarchaeology 16, 465-480. Amersfoort (Nederlandse langs de rijksweg N11 hearth-pits: fact or fantasy? Archeologische Rapporten 50), (Spookverlaat) ten behoeve van de A reassessment based on the Deeben, J., E. Drenth, 103-111. aanleg van het windturbinepark evidence from the sites of M.-F. van Oorsouw & Rijnwoude te Hazerswoude- Doel and Verrebroek (Belgium), L. Verhart (eds.) 2005: Deeben, J., & M.J.L.Th. Niekus Rijndijk (gem. Rijnwoude, prov. Journal of Archaeological Science De steentijd van Nederland, 2016b: Ahrensburger-cultuur, Zuid-Holland): een neolithische 61, 158-171. Zutphen (Archeologie 11/12). in: L. Amkreutz, F. Brounen, vindplaats langs de Oude Rijn, J. Deeben, R. Machiels, Nieuwekerk aan den IJssel, M.-F. van Oorsouw & B. Smit 244-253. 232 —

Demiddele, H. 2011: De Wolf, H. & P. Cleveringa Ducrocq, T., A. Bridault & Fontijn D.R. 2002: Sacrifcial Diatomeeën- en Testacae- 2006: Diatoms, in: L.P. Louwe A.V. Munaut 1991: Un gisement landscapes: cultural biographies of analyse van trechterbeker- en Kooijmans & P.F.B. Jongste mésolithique exceptionnel persons, objects and ‘natural’ places ‘Vlaardingen’-aardewerk, in: (eds.), Schipluiden: a Neolithic dans le nord de la France: le in the Bronze Age of the southern E. Lohof, T. Hamburg & setlement on the Dutch North Sea Petit Marais de la Chaussée- Netherlands, Leiden (PhD J. Flamman (eds.), Steentijd coast c. 3500 cal BC, Leiden Tirancourt (Somme), Bulletin de dissertation Leiden University). opgespoord: archeologisch (Analecta Praehistorica la Société préhistorique française onderzoek in het tracé van de Leidensia 37/38), 285-295. 88, 272-276. Fries, J.E., & S. Veil 2014: Hanzelijn-Oude Land, Fernkontakte späteiszeitlicher Amersfoort/Leiden (ADC report Diependaele, S. 2010: Fens, R.L., & J.P. Mendelts Jäger und Sammler in der 2576/Archol report 138), Resultaten: sporen en 2015: Interpretatie van de norddeutschen Tiefebene – 615-627. structuren, in: S. Diependaele vuursteenvindplaats, in: ein Klingkern aus rotem & E. Drenth (eds.), Archeologisch R.L. Fens & J.P. Mendelts (eds.), Holgoländer Flint von einem De Moor, J.J.W., J.A.A. Bos, onderzoek langs de rijksweg N11 Een steentijdcomplex op het Oberfächenfundplatz am M.T.I.J. Bouman, (Spookverlaat) ten behoeve van de Europapark te Groningen: resten uit Dümmer bei Damme, Ldkr. C. Moolhuizen, R. Exaltus, aanleg van het windturbinepark het laat-mesolithicum en de Diepholz, Siedlungs- und F.P.A. Maartense, Rijnwoude te Hazerswoude- trechterbekercultuur op de Küstenforschung im südlichen T.J.M. van der Linden & Rijndijk (gem. Rijnwoude, prov. oostelijke fank van de Hondsrug, Nordseegebiet 37, 125-140. B.I. Quadfieg 2009: Defnitief Zuid-Holland): een neolithische Groningen (Stadse Fratsen 36), archeologisch onderzoek in het vindplaats langs de Oude Rijn, 65-80. Geerts, R.C.A., A. Müller, tracé van de Hanzelijn in het Nieuwekerk aan den IJssel, M.J.L.Th. Niekus & F.J. Vermue Nieuwe Land: een interdisciplinaire 87-106. Fermin, H.A.C. 2006: Steentijd, 2016: Kampen in de steentijd: geo-archeologische waardering van in: H.M.P. Bouwmeester, mesolithisch onderzoek in het het begraven landschap van Dijk, N., P. Kleij & H. Peeters H.A.C. Fermin & M. Groothedde Reevediep, Archeobrief 4, 14-21. Oostelijk Flevoland, Utrecht. 1994: Mesolithische (eds.), Geschapen landschap: nederzetingsresten in de tienduizend jaar bewoning en Gendel, P.A. 1984: Mesolithic Devriendt, I. 2014: Swiferbant binnenstad van Tilburg (N.-Br.), ontwikkeling van het cultuur- social territories in northwestern stones: the Neolithic stone and fint Archeologie 4, 73-93. landschap op de Looërenk in Europe, Oxford (British industry at Swiferbant (the Zutphen, Deventer (BAAC report Archaeological Reports Netherlands): from stone typology Dijkstra, P., J. Groels & 00.068), 43-84. International Series 218). and fint technology to site function, A. van der Lee 2016: De Groningen (Groningen mesolithische woonstructuren Floss, H. 1994: Rohmaterial- Gendel, P.A. 1989: The analysis Archaeological Studies 25) van Westelbeers, in: F. Snijders versorgung im Paläolithikum des of lithic styles through (PhD dissertation University of & J. Broertjes (eds.), Bivak aan Mitelrheingebietes, Bonn distributional profles of Groningen). de Beerze, 2: laatpaleolithische en (Monographien Römisch- variation: examples from the mesolithische vindplaatsen te Germanisches Zentralmuseum European Mesolithic, in: Devriendt, I. 2015: Vuursteen Westelbeers, Veldhoven, 66-78. 21). C. Bonsall (ed.), The Mesolithic in Ede-Kernhem, in: A. Müller & Europe, Edinburgh, 40-47. I. Devriendt (eds.), Ede Kernhem Dorenbos, O. 2008: Fokkens, H., & R. Jansen 2002: vlek B: opgraving van een Ruimtelijke analyse, in: 2000 Jaar bewoningsdynamiek: Glasbergen, W., mesolithische vindplaats, H. Koot, L. Bruning & thema’s in het metaaltijden- W. Groenmanvan Waateringe Amersfoort (ADC report 3757), R.A. Houkes (eds.), Ypenburg- onderzoek, in: H. Fokkens & & G.M. HardenbergMulder 109-146. locatie 4: een nederzeting met R. Jansen (eds.), 2000 Jaar 1967: Setlements of the grafveld uit het midden-neolithicum bewoningsdynamiek, brons- en Vlaardingen Culture at De Vries, L.S. 2008: Artefacten in het West-Nederlandse ijzertijdbewoning in het Maas- Voorschoten and Leidschendam, van bot en gewei, in: H. Koot, kustgebied, Leiden, 405-441. Demer-Scheldegebied, Leiden, Helinium 7, 331, 97120. L. Bruning & R.A. Houkes (eds.), 1-22. Ypenburg-locatie 4: een Drenth, E., & M.J.L.Th. Niekus Gob, A. 1985: Extension nederzeting met grafveld uit het 2009: 14C-datierte Fokkens, H., B.J.W. Stefens, géographique et chronologique midden-neolithicum in het West- Geröllkeulen aus den S.F.M. van As 2016: Farmers, de la culture Rhein-Meuse- Nederlandse kustgebied, Leiden, Niederlanden, Archäologische fshers, fowlers, hunters, Schelde (RMS), Helinium 25, 263-267. Informationen 32, 91-94. Amersfoort (Nederlandse 23-36. Archeologische Rapporten 53). 233 —

Gordon, B.C. 2009: Groenewoudt, B.J., J. Deeben, Hamburg, T., C. Kruijshaar, Hiddink, H. 2005: Opgravingen Deciphering archaeological B. van Geel & R.C.G.M. J. Nientker, J.H.M. Peeters & op het Rosveld bij Nederweert, 1: palimpsests: an example from Lauwerier 2001: An Early A. Rast-Eicher 2001: landschap en bewoning in de a Canadian barrenland caribou Mesolithic assemblage with Grondsporen: antropogene ijzertijd, Romeinse tijd en hunting camp, in: faunal remains in a stream sporen en structuren, in: middeleeuwen, Amsterdam S.B. McCartan, R. Schulting, valley near Zutphen, the J.W.H. Hogestijn & J.H.M. Peeters (Zuid-Nederlandse G. Warren & P. Woodman Netherlands, Archäologisches (eds.), De mesolithische en vroeg- Archeologische Rapporten (eds.), Mesolithic horizons, Korrespondenzblat 31, 329-348. neolithische vindplaats Hoge 22/1). Oxford (papers presented at Vaart-A27 (Flevoland), Amersfoort the seventh International Groenewoudt, B.J., (Rapportages Archeologische Hodder, I. 1986: Reading the Conference on the Mesolithic M.C. Eerden, T. de Groot & Monumentenzorg 79). past: current approaches to in Europe, Belfast 2005), E.M. Theunissen 2017: Answers interpretation in archaeology, 122-128. to questions: the new National Hamburg, T., & L.P. Louwe Cambridge. Archaeological Research Agenda Kooijmans 2001: Sporen en Groenendijk, H.A. 1987: of the Netherlands, in: structuren, in: L.P. Louwe Hoebe, P. 2015: Hunters on the Mesolithic hearth-pits in the R.C.G.M. Lauwerier, M.C. Eerden, Kooijmans (ed.), Archeologie in Dutch tundra: an examination of Veenkoloniën (prov. B.J. Groenewoudt, M.A. Lascaris, de Betuweroute: Hardinxveld- Hamburgian artefacts excavated on Groningen, the Netherlands): E. Rensink, B.I. Smit, Giessendam Polderweg: een a sandy ridge near Deventer, the defning a specifc use of fre in B.P. Speleers & J. van Doesburg mesolithisch jachtkamp in het Netherlands (12,700-11,900 cal BC), the Mesolithic, Palaeohistoria (eds.), Knowledge for informed rivierengebied (5500-5000 v.C.), Groningen (unpublished report 29, 85-102. choices, Amersfoort (Nederlandse Amersfoort (Rapportage University of Groningen). Archeologische Rapporten 55), Archeologische Monumenten- Groenendijk, H.A. 1997: Op 179-194. zorg 83), 73-104. Högberg, A., & D. Olausson zoek naar de horizon: het landschap 2007: Scandinavian fint: an van Oost-Groningen en zijn Groenewoudt, B.J., & Hamburg, T., & L.P. Louwe archaeological perspective, bewoners tussen 8000 voor C. en J.H.M. Peeters 2006: Kooijmans 2006: Features, in: Aarhus. 1000 na C., Groningen (Regio- Assessment and selection in L.P. Louwe Kooijmans & en Landschapsstudies 4). archaeological heritage P.F.B. Jongste (eds.), Schipluiden: Hogestijn, J.W.H., & E. Drenth management in the Netherlands: a Neolithic setlement on the Dutch 2000/2001: In Slootdorp stond Groenendijk, H.A. 2004: past and future, Berichten van de North Sea coast c. 3500 cal BC, een trechterbeker-huis? Over Middle Mesolithic occupation Rijksdienst voor het Oudheidkundig Leiden (Analecta Praehistorica midden- en laat-neolithische on the extensive site NP3 in the Bodemonderzoek 46, 159-170. Leidensia 37/38), 39-65. huisplategronden uit peat reclamation district of Nederland, Archeologie 10, 42-79. Groningen, the Netherlands, in: Groenewoudt, B.J., P.A.C. Schut, Harsema, O.H. 1982: Setlement P. Crombé & P. Vermeersch F.J.G. van der Heijden, site selection in Drenthe in later Hogestijn, J.W.H., & (eds.), 7: Le mésolithique / The J.H.M. Peeters & prehistoric times: criteria and J.H.M. Peeters (eds.) 2001: De Mesolithic, Oxford (Acts of the M.H. Wispelwey 2006: Een considerations, Analecta mesolithische en vroeg-neolithische XIVth congress of the UISPP, inventariserend veldonderzoek bij de Praehistorica Leidensia 15, 145-160. vindplaats Hoge Vaart-A27 University of Liège, Belgium, Hunneschans (Uddel, Gelderland): (Flevoland), Amersfoort, 20 vols. 2-8 September, 2001) (BAR nieuwe gegevens over de Hartz, S., T. Terberger & (Rapportage Archeologische International Series 1302), steentijdbewoning bij het M. Zhilin 2010: New AMS-dates Monumentenzorg 79). 19-26. Uddelermeer en een beknopt overzicht on the Upper Volga Mesolithic van de onderzoeks geschiedenis van and the origin of microblade Holdaway, S., & L. Wandsnider Groenendijk, H.A., & J.L. Smit de Hunneschans, Amersfoort technology, Quartär 57, 155-169. (eds.) 2008: Time in archaeology: 1990: Mesolithischen (Rapportage Archeologische time perspectivism revisited, Salt Herdstellen: Erfahrungen eines Monumentenzorg 143). Heirbaut, E.N.A. 2010: Lake City. Brennversuchs, Archäologische Vuursteen, in: E.N.A. Heirbaut, Informationen 13, 213-220. Hamburg, T., & J. Hendriks & I. Hermsen (eds.), Holst, D. 2007: Subsistenz und S. Knippenberg 2005: Steentijd Onder een deken van zand: archeo- Landschafsnutzung im Groenendijk, M.J. 2011: op het spoor: proefsleuven op drie logisch onderzoek op de Meshallen en Frühmesolithikum: Nussröstplätze Archeologische basiskaart Gouda, locaties binnen het tracé van de de Kraanvogel te Wijchen, Nijmegen am Duvensee, Mainz (PhD Gouda. Hanzelijn ‘Oude Land’, Leiden (Archeologische Berichten dissertation Johannes (Archol report 54). Wijchen Rapport 10), 91-93. Gutenberg-Universität Mainz). 234 —

Houkes, R.A. 2008: Vuursteen, Ingold, T. 2000: The perception of Kamermans, H., M. van Leusen D.C.M. Raemaekers 2013: in: H. Koot, L. Bruning & the environment: essays on & P. Verhagen 2009: Synthesis – a mater of life and R.A. Houkes (eds.), Ypenburg- livelihood, dwelling and skill, Archaeological prediction and death at Mienakker, in: locatie 4: een nederzeting met London. risk management, in: J.P. Kleijne, O. Brinkkemper, grafveld uit het midden-neolithicum H. Kamermans, M. van Leusen & R.C.G.M. Lauwerier, B.I. Smit & in het West-Nederlandse Innes, J.B., & J.J. Blackford P. Verhagen (eds.), Archaeological E.M. Theunissen (eds.), A mater kustgebied, Leiden, 213-245. 2003: The ecology of Late prediction and risk management: of life and death at Mienakker (the Mesolithic woodland alternatives to current practice, Netherlands): Late Neolithic Houkes, R., & L. Bruning 2008: disturbances: model testing Leiden (Archaeological Series behavioural variability in a dynamic Grondsporen en structuren, in: with fungal spore assemblage Leiden University 17), 9-18. landscape (Nederlandse H. Koot, L. Bruning & data, Journal of Archaeological Archeologische Rapporten 45), R.A. Houkes (eds.), Ypenburg- Science 30, 185-194. Kamstra, H.K., J.H.M. Peeters 249-259. locatie 4: een nederzeting met & D.C.M. Raemaekers 2016: grafveld uit het midden-neolithicum Isaaks, E.H., & R.M. Srivastava The Neolithic stone cist at Knippenberg, S. 2016: The LBK in het West-Nederlandse 1989: An introduction to applied Heveskesklooster (prov. of fint material from the kustgebied, Leiden, 79-110. geostatistics, New York. Groningen, the Netherlands), Cannerberg, in: I.M. van Wijk Palaeohistoria 57/58, 37-54. (ed.), Setlement dynamics on the Hijma, M. 2009: From river valley Jansen J.B.H., & J.H.M. Peeters Cannerberg (Maastricht, the to estuary: the Early-Mid Holocene 2001: Geochemische aspecten: Karsten, P., & B. Knarrström Netherlands): archaeological transgression of the Rhine-Meuse verkenningen in enkele 2003: The Tågerup excavations, research of Bandkeramik and Iron valley, the Netherlands, Utrecht toepassingsmogelijkheden, in: Lund. Age setlements, Leiden (Archol (Netherlands Geographical J.W.H. Hogestijn & J.H.M. Peeters report 300), 161-196. Studies 389) (PhD dissertation (eds.), De mesolithische en vroeg- Katenberg, A., A. de Kraker, Utrecht University). neolithische vindplaats Hoge C. Soonius, P. Stassen, C. Suer, Knippenberg, S., & T. Hamburg Vaart-A27 (Flevoland), Amersfoort P. Verhagen, B. Groenewoudt, 2011: Sporen en structuren, in: Huisman, D.J., A.G. Jongmans (Rapportage Archeologische E. Rensink, H. Peeters, E. Lohof, T. Hamburg & & D.C.M. Raemaekers 2009: Monumentenzorg 79). H. Weerts, M. Vermeulen & J. Flamman (eds.), Steentijd Investigating Neolithic land use H. Fokkens 2008: 6: opgespoord: archeologisch in Swiferbant (NL) using Jensen, H.J. 1994: Flint tools and Archeologische prospectie, onderzoek in het tracé van de micromorphological plant working: hidden traces of Amersfoort (Nationale Hanzelijn-Oude Land, Leiden/ techniques, Catena 78, 185-197. Stone Age technology: a use wear Onderzoekagenda Archeologie Amersfoort (Archol report 138/ study of some Danish Mesolithic and 1.0, htps://archeologieinne- ADC report 2576), 115-207. Huisman, D.J., & B.J.H. van Os TRB implements, Aarhus. derland.nl/sites/default/ 2013: The origin of the red sand fles/6_prospectie.pdf). Knippenberg, S., & A. Verbaas in the Mariënberg graves, a Johansen, L., & D. Stapert 2012: Natuursteen, in: pedological and pedochemical 2004: Oldeholtwolde: Kiden, P. 2006: De evolutie van T. Hamburg, A. Müller & perspective, in: A.D. Verlinde & a Hamburgian family encampment de Beneden-Schelde in België B. Quadfieg (eds.), Mesolithisch R.R. Newell (eds.), The Mesolithic around a hearth, Lisse. en Zuidwest-Nederland na de Swiferbant: mesolithisch gebruik cemetery at Mariënberg (NL), laatste ijstijd, Belgeo 2006-3, van een duin ten zuiden van a rebutal to alternative Jordan, P. 2015: Technology as 279-294. Swiferbant (8300-5000 v.C.): een interpretations, Amersfoort human social tradition: cultural archeologische opgraving in het (Nederlandse Archeologische transmission among hunter- Klejn, L.S. 1982: Archaeological tracé van de N23/N307, provincie Rapporten 42), 33-38. gatherers, Oakland. typology, Oxford (BAR Flevoland, Leiden/Amersfoort International Series 153). (Archol report 174/ADC report Huisman, D.J. & Julien, M. 1987: Activités 3250), 269-298. D.C.M. Raemaekers 2014: saisonnières et déplacements Kleijne, J.P., S.M. Beckerman, Systematic cultivation of the des Magdaléniens dans le D.C. Brinkhuizen, V. García- Kooistra, L.I. 2008: Artefacten Swiferbant wetlands (The Bassin parisien, in: M. Ote Díaz, L. Kubiak-Martens, van hout en bast, in: H. Koot, Netherlands). Evidence from (ed.), Le Magdalénien en Europe: la G.R. Nobles, T.F.M. Oudemans, L. Bruning & R.A. Houkes (eds.), Neolithic tillage marks structuration du Magdalénien, J.T. Zeiler, O. Brinkkemper, Ypenburg-locatie 4: een nederzeting (c. 4300–4000 cal. BC), Journal Liège (Etudes et recherches R.C.G.M. Lauwerier, B.I. Smit, met grafveld uit het midden- of Archaeological Science 49, archéologiques de l’Université de E.M. Theunissen, A.L. van Gijn, neolithicum in het West-Nederlandse 572-584. Liège 38), 177-191. J.H.M. Peeters & kustgebied, Leiden, 269-276. 235 —

Kooistra, L.I. 2011: Houtskool Kooistra, L., & H. Peeters in Leiden (ADC report 2576/Archol The Mesolithic period at the Yangtze en houtgebruik, in: E. Lohof, prep. From land to water: report 138), 465-482. Harbour site – Roterdam T. Hamburg & J. Flamman geomorphological, hydrological Maasvlakte, the Netherlands: Early (eds.), Steentijd opgespoord: and ecological developments in Kubiak-Martens, L., Holocene landscape development archeologisch onderzoek in het Flevoland since the Late Glacial L.I. Kooistra & J.J. Langer 2011: and habitation, Roterdam tracé van de Hanzelijn-Oude Land, till the end of the Subboreal, in: Mesolithische teerproductie in (BOORrapporten 566), 223-286. Amersfoort/Leiden (ADC report H. Peeters, L. Kooistra, Hatemerbroek, in: E. Lohof, 2576/Archol report 138), D. Raemaekers, B. Smit & T. Hamburg & J. Flamman (eds.), Laan, W. 2011: Ruimtelijke 483-496. K. Waugh (eds.), The prehistory of Steentijd opgespoord: archeologisch analyse, in: T. Hamburg, new land: foraging and early farming onderzoek in het tracé van de E. Lohof & B. Quadfieg (eds.), Kooistra, L.I. 2012: Houtskool communities in the drowning Hanzelijn-Oude Land, Amersfoort/ Bronstijd opgespoord: archeologisch uit mesolithische kuilen, in: landscape of the Flevoland polders Leiden (ADC report 2576/Archol onderzoek van prehistorische T. Hamburg, A. Müller & (the Netherlands), Amersfoort. report 138), 497-512. vindplaatsen op Bedrijvenpark H2O B. Quadfieg (eds.), Mesolithisch – plandeel Oldebroek (provincie Swiferbant: mesolithisch gebruik Koot, H., L. Bruning & Kubiak-Martens, L., J.J. Langer Gelderland), Leiden/Amersfoort van een duin ten zuiden van R.A. Houkes (eds.) 2008: & L.I.Kooistra 2012: (Archol report 142/ADC report Swiferbant (8300-5000 v.C.): een Ypenburg-locatie 4: een Plantenresten en teer in 2627), 79-98. archeologische opgraving in het nederzeting met grafveld uit het haardkuilen, in: T. Hamburg, tracé van de N23/N307, provincie midden-neolithicum in het West- A. Müller & B. Quadfieg (eds.), Laarman, F.J. 2001: Flevoland, Leiden/Amersfoort Nederlandse kustgebied, Leiden. Mesolithisch Swiferbant: Archeozoölogie: aard en (Archol report 174/ADC report mesolithisch gebruik van een duin betekenis van de dierlijke 3250), 375-386. Kortekaas, G.L.G.A. & ten zuiden van Swiferbant resten, in: J.W.H. Hogestijn & M.J.L.Th. Niekus 1994: (8300-5000 v.C.): een J.H.M. Peeters (eds.), De Kooistra, L.I., L. Kubiak- Een vindplaats uit het vroege archeologische opgraving in het mesolithische en vroeg-neolithische Martens & J.J. Langer 2009: mesolithicum in de tracé van de N23/N307, provincie vindplaats Hoge Vaart-A27 Archeobotanisch onderzoek, Hooilandspolder, gemeente Flevoland, Leiden/Amersfoort (Flevoland), Amersfoort in: N.M. Prangsma & Slochteren (Gr.), Paleo-Aktueel 5, (Archol report 174/ADC report (Rapportages Archeologische D.A. Gerrets (eds.), Hanzelijn 27-31. 3250), 341-360. Monumentenzorg 79). tunnel Drontermeer: verbinding tussen Oude en Nieuwe Land: een Kranendonk, P., P. van der Kubiak-Martens, L., & Lange, A.G., E.M. Theunissen, archeologische begeleiding bij de Krof, J.J. Lanzing & T.F.M. Oudemans 2011: J.H.C. Deeben, J. van Doesburg, Sallanddijk en een compenserend B.H.F.M. Meijlink (eds.) 2006: Voedselresiduen in trechter- J. Bouwmeester & T. de Groot archeologisch onderzoek in gebied Wite vlekken ingekleurd: beker- en bronstijdaardewerk, (eds.) 2014: Huisplategronden in XVI, Amersfoort (ADC report archeologie in het tracé van de in: T. Hamburg, E. Lohof & Nederland: archeologische sporen 1601), 52-58. HSL-Zuid, Amersfoort B. Quadfieg (eds.), Bronstijd van het huis, Groningen. (Rapportage Archeologische opgespoord: archeologisch Kooistra, L.I., J.J. Langer & Monumentenzorg 113). onderzoek van prehistorische vind- Lanting, J.N., & J. van der Plicht L. Kubiak-Martens 2013: Het plaatsen op Bedrijvenpark H2O – 1998: Reservoir efects and ecologisch onderzoek, in: Kroezenga, P., J.N. Lanting, plandeel Oldebroek (provincie apparent 14C-ages, The Journal of P.J. Ilson (ed.), Jagers en R.J. Kosters, W. Prummel & Gelderland), Leiden/Amersfoort Irish Archaeology 9, 151-165. verzamelaars bij Scheemderzwaag- J.P. de Roever 1991: Vondsten (Archol report 142/ADC report Scheemda en -Opdiep: van de Swiferbantcultuur uit 2627), 453-468. Larsson, L. 1983: Ageröd V: aardgastransportleidingtracé het Voorste Diep bij Bronneger an Atlantic bog site in central Scania, Rysum-Scheemda (A-660) en (Dr.), Paleo-Aktueel 2, 32-36. Kubiak-Martens, L., Lund (Acta Archaeologica Midwolda-Tripscompagnie F. Verbruggen & L.I. Kooistra Lundensia 12). (A-666), KR-042 (A-660) & Kubiak-Martens, L. 2011: 2015: Archaeobotany: landscape KR-006 (A-666), Voedseleconomie: parenchym reconstruction and plant food Larsson, L. 2000: The passage catalogusnummers 16 & 17, en andere plantaardige subsistence economy on a of axes: fre transformation of gemeente Oldambt: archeologisch macroresten, in: E. Lohof, meso and micro scale, in: fint objects in the Neolithic of onderzoek: opgraving en T. Hamburg & J. Flamman (eds.), J.J. Moree & M.M. Sier (eds.), southern Sweden, begeleiding, Weesp (RAAP- Steentijd opgespoord: archeologisch Interdisciplinary Archaeological Antiquity 74, 602-610. rapport 2310), 107-118. onderzoek in het tracé van de Research Programme Maasvlakte 2, Hanzelijn-Oude Land, Amersfoort/ Roterdam, 1: twenty meters deep! 236 —

Lauwerier, R.C.M.G., Lock, G., & B. Molyneaux Louwe Kooijmans, L.P. 2006: Louwe Kooijmans, L.P. & J.G.A. Bazelmans, (eds.) 2006: Confronting scale in Schipluiden: a synthetic view, M. Nokkert 2001: Sporen en O. Brinkkemper, archaeology: issues of theory and in: L.P. Louwe Kooijmans & structuren, in: L.P. Louwe J.H.C. Deeben, J. van Doesburg practice, New York. P.F.B. Jongste (eds.), Schipluiden: Kooijmans (ed.), Archeologie in & P.A.M. Zoetbrood 2006: a Neolithic setlement on the Dutch de Betuweroute: Hardinxveld-De The quality of project outlines Louwe Kooijmans, L.P. 1986: North Sea coast c. 3500 cal BC, Bruin: een kampplaats uit het laat- produced by commercial Het loze vissertje of boerke Leiden (Analecta Praehistorica mesolithicum en het begin van de agencies, Berichten van de Naas? Het een en ander over Leidensia 37/38), 485-516. Swiferbant-cultuur (5500-4450 v. Rijksdienst voor het Oudheidkundig het leven van de steentijd- C.), Amersfoort (Rapportage Bodemonderzoek 46, 91-109. bewoners van het Rijnmond- Louwe Kooijmans, L.P. 2007: Archeologische gebied, in: M.C. van Trierum & Multiple choices: mortuary Monumentenzorg 88), 75-115. Lauwerier, R.C.M.G., & H.E. Henkes (eds.), Roterdam practices in the Low Countries J. Deeben 2011: Burnt animal papers V, a contribution to during the Mesolithic and Louwe Kooijmans, L.P., remains from Feddermesser prehistoric, Roman and medieval Neolithic, 9000-3000 cal BC, J.F.S. Oversteegen & sites in the Netherlands, archaeology, Roterdam, 7-25. Berichte Römisch-Germanische A.L. van Gijn 2001: Artefacten Archäologisches Kommission 88, 551-580. van been, gewei en tand, in: Korrespondenzblat 41, 1-20. Louwe Kooijmans, L.P. (ed.) L.P. Louwe Kooijmans (ed.), 2001a: Archeologie in de Louwe Kooijmans, L.P. 2012: Archeologie in de Betuweroute: Lauwerier, R.C.M.G., Betuweroute: Hardinxveld- Refections on the Mesolithic Hardinxveld-Giessendam M.C. Eerden, B.J. Groenewoudt, Giessendam Polderweg: een burial pits at Mariënberg Polderweg: een mesolithisch M.A. Lascaris, E. Rensink, mesolithisch jachtkamp in het (province of Overijssel), jachtkamp in het rivierengebied B.I. Smit, B.P. Speleers & rivierengebied (5500-5000 v. C.), the Netherlands, in: (5500-5000 v. C.), Amersfoort J. van Doesburg 2017: Knowledge Amersfoort (Rapportages M.J.L.Th. Niekus, R.N.E. Barton, (Rapportages Archeologische for informed choices: tools for more Archeologische M. Street & T. Terberger (eds.), Monumentenzorg 83), efective and efcient selection of Monumentenzorg 83). A mind set on fint: studies in 285-324. valuable archaeology in the honour of Dick Stapert, Groningen Netherlands, Amersfoort Louwe Kooijmans, L.P. (ed.) (Groningen Archaeological Louwe Kooijmans, L.P., (Nederlandse Archeologische 2001b: Archeologie in de Studies 16), 401-424. C.E. Vermeeren & Rapporten 55). Betuweroute: Hardinxveld-De A.M.I. Van Waveren 2001: Bruin: een kampplaats uit het laat- Louwe Kooijmans, L.P., Artefacten van hout en vezels, Leary, J. 2009: Perceptions of mesolithicum en het begin van de K. Hänninen & C.E. Vermeeren in: L.P. Louwe Kooijmans (ed.), and responses to the Holocene Swiferbant-cultuur (5500-4450 v. 2001: Artefacten van hout, in: Archeologie in de Betuweroute: fooding of the North Sea C.), Amersfoort (Rapportages L.P. Louwe Kooijmans (ed.), Hardinxveld-Giessendam lowlands, Oxford Journal of Archeologische Archeologie in de Betuweroute: Polderweg: een mesolithisch Archaeology 28(3), 227-237. Monumentenzorg 88). Hardinxveld-De Bruin: een jachtkamp in het rivierengebied kampplaats uit het laat- (5500-5000 v. C.), Amersfoort Leenders, K.A.H.W. 1996: Louwe Kooijmans, L.P. 2003: mesolithicum en het begin van de (Rapportage Archeologische Van Turnhoutervoorde tot The Hardinxveld sites in the Swiferbant-cultuur (5500-4450 v. Monumentenzorg 83), 379-418. Strienemonde: ontginnings- en Rhine/Meuse delta, the C.), Amersfoort (Rapportages nederzetingsgeschiedenis van het Netherlands, 5500-4500 cal BC, Archeologische Lovis, W.A. 2016: Is there a noordwesten van het Maas- in: L. Larsson, H. Kindgren, Monumentenzorg 88), research design role for Schelde-Demergebied (400-1350): K. Knutsson, D. Loefer & 435-477. sensitivity analysis (SA) in een poging tot synthese, Zutphen. A. Åkerlund (eds.), Mesolithic on archaeological modeling? in: the move: papers presented at the Louwe Kooijmans, L.P. & M. Brouwer Burg, H. Peeters & Lemonnier, P. 1992: Elements for sixth International Conference on L. Kooistra 2006: Wooden W. Lovis (eds.), Uncertainty and an anthropology of technology, the Mesolithic in Europe, Stockholm artefacts, in: L.P. Louwe sensitivity analysis in archaeological Michigan (Anthropological 2000, Oxford, 608-624. Kooijmans & P.F.B. Jongste computational modeling, papers of the Museum of (eds.), Schipluiden: a Neolithic New York (Interdisciplinary Anthropology 88). setlement on the Dutch North Sea Contributions to Archaeology), coast c. 3500 cal BC, Leiden 21-36. (Analecta Praehistorica Leidensia 37/38), 225-251. 237 —

Mellars, P. 1976: Fire ecology, Niekus, M.J.L.Th. 2009: een opgraving van een vindplaats uit Niekus, M.J.L.Th., A. Verbaas, animal populations and man: Trapeze shaped fint tips as het neolithicum op een rivierduin op H. de Kruyk & J.J. Boon 2015: a study of some ecological proxy data for occupation IJsselmonde (vindplaats 13-78), Flint and other stone, in: relationships in prehistory, during the Late Mesolithic and Roterdam (BOORrapporten J.M. Moree & M.M. Sier (eds.), Proceedings of the Prehistoric the Early to Middle Neolithic in 491), 55-79. Interdisciplinary Archaeological Society 42, 15-45. the northern part of the Research Programme Maasvlakte Netherlands, Journal of Niekus, M.J.L.Th., 2, Roterdam, 1: twenty meters Mellars, P. & P. Dark (eds.) Archaeological Science 36, P.H.J.I. Ploegaert, J.T. Zeiler & deep! The Mesolithic period at the 1998: Star Carr in context: 236-247. L. Smits 2016: A small Middle Yangtze Harbour site – Roterdam new archaeological and Mesolithic cemetery with Maasvlakte, the Netherlands: Early palaeoecological investigations at Niekus, M.J.L.Th., cremation burials from Holocene landscape development the Early Mesolithic site of Star Carr, G.R. Boekschoten & Roterdam, the Netherlands / Ein and habitation, Roterdam North Yorkshire, Oxford. J.H.C. Deeben in prep.: A Late kleiner mitelmesolithischer (BOORrapporten 566), 147-200. Preboreal site from Zwolle, Friedhof mit Brandbestatungen Merlo, S. 2010: Contextualising province of Overijssel, and von Roterdam, in: Nobles, G.R. 2012: Spatial intra-site spatial analysis: the role some remarks on the J.M. Grünberg, B. Gramsch, analysis, in: Smit, B.I., of three-dimensional GIS modelling Ahrensburgian in the L. Larsson, J. Orschiedt & O. Brinkkemper, J.P. Kleijne, in understanding excavation data, Netherlands, in: B.V. Eriksen, H. Meller (eds.), Mesolithic burials R.C.G.M. Lauwerier & Cambridge (PhD dissertation S. Harris & E. Rensink (eds.), – rites, symbols and social E.M. Theunissen (eds.), Cambridge University). Proceedings of the meetings of organization of early postglacial A kaleidoscope of gathering at UISPP commission ‘The fnal communities / Mesolithische Keinsmerbrug (the Netherlands): Modderman, P.J.R. 1970: Palaeolithic of northern Eurasia’ Bestatungen – Riten, Symbole und Late Neolithic behavioural Linearbandkeramik aus Elsloo und held in Amersfoort (2012) and soziale Organisation früher variability in a dynamic landscape, Stein, Leiden/Amersfoort Schleswig (2013), Schleswig. postglazialer Gemeinschafen, Halle Amersfoort (Nederlandse (Analecta Praehistorica Leidensia (proceedings of the International Archeologische Rapporten 43), 3/Nederlandse Oudheden 3). Niekus, M.J.L.Th., Conference Halle (Saale), 149-210. H.H. Huisman & Germany, 18th-21st September Musch, J., & H. Peeters 1993: J.H.M. Peeters, in prep.: 2013) (Tagungen des Nobles, G.R. 2013: Spatial Signalement: klingproductie Antiligent design: on the Landesmuseums für analysis, in: J.P. Kleijne, met de druktechniek in het anthropogenic nature of Vorgeschichte Halle 13), 569-592. O. Brinkkemper, Nederlandse mesolithicum? Mesolithic pit hearths and the R.C.G.M. Lauwerier, B.I. Smit & Archeologie 4, 83-91. burnt ant nest hypothesis Niekus, M.J.L.Th. & B. Smit E.M. Theunissen (eds.), A mater (working title). 2006: Wie het kleine niet eert … of life and death at Mienakker (the Newell, R.R. 1973: The post- Micro-driehoeken in het Netherlands): Late Neolithic glacial adaptations of the Niekus, M.J.L.Th., mesolithicum van Noord- behavioural variability in a dynamic indigenous population of the S. Knippenberg & I. Devriendt Nederland, Paleo-Aktueel 17, landscape, Amersfoort northwest European Plain, in: 2012: Vuursteen, in: 46-55. (Nederlandse Archeologische S.K. Kozlowski (ed.), The Mesolithic T. Hamburg, A. Müller & Rapporten 45),185-240. in Europe, Warsaw, 399-440. B. Quadfieg (eds.), Mesolithisch Niekus, M.J.L.Th., Swiferbant: mesolithisch gebruik S. Knippenberg & Nobles, G.R. 2014: Spatial Newell, R.R. 1975: van een duin ten zuiden van I.I.J.A.L.M. Devriendt 2012: 7. analysis, in: E.M. Theunissen, Mesolithicum, in: G.J. Verwers Swiferbant (8300-5000 v.C.): een Vuursteen, in: T. Hamburg, O. Brinkkemper, (ed.), Noord-Brabant in pre- en archeologische opgraving in het A. Müller & B. Quadfieg R.C.G.M. Lauwerier, B.I. Smit & protohistorie, Oosterhout, 39-54. tracé van de N23/N307, provincie (eds.), Mesolithisch Swiferbant: I.M.M. van der Jagt (eds.), Flevoland, Leiden/Amersfoort mesolithisch gebruik van een duin A mosaic of habitation at Zeewijk Niekus, M.J.L. Th. 2006: (Archol report 174/ADC report ten zuiden van Swiferbant (the Netherlands): Late Neolithic A geographically referenced 14C 3250), 157-243. (8300-5000 v.Chr.). Een behavioural variability in a dynamic database for the Mesolithic and archeologische opgraving in het landscape, Amersfoort the early phase of the Niekus, M.J.L.Th., & R. Machiels tracé van de N23/N307, provincie (Nederlandse Archeologische Swiferbant culture in the 2013: Vuursteen en natuursteen, Flevoland. Leiden/Amersfoort Rapporten 47), 197-255. northern Netherlands, in: D.E.A. Schiltmans (ed.), (ARCHOL-rapport 174; Palaeohistoria 47/48 (2005/2006), Roterdam Groenenhagen- ADC-rapport 3250), 157-241. 41-99. Tuinenhoven, De Zwanen-Rietpark: 238 —

Nobles, G.R. 2016: Dwelling on G. Kulcsár, V. Kiss, A. Czene, Peeters, J.H.M. 2001: Het Peeters, H. 2010: Early the edge of the Neolithic: R. Patay, A. Endródi, K. Köhler, vuursteenmateriaal van de Swiferbant potery from Hoge investigating human behaviour T. Hajdu, J.L. Cardoso, trechterbekervindplaats Vaart-A27 (Almere, the through the spatial analysis of C. Liesau, M. Parker Pearson, Bouwlust bij Slootdorp (gem. Netherlands), in: Corded Ware setlement material in P. Włodarczak, T.D. Price, Wieringermeer, prov. N.H.), in: B. Vanmontfort, L. Amkreutz, the Dutch coastal wetlands P. Prieto, P.-J. Rey, P. Ríos, R.M. van Heeringen & L.P. Louwe Kooijmans & (2900-2300 cal BC), Groningen R. Risch, M.A. Rojo Guerra, E.M. Theunissen (eds.), L. Verhart (eds.), Early potery (PhD dissertation University of A. Schmit, J. Serralongue, Kwaliteitsbepalend onderzoek ten traditions in the lower Rhine area, Groningen). A.M. Silva, V. Smrčka, behoeve van duurzaam behoud van Leiden, 151-160. L. Vergnaud, J. Zilhão, neolithische terreinen in West- Noens, G. 2011: Een afgedekt D. Caramelli, T. Higham, Friesland en de Kop van Noord- Peeters, J.H.M. 2016: mesolithisch nederzetingsterrein te V. Heyd, A. Sheridan, Holland, 3: archeologische Trechterbeker-cultuur, in: Hempens/N31 (gemeente K.-G. Sjögren, M.G. Thomas, onderzoeksverslagen, Amersfoort L. Amkreutz, F. Brounen, Leeuwarden, provincie Friesland, P.W. Stockhammer, R. Pinhasi, (Nederlandse Archeologische J. Deeben, R. Machiels, Nl.), Ghent (Archaeological J. Krause, W. Haak, I. Barnes, Rapporten 21), 661-716. M.F. van Oorsouw & B. Smit Reports Ghent University 7). C. Lalueza-Fox & D. Reich 2017: (eds.): Vuursteen verzameld: over The beaker phenomenon and Peeters, J.H.M. 2007: Hoge het zoeken en onderzoeken van Olalde, I., S. Brace, the genomic transformation of Vaart-A27 in context: towards a steentijdvondsten en -vindplaatsen, M.E. Allentof, I. Armit, northwest Europe, Antiquity 81, model of Mesolithic-Neolithic land Amersfoort (Nederlandse K. Kristiansen, N. Rohland, 343-352. use dynamics as a framework for Archeologische Rapporten 50), S. Mallick, T. Booth, archaeological heritage 176-179. A. Szécsényi-Nagy, Opbroek, M. & T. Hamburg management, Amersfoort. A. Mitnik, E. Altena, 2011: Sporen en structuren, in: Peeters, J.H.M., M. Lipson, I. Lazaridis, T. Hamburg, A. Müller & Peeters, H. 2009a: Early D.C. Brinkhuizen, K.M. Cohen, N.J. Paterson, B. Quadfieg (eds.), Mesolithisch Holocene landscape dynamics L.I. Kooistra, L. Kubiak- N. Broomandkhoshbacht, Swiferbant: mesolithisch gebruik and forager land-use diversity: Martens, J.M. Moree, Y. Diekmann, Z. Faltyskova, van een duin ten zuiden van the example of Hoge Vaart-A27 M.J.L.Th. Niekus, D.M. Fernandes, M. Ferry, Swiferbant (8300-5000 v.C.): een (Almere, the Netherlands), in: D.E.A. Schiltmans, A.Verbaas, E. Harney, P. de Knijf, archeologische opgraving in het S.B. McCartan, R. Schulting, F. Verbruggen, P.C. Vos & M. Michel, J. Oppenheimer, tracé van de N23/N307, provincie G. Warren & P. Woodman J.T. Zeiler 2015: Synthesis, in: K. Stewardson, A. Barclay, Flevoland, Leiden/Amersfoort (eds.), Mesolithic horizons: J.J. Moree & M.M. Sier (eds.), K.W. Alt, A. Avilés Fernández, (Archol report 174/ADC report papers presented at the Interdisciplinary Archaeological E. Bánfy, M. Bernabò-Brea, 3250), 113-146. seventh International Research Programme Maasvlakte D. Billoin, C. Blasco, C. Bonsall, Conference on the Mesolithic 2, Roterdam, 1: twenty meters L. Bonsall, T. Allen, L. Büster, Opbroek, M. & E. Lohof 2012: in Europe, Belfast 2005, deep! The Mesolithic period at the S. Carver, L. Castells Navarro, Tijd in centimeters: een kijkje in het Oxford, 269-276. Yangtze Harbour site – Roterdam O.E. Craig, G.T. Cook, landschap van een dekzandrug te Maasvlakte, the Netherlands: Early B. Cunlife, A. Denaire, Almere: een inventariserend Peeters, H. 2009b: ‘Occupation Holocene landscape development K. Egging Dinwiddy, veldonderzoek in de vorm van continuity’ and ‘behavioural and habitation, Roterdam N. Dodwell, M. Ernée, proefsleuven en een hoogwaardig discontinuity’: abrupt changes (BOORrapporten 566), 287-318. C. Evans, M. Kuchařík, booronderzoek, Amersfoort in forager land use at the Mid/ J. Francès Farré, H. Fokkens, (ADC report 2662). Late Atlantic boundary in the Peeters, J.H.M. & I. Devriendt C. Fowler, M. Gazenbeek, Flevoland polders (the 2016: Swiferbant-cultuur, in: R. Garrido Pena, M. Haber- Out, W.A. 2009: Sowing the Netherlands), in: P. Crombé & L. Amkreutz, F. Brounen, Uriarte, E. Haduch, G. Hey, seed? Human impact and plant M. van Strydonck (eds.), J. Deeben, R. Machiels, N. Jowet, T. Knowles, subsistence in Dutch wetlands Chronology and evolution in the M.-F. van Oorsouw & B. Smit K. Massy, S. Pfrengle, during the Late Mesolithic and Early Mesolithic of N(W) Europe, (eds.), Vuursteen verzameld: over P. Lefranc, O. Lemercier, and Middle Neolithic (5500-3400 Cambridge, 721-736. het zoeken en onderzoeken van A. Lefebvre, J. Lomba cal BC), Leiden (Archaeological steentijdvondsten en -vindplaatsen, Maurandi, T. Majó, Series Leiden University 18) Amersfoort (Nederlandse J.I. McKinley, K. McSweeney, (PhD dissertation Leiden Archeologische Rapporten 50), M. Balázs Gusztáv, A. Modi, University). 150-155. 239 —

Peeters, J.H.M. & Peeters, H. & J.W. Romeijn modèles régionaux de peuplement: Prummel, W. & J.W.H. Hogestijn 2001: Deel 20: 2016: Epistemic considerations actes de la table-ronde de Nemours, M.J.L.Th. Niekus 2011: Late Op de grens van land en water: about uncertainty and model mai 1997, Nemours (Mémoire Mesolithic hunting of a small jagers-vissers-verzamelaars in selection in computational du Musée de Préhistoire d’Ile- female aurochs in the valley of een verdrinkend landschap, in: archaeology: a case study on de-France 7), 73-86. the River Tjonger (the J.W.H. Hogestijn & J.H.M. Peeters exploratory modeling, in: Netherlands) in the light of (eds.): De mesolithische en vroeg- M. Brouwer Burg, H. Peeters & Perry, D. 1997: The archaeology Mesolithic aurochs hunting in neolithische vindplaats Hoge W.A. Lovis (eds.), Uncertainty and of hunter-gatherers: plant use in the NW Europe, Journal of Vaart-A27 (Flevoland), Amersfoort sensitivity analysis in archaeological Dutch Mesolithic, New York Archaeological Science 38, (Rapportage Archeologische computational modelling, Cham, (unpublished PhD dissertation 1456-1467. Monumentenzorg 79). 37-58. New York University). Raemaekers, D.C.M. 2001: Peeters, H., L. Kooistra, Peeters, J.H.M., J. Schreurs & Perry, D. 2002: Preliminary Aardewerk en verbrande klei, D. Raemaekers, B. Smit & S.M.J.P. Verneau 2001: results of an archaeobotanical in: L.P. Louwe Kooijmans (ed.), K. Waugh in prep.: The prehistory Vuursteen: typologie, analysis of Mesolithic sites in Archeologie in de Betuweroute: of new land: foraging and early technologische organisatie en the Veenkoloniën, province of Hardinxveld-De Bruin: een farming communities in the drowning gebruik, in: J.W.H. Hogestijn & Groningen, the Netherlands, in: kampplaats uit het laat- landscape of the Flevoland polders J.H.M. Peeters (eds.), De S.L.R. Mason & J.G. Hather mesolithicum en het begin van de (the Netherlands), Amersfoort. mesolithische en vroeg-neolithische (eds.), Hunter-gatherer Swiferbant-cultuur vindplaats Hoge Vaart-A27 archaeobotany: perspectives from (5500-4450 v. C.), Amersfoort Peeters, H. & M.J.L.Th. Niekus (Flevoland), Amersfoort the temperate zone, London, (Rapportage Archeologische 2005: Het mesolithicum in (Rapportage Archeologische 108-116. Monumentenzorg 88), 117-152. Noord-Nederland, in: J. Deeben, Monumentenzorg 79). E. Drenth, M.-F. van Oorsouw & Peters, F.J.C. & J.H.M. Peeters Raemaekers, D.C.M. 2008: L. Verhart (eds.), De steentijd van Peeters, H., S. Verneau & 2001: De opgraving van de The Schipluiden potery: Nederland, Zutphen (Archeologie M. Admiraal 2015: mesolithische en neolithische mobility, exchange and mode 11/12), 201-234. Vuursteenonderzoek in Epse: vindplaats Urk-E4 (Domineesweg, of production, in: H. Fokkens, opgravingscampagne Olthof- gemeente Urk), Amersfoort B.J. Coles, A.L. van Gijn, Peeters, H. & M.J.L.Th. Niekus Zuid 2010, in: I. Hermsen, (Rapportage Archeologische J.P. Kleijne, H.H. Ponjee & 2017: Mesolithic pit hearths in M. van der Wal & H. Peeters Monumentenzorg 93). C.G. Slappendel (eds.), Between the northern Netherlands: (eds.), Afslag Olthof: archeologisch foraging and farming: an extended function, time-depth and onderzoek naar de vroeg- Pétrequin, A.M. & P. Pétrequin broad spectrum of papers presented behavioural context, in: prehistorische vindplaatsen op de 1990: Flèches de chasse, to Leendert Louwe Kooijmans, N. Achard-Corompt, locaties Olthof-Noord en Olthof- fèches de guerre: le cas des Leiden (Analecta Praehistorica E. Ghesquière & V. Riquer (eds.), Zuid in Epse-Noord, Deventer Danis d’Irian Jaya (Indonésie), Leidensia 40), 131-138. Creuser au Mésolithique / Digging in (Rapportage Archeologie Bulletin de la Société Préhistorique the Mesolithic: actes de la séance de Deventer 34), 143-162. Française, 484-511. Raemaekers, D.C.M., la Société préhistorique française de C.C. Bakels, B. Beerenhout, Châlons-en-Champagne (29-30 Pélegrin, J., C. Karlin & P. Bodu Price, T.D. 1981: Swiferbant, A.L. van Gijn, K. Hänninen, mars 2016), Paris (Séances de la 1988: Chaînes opératoires: un Oost Flevoland, Netherlands: S. Molenaar, D. Paalman, Société préhistorique française outil pour le préhistorien, in: excavations at the river dune M. Verbruggen & 12), 225-239. J. Tixier (ed.), Technologie sites, S21-S24, 1976: fnal C. Vermeeren 1997: préhistorique, Paris, 55-62. reports on Swiferbant III, Wateringen 4: a coastal Peeters, H. & E. Rensink, in Palaeohistoria 23, 75-104. setlement of the Middle prep.: Hamburgian hunter- Pélegrin J. 2000: Les Neolithic Hazendonk 3 Group, gatherers in the ice-pushed techniques de débitage Prummel, W. 2016: Mens en Analecta Praehistorica Leidensia hills of the central Netherlands, laminaire au Tardiglaciaire: dier in de Swiferbantcultuur, 29, 143-191. in: B.V. Eriksen, S. Harris & critères de diagnose et in: W. Prummel, J.P. de Roever E. Rensink (eds.), Proceedings of quelques réfexions, in: & A.F.L. van Holk (eds.), the meetings of UISPP Commission B. Valentin, P. Bodu & Swiferbant: pionieren in Flevoland ‘The Final Palaeolithic of northern M. Christensen (eds.), L’Europe 6500 jaar geleden, Groningen, Eurasia’ held in Amersfoort (2012) centrale et septentrionale au 51-61. and Schleswig (2013), Schleswig. Tardiglaciaire: confrontation des 240 —

Raemaekers, D.C.M., Rensink, E. 1993: Moving into the Rigaud, J. & J.F. Simek 1991: the interpretation of past fre L. Kubiak-Martens, North: Magdalenian occupation Interpreting spatial paterns at practices in the landscape, T.F.M. Oudemans 2013: and exploitation of the loess the Grote XV: a multiple Current Anthropology 56(3), New food in old pots – charred landscapes of northwestern Europe, method approach, in: 299-326. organic residues in Early Leiden (unpublished PhD E.M. Kroll & T.D. Price (eds.), Neolithic ceramic vessels from dissertation Leiden University). The interpretation of archaeological Schifer, M. 1999: The material life Swiferbant, the Netherlands spatial paterning, New York, of human beings: artifacts, behavior (4300-4000 cal. BC), Rensink, E. 2010: Eyserheide: a 199-220. and communication, London. Archäologisches Korrespondenzblat Magdalenian open-air site in the 43(3), 315-334. loess area of the Netherlands and Robinson, E., M. Van Strydonck, Schlanger, N. 1994: Mindful its archaeological context, Leiden V. Gelorini & P. Crombé 2013: technology: unleashing the Raemaekers, D.C.M., (Analecta Praehistorica Radiocarbon chronology and the chaîne opératoire for an H.M. Molthof & L. Smits 2007: Leidensia 42). correlation of hunter-gatherer archaeology of mind, in: The textbook ‘dealing with sociocultural change with abrupt C. Renfrew & E. Zubrow (eds.), death’ from the Neolithic Rensink, E. & J.W. de Kort palaeoclimate change: the The ancient mind: elements of Swiferbant Culture (5000-3400 2012: A new Hamburgian site Middle Mesolithic in the Rhine- cognitive archaeology, Cambridge BC), the Netherlands, Berichte near Stroe (municipality of Meuse-Scheldt area of (New Directions in Archaeology), Römisch-Germanische Kommission Barneveld), the Netherlands, northwest Europe, Journal of 143-151. 88, 479-500. in: M.J.L.Th. Niekus, Archaeological Science 40, 755-763. R.N.E. Barton, M. Street & Schlanger, N. 2005: The history Raemaekers, D., & M. Rooke T. Terberger (eds.), A mind set on Roebroeks, W. 2014: of a special relationship: 2006: The Schipluiden potery, fint: studies in honour of Dick Terra incognita: the Palaeolithic prehistoric terminology and in: L.P. Louwe Kooijmans & Stapert, Groningen (Groningen record of northwestern Europe lithic technology between the P.F.B. Jongste (eds.), Schipluiden: Archaeological Studies 16), and the information potential French and South African a Neolithic setlement on the Dutch 235-250. of the southern North Sea, research traditions, in: F. d’Errico North Sea coast c. 3500 cal. BC, Netherlands Journal of Geosciences & L. Backwell (eds.), From tools to Leiden (Analecta Praehistorica Rensink, E. & J.W. de Kort 2013: 93(1/2), 43-54. symbols: from early hominids to Leidensia 37/38), 113-128. Archeologische boringen langs Groot modern humans, Johannesburg, Zeilmeer: waarderend onderzoek van Rossignol, J., & L. Wandsnider 9-37. Regert, M. 2004: Investigating een vindplaats van de Hamburg- (eds.) 1992: Space, time, and the history of prehistoric glues cultuur bij Meerveld (gemeente archaeological landscapes, Schulting, R.J. 2005: ‘… by gas chromatography-mass Apeldoorn), Amersfoort New York (Interdisciplinary pursuing a rabbit in Burrington spectrometry, Journal of (Rapportage Archeologische Contributions to Archaeology). Combe’: new research on the Separation Science 27, 244-254. Monumentenzorg 216). Early Mesolithic burial cave of Sacket, J.R. 1982: Approaches Aveline’s Hole, Proceedings of the Regnell, M. 2012: Plant Rensink, E. & M.J.L.Th. Niekus to style in lithic archaeology, University of Bristol Spelaeological subsistence and environment 2016: Hamburg-cultuur, in: Journal of Anthropological Society 23,171-265. at the Mesolithic site Tågerup, L. Amkreutz, F. Brounen, Archaeology 1, 59-112. southern Sweden: new insights J. Deeben, R. Machiels, Schulting, R.J. 2010: Holocene on the ‘nut age’, Vegetation M.-F. van Oorsouw & B. Smit Schepers, M. 2014: environmental change and the History and Archaeobotany 21, (eds.), Vuursteen verzameld: over Reconstructing vegetation diversity Mesolithic-Neolithic transition 1-16. het zoeken en onderzoeken van in coastal landscapes, Groningen in northwest Europe: revisiting steentijd vondsten en -vindplaatsen, (Advances in Archaeobotany 1) two models, Environmental Regnell, M., M.J. Gaillard, Amersfoort (Nederlandse (PhD dissertation University of Archaeology 15,160-172. T.S. Bartholin & P. Karsten Archeologische Rapporten 50), Groningen). 1995: Reconstruction of 86-93. Schulting, R.J. & M.P. Richards environment and history of Scherjon, F., C. Bakels, 2001: Dating women and plant use during the Late Rensink, E., H.J.T. Weerts, K. MacDonald & becoming farmers: new Mesolithic (Ertebølle Culture) M. Kosian, H. Feiken & W. Roebroeks 2015: Burning palaeodietary and AMS data at the inland setlement of B.I. Smit 2015: Archeologische the land: an ethnographic study from the Breton Mesolithic Bökeberg III, southern Sweden, landschappenkaart van Nederland: of of-site fre use by current cemeteries of Téviec and Hoëdic, Vegetation History and methodiek en kaartbeeld, and historically documented Journal of Anthropological Archaeobotany 4, 67-91. Amersfoort. foragers and implications for Archaeology 20, 314-344. 241 —

Schulting, R.J. & M.P. Richards Late Neolithic behavioural Lithics in context: suggestions for Tixier, J., M.-L. Inizan & 2002: The wet, the wild and the variability in a dynamic landscape, the future direction of lithic studies, H. Roche 1980: Préhistoire de la domesticated: the Mesolithic- Amersfoort (Nederlandse London (Lithic Studies Society pierre taillée, 1: terminologie et Neolithic transition on the west Archeologische Rapporten 43), Occasional Paper 5), 125-137. technologie, Paris. coast of Scotland, European 211-222. Journal of Archaeology 5,147-189. Stapert, D. 1981: A site of the Theunissen, E.M., Smith, O., G. Momber, Hamburg tradition on the S.M. Beckerman, Sergant, J., P. Crombé & R. Bates, P. Garwood, S. Fitch, Wadden island of Texel D.C. Brinkhuizen, V. García- Y. Perdaen 2006: The ‘invisible’ M. Pallen, V. Gafney & (province of North Holland, Díaz, L. Kubiak-Martens, hearths: a contribution to the R.G. Allaby 2015: Sedimentary Netherlands), Palaeohistoria 23, G.R. Nobles, T.F.M. Oudemans, discernment of Mesolithic DNA from submerged site 1-27. J.T. Zeiler, O. Brinkkemper, non-structured surface provides evidence of wheat in I.M.M. van der Jagt, J.P. Kleijne, hearths, Journal of Archaeological the British Isles 8000 years Stapert, D. 1992: Rings and R.C.G.M. Lauwerier, B.I. Smit, Science 33, 999-1007. before present, Science 347, sectors: intrasite spatial analysis of A.L. van Gijn, J.H.M. Peeters & 998-1001. Stone Age sites, Groningen (PhD D.C.M. Raemaekers 2014: Simmons, I.G. 1996: The dissertation University of Synthesis –a mosaic of environmental impact of later Smits, L. & L. Louwe Kooijmans Groningen). habitation at Zeewijk, in: Mesolithic cultures: the creation of 2006: Graves and human E.M. Theunissen, moorland landscape in England remains, in: L.P. Louwe Stäuble, H. 1997: Häuser, O. Brinkkemper, and Wales, Edinburgh. Kooijmans & P.F.B. Jongste Gruben und Fundverteilung, in: R.C.G.M. Lauwerier, B.I. Smit & (eds.), Schipluiden: a Neolithic J. Lüning (ed.): Ein Siedlungplatz I.M.M. van der Jagt (eds.), Slon, V., C. Hopfe, C.L. Weiss, setlement on the Dutch North Sea der ältesten Bandkeramik in A mosaic of habitation at Zeewijk F. Mafessoni, M. de la Rasilla, coast c. 3500 cal. BC, Leiden Bruchenbrücken, Stadt Friedberg/ (the Netherlands): Late Neolithic C. Laluezza-Fox, A. Rosas, (Analecta Praehistorica Hessen, Bonn, 17-150. behavioural variability in a dynamic M. Soressi, M.V. Knul, R. Miller, Leidensia 37/38), 91-112. landscape, Amersfoort J.R. Stewart, A.P. Derevianko, Ten Anscher, T.J. 2012: Leven (Nederlandse Archeologische Z. Jacobs, B. Li, R.G. Roberts, Smits, L. & J. van der Plicht met de Vecht: Schokland-P14 en de Rapporten 47), 257-266. M.V. Shunkov, H. de Lumley, 2009: Mesolithic and Neolithic Noordoostpolder in het neolithicum C. Perrenoud, I. Gusi, Z. Kucan, human remains in the en de bronstijd, Amsterdam (PhD Tol, A., P. Verhagen, P. Rudan, A. Aximu-Petri, Netherlands: physical dissertation University of A. Borsboom & M. Verbruggen E. Essel, S. Nagel, B. Nickel, anthropological and stable Amsterdam). 2004: Prospectief boren: een studie A. Schmidt, K. Prüfer, J. Kelso, isotope investigations, Journal naar de betrouwbaarheid en H.A. Burbano, S. Pääbo & of Archaeology in the Low Countries Terberger, T., H. Floss, toepasbaarheid van booronderzoek M. Meyer 2017: Neanderthal 1, 58-85. Ph. Heinzemann, A. Kotula & in de prospectiearcheologie, and Denisovan DNA from J. Serangeli 2013: Down the Amsterdam (RAAP-rapport Pleistocene sediments, Science Sørensen, M., T. Rankama, Rhine River ca. 16,000 years ago: 1000). 356, 605-608. J. Kankaanpää, K. Knutsson, new evidence from the site H. Knutsson, St.A. Melvold, Dreieich-Götzenhain, Hesse, in: Tolan-Smith, C. 2008: Smit, B.I., S.M. Beckerman, B.V. Eriksen & H. Glørstad A. Pastorrs & B. Aufermann Mesolithic Britain, in: G. Bailey D.C. Brinkhuizen, V. Garcia- 2013: The frst eastern (eds.), Pleistocene Foragers: their & P. Spikins (eds.), Mesolithic Diaz, L. Kubiak-Martens, migrations of people and culture and environment: Festschrif Europe, Cambridge, 132-157. G.R. Nobles, T.F.M. Oudemans, knowledge into Scandinavia: G.-Ch. Weniger zum 65. Geburtstag J.T. Zeiler, O. Brinkkemper, evidence from studies of (Wissenschafliche Tolksdorf, J.F., K. Kaiser, J.P. Kleijne, R.C.G.M. Lauwerier, Mesolithic technology, 9th-8th Schrifenreihe des Neanderthal S. Veil, N. Klasen & E.M. Theunissen, A.L. van Gijn millennium BC, Norwegian Museums 6), 101-115. H. Brückner 2009: The Early & D.C.M. Raemaekers 2012: Archaeological Review 46, 19-56. Mesolithic Haverbeck site, Synthesis – Keinsmerbrug: Tixier, J. 1984: Le débitage par northwest Germany: evidence a kaleidoscope of gathering, in: Spikins, P., H. Ayestaran & pression, in: J. Tixier (ed.), for Preboreal setlement in the B.I. Smit, O. Brinkkemper, C. Conneller 1995: ‘Meaningful Préhistoire de la pierre taillée, 2: western and central European J.P. Kleijne, R.C.G.M. Lauwerier paterns or mere pictures’: the economie du débitage laminaire, plain, Journal of Archaeological & E.M. Theunissen (eds.), infuence of collection strategy Paris, 57-70. Science 36, 1466-1476. A kaleidoscope of gathering at on the interpretation of lithic Keinsmerbrug (the Netherlands): scaters, in: A.J. Schofeld (ed.), 242 —

Toussaint, M., L. Brou, Van der Plicht, J., Van Gijn, A.L. 2006: Ornaments Van Gijn, A.L., V. van Betuw, F. Brun-Ricalens & F. Spier 2009: L.W.S.W. Amkreutz, of jet, amber and bone, in: A. Verbaas & K. Wentink 2006: The Mesolithic Site of Hefngen- M.J.L.Th. Niekus, L.P. Louwe Kooijmans & Flint, procurement and use, in: Loschbour (Grand Duchy of J.H.M. Peeters & B.I. Smit P.F.B. Jongste (eds.), Schipluiden: L.P. Louwe Kooijmans & Luxembourg). A yet undescribed 2016: Surf’n turf in Doggerland: a Neolithic setlement on the Dutch P.F.B. Jongste (eds.), Schipluiden: Human Cremation possibly from dating, stable isotopes and diet North Sea coast c. 3500 cal BC, a Neolithic setlement on the Dutch the Rhin-Meuse-Schelde of Mesolithic human remains Leiden (Analecta Prehistorica North Sea coast c. 3500 cal BC, Culture: Anthropological, from the southern North Sea, Leidensia 37/38), 195-206. Leiden (Analecta Praehistorica Radiometric and Archaeological Journal of Archaeological Science Leidensia 37/38), 129-166. Implications, in: P. Crombé & Reports 10, 110-118. Van Gijn, A.L. 2008: M. van Strydonck (eds.), De ornamenten van Ypenburg, Van Gijn, A.L. & Chronology and evolution in the Van der Reijden, H., G. Keers in: H. Koot, L. Bruning & D.C.M. Raemaekers 2014: Mesolithic of N(W) Europe, & H. van Rossum 2011: Ruimte R.A. Houkes (eds.), Ypenburg- Choosy about stone: the Cambridge, 239-260. voor archeologie: synthese van locatie 4: een nederzeting met signifcance of the colour red in themaveldrapportages, grafveld uit het midden-neolithicum the Dutch Funnel Beaker Culture, Van Beek, R., M.T.I.J. Gouw- Amsterdam (RIGO Research en in het West-Nederlandse Siedlungs- und Küstenforschung im Bouman, J.A.A. Bos & Advies rapport P18090). kustgebied, Leiden, 277-288. südlichen Nordseegebiet 37, 195-202. M.H. Kriek 2017: A glimpse into the past: mapping regional Van de Velde, P. 2014: Van Gijn, A.L. 2010: Flint in focus: Van Haaf, G., M.J.A. de Haan, vegetation developments since Aardewerktradities per lithic biographies in the Neolithic J.W.H. Hogestijn & the Late Glacial in Twente vindplaats en regio, in: and Bronze Age, Leiden. W.J. van Tent 1988: Baarn, (the Netherlands), in: I.M. van Wijk, L. Amkreutz & gem. Baarn, Jaarverslag R.C.G.M. Lauwerier, P. van de Velde (eds.), ‘Vergeten’ Van Gijn, A.L., V. Beugnier & Rijksdienst voor het Oudheidkundig M.C. Eerden, B.J. Groenewoudt, bandkeramiek: een odyssee naar de Y. Lammers-Keijsers 2001: Bodemonderzoek 1988, 60-61. M.A. Lascaris, E. Rensink, oudste neolithische bewoning in Vuursteen, in: L.P. Louwe B.I. Smit, B.P. Speleers & Nederland, Leiden, 498-505. Kooijmans (ed.), Archeologie in de Van Haaster, H. 2011: J. van Doesburg (eds.), Betuweroute: Hardinxveld- Palynologisch onderzoek, in: Knowledge for informed choices, Van der Velde, H. & N. Bouma Giessendam Polderweg: een mesoli- E. Lohof, T. Hamburg & Amersfoort (Nederlandse 2016: Oosterdalfsen: een rijk thisch jachtkamp in het rivierengebied J. Flamman (eds.), Steentijd Archeologische Rapporten 55), grafveld van de trechterbeker- (5500-5000 v. C.), Amersfoort opgespoord: archeologisch 64-72. cultuur, in: L. Amkreutz, (Rapportage Archeologische onderzoek in het tracé van de F. Brounen, J. Deeben, Monumentenzorg 83), 119-161. Hanzelijn-Oude Land, Amersfoort/ Van den Broeke, P., R. Machiels, M.-F. van Oorsouw Leiden (ADC report 2576/Archol H. Fokkens & A.L. van Gijn & B. Smit (eds.), Vuursteen Van Gijn, A. & R. Houkes 2006: report 138), 423-464. 2005: A prehistory of our time, verzameld: over het zoeken en Stone, procurement and use, in: L.P. Louwe Kooijmans, onderzoeken van steentijdvondsten in: L.P. Louwe Kooijmans & Van Hoof, L.G.L. 2009: P.W. Van den Broeke, en -vindplaatsen, Amersfoort P.F.B. Jongste (eds.), Schipluiden: Resultaten: sporen en H. Fokkens & A.L. van Gijn (Nederlandse Archeologische a Neolithic setlement on the Dutch structuren, in: T.A. Goossens (eds.), The prehistory of the Rapporten 50), 359-362. North Sea coast c. 3500 cal. BC, (ed.), Opgraving Hellevoetsluis- Netherlands, Amsterdam, 17-32. Leiden (Analecta Praehistorica Ossenhoek: een nederzeting van de Van Dijk, J. 2009: Zoogdieren, Leidensia 37/38), 167-193. Vlaardingen-groep op een Van der Leeuw, S. 2016: vogels en reptielen, in: kwelderrug in de gemeente Uncertainties, in: M. Brouwer T.A. Goossens (ed.), Opgraving Van Gijn, A.L., Y. Lammers- Hellevoetsluis, Leiden (Archol Burg, H. Peeters & W. Lovis Hellevoetsluis-Ossenhoek: een Keijsers & R. Houkes 2001: report 87), 51-69. (eds.), Uncertainty and sensitivity nederzeting van de Vlaardingen- Vuursteen, in: L.P. Louwe analysis in archaeological groep op een kwelderrug in de Kooijmans (ed.), Archeologie in de Van Hoof, L.G.L., I.M. Van Wijk computational modeling, gemeente Hellevoetsluis, Leiden Betuweroute: Hardinxveld-De Bruin: & C.M. van der Linde 2013: New York (Interdisciplinary (Archol report 87), 113-144. een kampplaats uit het laat- Zwervende erven op de löss? Contributions to Archaeology), mesolithicum en het begin van de Onderzoek van een nederzeting uit 157-169. Swiferbant-cultuur de vroege ijzertijd en van sporen uit (5500-4450 v. C.), Amersfoort de Stein-groep te Hof van Limburg (Rapportage Archeologische (gemeente Sitard-Geleen), Leiden Monumentenzorg 88), 153-191. (Archol report 33). 243 —

Van Leusen, M. & Van Zijverden, W. 2011: Verhagen, P., E. Rensink, Verlinde, A.D. 1974: A Mesolithic H. Kamermans (eds.) 2005: Landschap, in: E. Lohof, M. Bats & P. Crombé 2013: setlement with cremation at Predictive modelling for T. Hamburg & J. Flamman Establishing discovery Dalfsen, Berichten van de archaeological heritage (eds.), Steentijd opgespoord: probabilities of lithic artefacts Rijksdienst voor het Oudheidkundig management: a research agenda, archeologisch onderzoek in het in Palaeolithic and Mesolithic Bodemonderzoek 24, 113-117. Amersfoort (Nederlandse tracé van de Hanzelijn-Oude Land, sites with core sampling, Archeologische Rapporten 29). Amersfoort/Leiden (ADC report Journal of Archaeological Science Verlinde, A.D. & R.R. Newell 2576/Archol report 138), 37-77. 40, 240-247. 2006: A multi-component Van Noten, F.L. & D. Cahen complex of Mesolithic 1978: Les chasseurs de Meer, Veerman, J.P. 1971: Verhagen, P. & T.G. Whitley setlements with Late Mesolithic Bruggen. Archeologische vondsten te 2012: Integrating grave pits at Mariënberg in Cadzand-Bad, Westerheem archaeological theory and Overijssel, in: B.J. Groenewoudt, Van Oorsouw, M.-F., 1993: 20(2), 111-127. predictive modelling: a live R.M. van Heeringen & Wommersom revisited: een analyse report from the scene, Journal G.H. Scheepstra (eds.), van de verspreiding en het gebruik Verbaas, A., M.J.L.Th. Niekus, of Archaeological Method and Het zandeilandenrijk van Overijssel: van Wommersomkwartsiet in A.L. van Gijn, S. Knippenberg, Theory 19(1), 49-100. bundel verschenen ter gelegenheid Nederland, Leiden (unpublished Y.L. Lammers-Keijsers & van de pensionering van A.D. Verlinde MA thesis Leiden University). P.C. van Woerdekom 2011: Verhart, L.B.M. 1992: Setling als archeoloog in, voor en van Vuursteen uit de Hanzelijn, in: or trekking? The Late Neolithic Overijssel, Amersfoort Van Rijn, P. & L. Kooistra 2001: E. Lohof, T. Hamburg & house plans of Haamstede- (Nederlandse Archeologische Deel 15: Hout en houtskool: het J. Flamman (eds.), Steentijd Brabers and their counterparts, Rapporten 22), 83-270. gebruik van hout als opgespoord: archeologisch Oudheidkundige Mededelingen uit constructiemateriaal en onderzoek in het tracé van de het Rijksmuseum van Oudheden Verlinde, A.D. & R.R. Newell brandstof, in: J.W.H. Hogestijn Hanzelijn-Oude Land, Leiden/ 72, 73-99. 2013: The Mesolithic cemetery at & J.H.M. Peeters (eds.), De Amersfoort (Archol report 138/ Mariënberg (NL), a rebutal to mesolithische en vroeg-neolithische ADC report 2576), 335-393. Verhart, L.B.M. 2000: alternative interpretations, vindplaats Hoge Vaart-A27 Times fade away: the neolithization Amersfoort (Nederlandse (Flevoland), Amersfoort Verhagen, P. 2005: Prospection of the southern Netherlands in an Archeologische Rapporten 42). (Rapportages Archeologische strategies and archaeological anthropological and geographical Monumentenzorg 79). predictive modelling, in: M. van perspective, Leiden (PhD Verneau, S. & H. Peeters 2001: Leusen & H. Kamermans (eds.), dissertation Leiden University). Het topje van de ijsberg: een Van Wijk, I.M. (ed.) 2016: Predictive modelling for klein mesolithisch jachtkamp in Setlement dynamics on the archaeological heritage Verhart, L. 2008: de Ooyerhoek te Zutphen, Cannerberg (Maastricht, the management: a research agenda, New developments in the Archeologie, 10, 20-41. Netherlands): archaeological Amersfoort (Nederlandse study of the Mesolithic of the research of Bandkeramik and Iron Archeologische Rapporten 29), Low Countries, in: G. Bailey & Verneau-Peeters, S.M.J.P. & Age setlements, Leiden (Archol 109-121. P. Spikins (eds.), Mesolithic H.A.C. Fermin 2007: Zutphen- report 300). Europe, Cambridge, 158-181. Looërenk (werkput 167): een Verhagen, P. 2007: Case studies in vuursteenconcentratie uit het Van Wijk, I.M. & archaeological predictive modelling, Verhart, L. & H. Groenendijk mesolithicum, Deventer P. van de Velde 2016: Leiden (Archaeological Studies 2005: Living in abundance, (BAAC report 03.196). Setlement dynamics, in: Leiden University 17). Middle and Late Mesolithic, in: I.M. Van Wijk 2016 (ed.), L.P. Louwe Kooijmans, Visser, D., C. Whiton, Setlement dynamics on the Verhagen, J.W.H.P., E. Rensink, P.W. van den Broeke, O. Brinkkemper & Cannerberg (Maastricht, the M. Bats & P. Crombé 2011: H. Fokkens & A.L. van Gijn J.W.H. Hogestijn 2001: Netherlands): archaeological Optimale strategieën voor het (eds.), The prehistory of the Archeobotanie: de analyse van research of Bandkeramik and Iron opsporen van steentijdvindplaatsen Netherlands, Amsterdam, botanische macroresten, in: Age setlements, Leiden (Archol met behulp van booronderzoek: een 161-178. J.W.H. Hogestijn & J.H.M. Peeters report 300), 130-139. statistisch perspectief, Amersfoort (eds.), De mesolithische en vroeg- (Rapportage Archeologische neolithische vindplaats Hoge Monumentenzorg 197). Vaart-A27 (Flevoland), Amersfoort (Rapportages Archeologische Monumentenzorg 79). 244 —

Vos, P.C. 2015: Origin of the Dutch Wansleeben, M. & W. Laan radiocarbon evidence, in: Whitley, T.G. 2016: coastal landscape: long-term 2012: Ruimtelijke analyse, in: R. Fellmann, G. Germann & Archaeological simulation and landscape evolution of the T. Hamburg, A. Müller & K. Zimmermann (eds.), Jagen the testing paradigm, in: Netherlands during the Holocene, B. Quadfieg (eds.), Mesolithisch und Sammeln: Festschrif für M. Brouwer Burg, H. Peeters & described and visualized in national, Swiferbant: mesolithisch gebruik Hans-Georg Bandi zum 65. W. Lovis (eds.), Uncertainty and regional and local palaeogeo- van een duin ten zuiden van Geburtstag, Bern (Jahrbuch des sensitivity analysis in archaeological graphical map series, Groningen. Swiferbant (8300-5000 v.C.): een Bernische Historische Museum computational modeling, archeologische opgraving in het 63/64), 273-281. New York (Interdisciplinary Vos, P.C., & K.M. Cohen 2015: tracé van de N23/N307, provincie Contributions to Archaeology), Landscape genesis and Flevoland, Leiden/Amersfoort Waterbolk, H.T. 1999: De mens 131-156. palaeogeography, in: J.J. Moree (Archol report 174/ADC report in het preboreale, boreale en & M.M. Sier (eds.), Interdisciplinary 3250), 85-112. atlantische bos, Paleo-Aktueel Wiessner, P. 1983: Style and Archaeological Research Programme 10, 68-73. social information in Kalahari Maasvlakte 2, Roterdam, 1: twenty Wansleeben, M., W. Laan & San projectile points, American meters deep! The Mesolithic period at S. Knippenberg 2011: Waterbolk, H.T. 2009: Antiquity 48, 253-276. the Yangtze Harbour site – Roterdam Ruimtelijke analyse, in: Getimmerd verleden: sporen van Maasvlakte, the Netherlands: Early E. Lohof, T. Hamburg & voor- en vroeghistorische houtbouw Willems, W.J.H. & R.W. Brandt Holocene landscape development and J. Flamman (eds.), Steentijd op de zand- en kleigronden tussen 2004: Dutch archaeology quality habitation, Roterdam opgespoord: archeologisch Eems en IJssel, Groningen standard, Den Haag. (BOORrapporten 566), 63-146. onderzoek in het tracé van de (Groningen Archaeological Hanzelijn-Oude Land, Studies 8). Woldring, H., M. Schepers, Vroomans, M.A.K., Amersfoort/Leiden (ADC report J. Mendelts & R. Fens 2012: L.A. Tebbens & M. Tump 2014: 2576/Archol report 138), 79-114. Weber, M.-J. 2012: Camping and foraging in Boreal Resultaten van de opgraving, From technology to tradition – hazel woodland – the in: M. Tump, M.A.K. Vroomans, Wansleeben, M., & L.P. Louwe re-evaluating the Hamburgian- environmental impact of L.A. Tebbens & C. Kasse (eds.), Kooijmans 2006: The Magdalenian relationship, Mesolithic hunter-gatherers Een Ahrensburg-site uit de eerste archaeological remains: Neumunster (Untersuchungen near Groningen, in: helf van de late Dryas langs de A2 a critical spatial approach, in: und Materialien zur Steinzeit in M.J.L.Th. Niekus, R.N.E. Barton, bij Geldrop (gemeente Heeze en L.P. Louwe Kooijmans & Schleswig-Holstein und im M. Street & T. Terberger (eds.), Leende): A2 Aalsterhut, Den Bosch P.F.B. Jongste (eds.), Schipluiden: Ostseeraum 5). A mind set on fint: studies in (BAAC report A-08.0480/A- a Neolithic setlement on the Dutch honour of Dick Stapert, Groningen 09.0116/A-09.0210/A-09.0386), North Sea coast c. 3500 cal BC, Weiner, J. 2005: Another word (Groningen Archaeological 93-138. Leiden (Analecta Praehistorica on pitch: some comments on a Studies 16), 381-392. Leidensia 37/38), 67-87. ‘sticky issue’ from old Europe, Waddington, C. (ed.) 2007: Bulletin of Primitive Technology 29, Woodman, P.C. 1985: Mesolithic setlement in the North Waterbolk, H.T. 1960: 20-27. Excavations at Mount Sandel Sea basin: a case study from Howick, Preliminary report on the 1973-1977, Belfast (Northern north-east England, Oxford. excavations at Anlo in 1957 and Wentink, K. 2008: Crafing Ireland Archaeological 1958, Palaeohistoria 8, 59-90. axes, producing meaning: Monographs 2). Waddington, C. & C. Bonsall Neolithic axe depositions in the 2015: Archaeology and Waterbolk, H.T. 1983: northern Netherlands, Zeiler, J.T. 2006: Mammals, in: environment on the North Sea Ten guidelines for the Archaeological Dialogues, 15(2), L.P. Louwe Kooijmans & litoral: a case study from Low archaeological interpretation 151-173. P.F.B. Jongste (eds.), Schipluiden: Hauxley, Bakewell. of radiocarbon dates, in: a Neolithic setlement on the Dutch W.G. Mook & H.T. Waterbolk Whitley, T.G. 2000: Dynamical North Sea coast c. 3500 cal BC, Walczak, J. 1998: La question (eds.), Proceedings of the frst systems modelling in archaeology: Leiden (Analecta Praehistorica des styles techniques durant le International Symposium 14C and a GIS approach to site selection Leidensia 37/38), 375-420. Mésolithique: remarques Archaeology, Groningen, 57-70. processes in the greater Yellowstone générales sur le style region, Pitsburgh (unpublished tardenoisien de Coincy et sur sa Waterbolk, H.T. 1985: The PhD dissertation University of “valeur humaine”, Bulletin de la Mesolithic and Early Neolithic Pitsburgh). Société préhistorique française, setlement of the northern 95-2, 203-220. Netherlands in the light of 245 —

Zeiler, J.T. & D.C. Brinkhuizen 2015: Fauna, in: J.M. Moree & M.M. Sier (eds.), Interdisciplinary Archaeological Research Programme Maasvlakte 2, Roterdam, 1: twenty meters deep! The Mesolithic period at the Yangtze Harbour site – Roterdam Maasvlakte, the Netherlands: Early Holocene landscape development and habitation, Roterdam (BOORrapporten 566), 201-221.

Zvelebil, M. 1994: Plant use in the Mesolithic and its role in the transition to farming, Proceedings of the Prehistoric Society 60, 3574.

Zvelebil, M. & J. Moore 2006: Assessment and representation: the information value of Mesolithic landscapes, in: E. Rensink & H. Peeters (eds.), Preserving the early past: investigation, selection and preservation of Palaeolithic and Mesolithic sites and landscapes, Amersfoort (Nederlandse Archeologische Rapporten 31), 151-165.

247 Appendices —

I Overview of the relationship between research questions, overarching themes, analytical dimensions, and NOaA themes II List of reports 248 — Appendix I Overview of the relationship between research questions, overarching themes, analytical dimensions, and NOaA themes

Analytical themes Spatio-temporal NOaA 1.0 themes dimensions

QID Question P&M P&S P&E L R S-T L-T N1 N2 N3 N4 N5 Method

1 What is the nature and use of fint, and what is the origin of +++ + the raw materials?

2 What is the quality of fint production and is there a +++ + relationship between quality of fint and length and breadth of the end products?

3 Is any worked stone found? What was the function of such +++ + artefacts, what is the origin of the raw material?

4 What is the evolution of form and decoration in ceramics? Is +++ + there evidence for gradual regional diferences?

5 What is the nature and origin of resoures used in ceramics +++ + production?

6 What is the origin, use and archaeological context of jewellry + + + + +

7 Can typological artefacts be seen as culturally specifc +++ indicators?

8 Is there evidence for typological change or developments +++++ + within a culture? Is the current typochronology adequate for dating purposes, if not, is it possible to refne these typochronological with the new material evidence and dating?

9 Does the chronological and spatial variation in the use of + + ++ ++ certain stone resources (e.g. Wommersomquarzite) for certain arrowhead types relate to more than just diferences in cultural tradition?

10 What practical uses of organic material other than food have +++++++ been recognized and which traces of the associated crafs have been identifed? (Evidence for rope, dye, woodworking, tar used for hafing?)

11 What is the function of sites and to what periods/cultures do + + + ++++ the phases of habitation date? Is there evidence for continuity or dinscontintuity?

12 What is the function of hearth-pits (any evidence related to + + + ++++ ++ food preparation and cooking?) and how can their large numbers and long history of activity at one location be explained (int he mesolithic)?

13 What can be said on the division between the North and + +++++ South of the Netherlands in relation to arrowhead styles and stone resource use during the Late Palaeolithic and Mesolithic, how does the dividing line change int time and how does this realte to a changing landscape?

14 Does the dispersal of deep hearthpits show a similar spatial +++++ ++ division as resource and arrowhead use?

15 What can be said on the colonisation and dispersion of +++++ cultures and traditions in relation to environmental changes? Are changes in material culture related to these processes?

16 Did research on organics yield any signs of natural or cultural + ++++ ++ + disasters or changes, and if this is the case, are there any indications on how people dealt with them?

17 What is the importance (shares of the food economy) of + + ++++ + hunting, fshing, acquatic resources and plant foods in hunter-gatherer societies versus neolithic societies?

18 Which traces of food preparation, processing and storage +++++++ have been identifed?

19 What can be said about the origin of food sources, the balance between local production and import, including trade, and to what extent did the availability and seasonality of food sources infuence human mobility and setlement location choices?

20 Is there evidence for products that were especially important +++++ in the food supply? What was their signifcance (eg prestigious products)? 249 —

Analytical themes Spatio-temporal NOaA 1.0 themes dimensions

QID Question P&M P&S P&E L R S-T L-T N1 N2 N3 N4 N5 Method

21 Are there human remains, what is the character of the ++++ ++ remains (i.e. sex, age, health, demographic information) and the deposition (i.e inhumation, cremation, disarticulation, cultic or religious character)?

22 Have any grave goods been found, what is their character + + ++++ +++ (i.e exotic or local grave goods? Common or elite? Inidividualistic or defned set?)

23 What role did organic materials (plants and animals) and the + + ++++ +++ landscape play in (religious) rituals and burial customs?

24 What are the spatial trends of mortuary tradition (relation to + + ++++ + ++ setlement, position of the body, consistency of the grave structure)?

25 Do activity areas or features indicate the presence of ++ + + + structures?

26 What is the signifcance of empty and flled spaces + + + + + +

27 Is there evidence for social use of space? + + + + +

28 What activity areas and phases of habitation can be ++++++ + discerned in palimpsest sites? Are there specifc zones of artefacts, features and structures?

29 What impact does site formation processes and post ++ + + depositional processes have on the archaeological record and subsequent archaeological interpretation? (incl absence of features in concentrations, erosion, relation between surface fnds and underlying features)

30 What (spatial) variation exists between sites + + + + + +

31 Does the zoological or physical anthropological evidence + + + + +++ indicate social or economical diferences between or within sites?

32 Is there evidence for hierarchic (or some other) relations ++++++ + + between or within setlements based on activity within the setlements?

33 Is data collected in a suitable manner for spatial research + + + +

34 Is additional spatial data available online (DANS)? + +

P&M People & Material

P&S People & Space

P&E People & Environment

L Local

R Regional

S-T Short-Term

L-T Long-Term

N1 Colonisation and earliest occupation history of the Netherlands

N2 Landscape use and setlement systems

N3 Food economy, relationship Man and environment

N4 Burial and deposition of human remains

N5 Cultural traditions, social relationships and interaction 250 — Appendix II List of reports

The following list of publications contains all 61: Roessingh,W., 2008: Graven op een zandkop. reports that were selected for analysis. The Een archeologische opgraving langs de Geranium analysis started with a long list of reports on te Bergerden, gemeente Lingewaard, feldwork of potential interest for the purpose of Amersfoort (ADC-rapport 837). this study. These were numbered and added to 65: Weten H.J.W.C. van 2009: Herinrichtingsplan the research database. In the further assessment Keersop te Bergeijk. Een Archeologische Beekdal a substantial number of reports were discarded Begeleiding, Doetinchem (Synthegra- (see section 2.3.1). Hence, de sequence of report rapport S083301). number (PUBID) is discontinuous. 71: Hermsen, I., 2009: Vissen in het verleden van Bergharen, Nijmegen (Archeologische 1: Muller, A., T. Obdam, M.T.I.J. Bouman & Berichten Wijchen 7). J.A.A. Bos 2012: A2 Viaduct Aalsterhut fase 2 74: Groot, R.W. de, 2013: Agrarisch en ritueel Aalst, viaduct Aalsterhut (Gemeente Waalre), landgebruik op de vindplaats Beuningen- Amersfoort (ADC-rapport 3181). Hogewaldstraat, Weesp (RAAP-rapport 7: Bouma, N. & A. Müller 2014: Tienduizend jaar 2349). landschaps- en bewoningsgeschiedenis in het 80: Steeg, J.F. van de & J.M.J. van de Steeg- Maasdal tussen Well en Aijen, Amersfoort Oorthuis 2007: Drie opmerkelijke vondsten (ADC-rapport 3472). uit Het Hambroek, Borculo, Cranium 24, 9: Hielkema, J.B., 2014: Sporen uit de Prehistorie 25-30. op de vindplaats Almen-Groot Besselink, Weesp 86: Linden, B.A. van der, J.S. Krist & (RAAP-rapport 2333). P. Wemerman 2009: Archeologische 10: Scholte Lubberink, H.B.G., 2014: Begeleiding conform KNA protocol Opgraven, Nederzetingssporen uit de Prehistorie en Perceel 85 en 86, plangebied Trierveld te Born, Romeinse tijd op de Whemerenk te Almen, Doetinchem (Synthegra-rapport S090413). Weesp (RAAP-rapport 2332). 87: Linden, B.A. van der, 2010: Klein Trierveld te 15: Geer, P. van de, 2013: Steentijd op de Born, percelen 131, 132, 133, 134, 135, 172. Stichtsekant. Defnitieve opgraving van drie Een Archeologische Begeleiding, Zelhem vindplaatsen op bedrijventerrein Stichtsekant, (Synthegra-rapport P0502090). gemeente Almere, Leiden (ARCHOL-rapport 88: Hensen, G. 2006: Archeologische Begeleiding 212). (AB), Klein Trierveld te Born, Zelhem 16: Hamburg, T., A. Tol, J. de Moor & (Synthegra-briefrapport 176055). Y. Lammers-Keijsers 2014: Afgedekt Verleden, 94: Hoegen, R.D., H.B.G. Scholte Lubberink, Leiden (ARCHOL-rapport 244). E. Verhelst & N.W. Willemse 2008: Tussen 24: Van der Kuijl, E.E.A. 2006: Archeologische bekken en stuwwal: Plangebied opgraving, Julianastraat 27 te Alphen aan den Parachutistenstraat te Breedeweg; gemeente Rijn, Zelhem (Synthegra-rapport 174137). Groesbeek: een archeologische opgraving, 43: Zielman, G., 2013: Van oale groond. Nieuwe Weesp (RAAP-rapport 1744). vondsten op de Azeler Esch. 95: Linde, C.M. van der, 2013: Prehistorische Aardgastransportleidingtracé Bornerbroek-Epe resten en een middeleeuwse landweer op (A670), catalogusnummer 1. Gemeente Hof van bedrijventerrein Rode Beek, Leiden (ARCHOL- Twente, Weesp (RAAP-rapport 2395). rapport 211). 49: Weerden, J.F. van der, 2012: Barneveld 98: Hensen, G., M. Janssens, H. Leuvering & Harselaar West-west Archeologisch M. Kappers 2006: Opgraving, DO Berikstraat onderzoek, ‘s Hertogenbosch/Deventer te Buggenum, Zelhem (Synthegra-rapport (BAAC-rapport A-09.0252). 175222). 51: Zee, K., 2009: Archeologische onderzoek aan de 103: Weerden, J.F. van der, A.C. van de Venne & Waterstraat in Beek - Gem. Ubbergen, P.Dijkstra 2009: Buurse, plangebied Buurse- Nijmegen (Archeologische Berichten Zuid; Archeologisch onderzoek, Ubbergen 3). ‘s Hertogenbosch/Deventer (BAAC-rapport 53: Lohof, E. & S. Wyns 2009: Beek Kerkeveld, de A-09.0040). periferie van een Bandkeramische nederzeting 104: Veken, B. van der & A. Müller 2013: Venray, Een Defnitief Archeologisch Onderzoek, Castenray, De Diepeling. Een archeologische Amersfoort (ADC-rapport 1292). opgraving, Amersfoort (ADC-rapport 3414). 251 —

105: Moerman, S., 2012: Uitwerking van een 134: Mitendorf, E.S., 2013: Archeologische archeologische begeleiding, eindrapport, begeleiding aanleg nutsleidingen Natuurderij, Wagenbroeksloopje, Casteren gemeente Bladel, gemeente Deventer (project 454), Deventer Noordwijk (IDDS-Archeologierapport 1417). (Interne rapportages archeologie Deventer 106: Wal, A. ter, 2006: Casteren De Vloed. 59). Defnitief Archeologisch Onderzoek, 136: Eimermann, E., 2012: Opgraving keerwand ‘s Hertogenbosch/Deventer (BAAC-rapport grondwal Olthof-Zuid Bedrijvenpark A1 (project 05.286). 450), Deventer (Interne rapportages 109: Beek, J.L. van, 2006: Plangebied ‘t klooster te Archeologie Deventer 58). Coevorden, Gemeente Coevorden, Een 141: Verhoeven, M., 2013: Plangebied Lange archeologische begeleiding, Weesp (RAAP- Akkersweg-Doorderweg bij Dieteren. Gemeente rapport 1443). Echt-Susteren. Archeologisch vooronderzoek: een 113: Heirbaut, E.N.A. 2005: Bewoning van archeologische begeleiding, Weesp (RAAP- prehistorie tot middeleeuwen in het rapport 2624). buitengebied van Cuijk, Leiden (ARCHOL- 142: Hoof, B.I. van, 2011: Plangebied uitbreiding rapport 34). begraafplaats Diever, gemeente Westerveld; 120: Mijle Meijer, R.A. van der, E.E.B. Bulten, archeologisch onderzoek: een archeologische P.J.A. Stokkel & E.C. Riefe 2010: begeleiding, Weesp (RAAP-rapport 2419). Westvlietweg, Gemeente Den Haag, onderzoek 143: Kruining, M.E. van & J.B. Hielkema 2009: naar bewoningssporen uit het midden- en laat- Uitbreiding begraafplaats Diever, gemeente neolithicum op een duintje achter de oudste Westerveld; archeologisch voor onder zoek: een strandwal, Den Haag (Rapport Afdeling inventariserend veldonderzoek proefsleuven, Archeologie Dienst Stadsbeheer Den Haag Weesp (RAAP-rapport 1991). 0919). 146: Oosterhout, F. van, 2014: Plangebied 122: Stokkel, P., 2013: Steynhof, gemeente Holterhoek 2, deelgebied Fatima (weide Den Haag. Inventariserend veldonderzoek- Annahoeve), gemeente Doetinchem. proefsleuven en archeologische begeleiding. Archeologisch onderzoek: een opgraving, Weesp Vondsten en sporen uit het neolithicum, (RAAP-rapport 2697). Den Haag (Haagse Archeologische 147: Pronk, E.C., 2011: Plangebied Veemarkterrein, Rapportage 1225). gemeente Doetinchem. Een opgraving met resten 124: Zoolingen, R.J. van, 2014: uit het Meso- en Neolithicum, de Brons- en Zevenwoudenlaan, gemeente Den Haag. IJzertijd en de Vroege en Late Middeleeuwen, Archeologisch onderzoek naar sporen van archeologisch onderzoek: een opgraving, Weesp bewoning uit de prehistorie, Romeinse tijd (RAAP-rapport 2217). en middeleeuwen, Den Haag (Rapport 148: Diependaal, S., S.M. Koeman, Afdeling Archeologie Dienst Stadsbeheer P. Wemerman & J.S. Krist, 2012: Defnitieve Den Haag 1404). Opgraving Holterhoek te Doetinchem, Gemeente 126: Ridder, J.A.A. de, 2013: Informatie over het Doetinchem. Sporen van menselijke verleden onder het informatiecentrum. Een aanwezigheid in een rivierduinlandschap langs de archeologische opgraving in Den Nul, gemeente Oude IJssel van het vroeg mesolithicum tot Olst-Wijhe, Amersfoort (ADC-rapport 2487). nieuwe tijd, Doetinchem (Synthegra-rapport 129: Hiddink, H.A., 2008: Archeologisch onderzoek S100306). op de Groot Botelsche Akker bij Deurne. 161: Prangsma, N.M. & D.A. Gerrets, 2009: Bewoning uit de Steentijd, IJzertijd, Romeinse Hanzelijn Tunnel Drontermeer: verbinding tussen tijd, Vroege en Volle Middeleeuwen, Amsterdam Oude en Nieuwe Land. Een archeologische (Zuidnederlandse Archeologische begeleiding bij de Sallanddijk en een Rapporten 33). compenserend archeologisch onderzoek in gebied 130: Norde, E.H.L.D., 2013: Aan de rand van de XVI, Amersfoort (ADC-rapport 1601). Deurninger Beek. Gemeente Dinkelland. 162: Veldman, H.A.P., 2006: Duiven de Eng Archeologisch onderzoek: een opgraving op de ‘Graafstaete’ vindplaats 1, een archeologische Deurninger Es, Weesp (RAAP-rapport 2759). opgraving, Amersfoort (ADC-rapport 537). 131: Hermsen, I., 2008: Deventer, VV IJsselstreek, Deventer (Overijssels Erfgoed 2007). 252 —

163: Kooi, M., 2014: Duiven, Ploen Zuid. 211b: Schrijer, E, & W. de Neef, 2007: Fluitenberg, Vindplaatsen uit het mesolithicum en sporen uit Begraafplaats Zevenberg (Dr.) Steentijd en de ijzertijd - Romeinse tijd en late middeleeuwen IJzertijdbewoning in Fluitenberg. Deel 2: - nieuwe tijd, ’s-Hertogenbosch (BAAC- Opgraving en Archeologische begeleiding, rapport A-10.0287). Zuidhorn (Steekproefrapport 2007-02/02). 167: Keijers, D.M.G., 2006: Plangebied 216: Tump, M., M.A.K. Vroomans, L.A. Tebbens & Bisschoppelijk College Echt, Gemeente Echt- C. Kasse, 2014: Een Ahrensburg-site uit de eerste Susteren. Een archeologische begeleiding, Weesp helf van de late Dryas langs de A2 bij Geldrop (RAAP-rapport 1392). (gemeente Heeze en Leende) A2 Aalsterhut. 170: Brouwer, M.C., 2010: Ede Uitvindersbuurt fase ’s-Hertogenbosch (BAAC-rapport A08.0480 1&2. Defnitief archeologisch onderzoek, / A-09.0116 / A-09.0210 / A-09.0386). ’s-Hertogenbosch (BAAC-rapport 219: Hoof, L.G.L. van, I.M. Van Wijk & C.M. van A-07.0136). der Linde, 2013: Zwervende erven op de löss? 174: Roessingh, W., 2008: Haardkuilen uit het Onderzoek van een nederzeting uit de vroege Mesolithicum en een erf uit de Romeinse tijd aan ijzertijd en van sporen uit de Stein-groep te Hof de Pascalstraat in Ede. Een archeologische van Limburg (gemeente Sitard-Geleen), Leiden Opgraving, Amersfoort (ADC-rapport 1172). (ARCHOL-rapport 33). 175: Wal, A. Ter & N. Wite, 2013: Ede 224: Mooren, J.R., 2011: Het Genneperhuis: de Uitvindersbuurt fase III. Een opgraving van een vestinggrachten. Inventariserend veldonderzoek laat-Romeins grafveld en bewoningssporen uit door middel van proefsleuven, gevolgd door een de ijzertijd tot voege middeleeuwen, archeologische begeleiding, ’s-Hertogenbosch ’s-Hertogenbosch (BAAC-rapport A11.0017). (BAAC-rapport A-09.0235). 178: Tulp, C., 2014: Eelde, Groote Veen (Gemeente 227: Boon, H., 2012: ARC-heologisch onderzoek Tynaarlo, Dr.). Een defnitief archeologisch Torenveen. Archeologische begeleiding, onderzoek (DAO), Zuidhorn Groningen (Grontmij Archeologische (Steekproefrapport 2008-07/17). Rapporten 1181). 190: Debruyne, S. & J.M. Vroomans, 2012: 238: Wieringa, A.R., 2013: Een archeologische Archeologische opgravingen Eindhoven – Puten. opgraving op het terrein van de Coendershof aan Grootschalig archeologisch onderzoek in het de Helperbrink te Groningen (Gr.), Groningen zuidoosten van Eindhoven. Eindhoven (ARC--rapport 2013). (Eindhoven Archeologisch Centrum rapport 239: Wieringa, A.R., 2011: Sporen uit de ijzertijd 46). langs de Hunze. Een opgraving aan de Kielerbocht 194: Porreij, A., 2009: 100 bandkeramische huizen te Groningen. Groningen (Stadse Fratsen 29). in Elsloo. Opgravingsverslag van het onderzoek 240: Overeem, M.J., 2006: Trechterbekeraardewerk Joannes Riviusstraat fase 1 te Elsloo. Leiden uit een bouwput op een terrein van het (ARCHOL-rapport 132). Universitair Medisch Centrum Groningen. 206: Houkes, M.C.E. & H.M. van der Velde, 2008: Groningen (Stadse Fratsen 6). Enterse akkers onderzocht. Enter de Akkers fase 2, 248: Tuinstra, S.J., J.R. Veldhuis & J.A.W. Nicolay, Gemeente Wierden. Een Archeologische 2009: Hallum, een welvarend dorp aan de Opgraving, Amersfoort (ADC-rapport 993). monding van de Middelzee. Een archeologische 210: Ilson, P.J. & G. Aalbersberg, 2012: opgraving te Hallum, gemeente Ferwerderadeel Mesolithische en Middeleeuwse resten bij (Fr.), Groningen (ARC--Publicaties 205). Finsterwolde-Ganzedijk. Aardgastransport- 253: Knippenberg, S., 2013: Neolithicum en ijzertijd leidingtracé Oude Statenzijl-Midwolda (A-677) in de Maaskant. Opgravingen van een midden- KR-012, catalogusnummer 3, Gemeente neolithische nederzeting en een midden- en late Oldambt. Archeologisch onderzoek: proefsleuven, ijzertijd crematiegrafveld te Haren (N-Br.), Weesp (RAAP-rapport 2303). Leiden (ARCHOL-rapport 214). 211a: Schrijer, E., W. de Neef & A. Vissinga, 2007: 254: Ringenier, H., 2013: Archeologisch onderzoek Fluitenberg, Begraafplaats Zevenberg (Dr.) op locatie Harfsen-Nijland. Aardgastransport- Steentijd en IJzertijdbewoning in Fluitenberg. leidingtracé Ommen-Esveld (A662), Deel 1: Het inventariserende onderzoek, catalogusnummer 10, Gemeente Lochem. Zuidhorn (Steekproefrapport 2007-02/02). Archeologisch onderzoek: opgraving en begeleiding, Weesp (RAAP-rapport 2331). 253 —

255: Arnoldussen, S., J.P. Mendelts, R.L. Fens, & 330: Rondags, E.J.N., 2013: Een prehistorisch grafveld J.H.M. Peeters, 2012: Een Mesolithisch in Langenboom; aardgastransportleidingtracé kampement te Meerstad – vindplaats 2a, Ravenstein-Odiliapeel (A-665), catalogusnummer Groningen (Grondsporen 12). 15, gemeente Mill en Sint-Hubert; Archeologisch 258: Scholte Lubberink, H.B.G., 2010: Opgraving onderzoek: opgraving, Weesp (RAAP-rapport Hasselo-’t Oosterveld, Gemeente Hengelo. Een 2354). opgraving met archeologische resten uit het 332: Opbroek, M., F. Wijsenbeek & Meso- en Neolithicum, de Bronstijd en de S. Beckerman, 2008: Leek, Leeksterveld. Een Middeleeuwen, Weesp (RAAP-rapport 2191). Archeologische begeleiding, Amersfoort 259: Lohof, E., T. Hamburg & J. Flamman, 2011: (ADC-rapport 918). Steentijd opgespoord. Archeologisch onderzoek in 333: Vissinga, A., 2007: Leek, Mensumaweg (Gr.). het tracé van de Hanzelijn-Oude Land, Leiden/ Een waarderend archeologisch veldonderzoek, Amersfoort (ARCHOL-rapport 138 / Zuidhorn (Steekproefrapport 2007-02/19). ADC-rapport 2576). 337: Broeke, P.W. van den, J.A. den Braven & 260: Hamburg,T., E. Lohof & B. Quadfieg, 2011: E.A.G. Ball, 2010: Midden-neolithicum tot en Bronstijd opgespoord. Archeologisch onderzoek met de vroeg-Romeinse tijd in het Zuiderveld. van prehistorische vindplaatsen op Bedrijvenpark Onderzoek van nederzetingssporen en graven te H2O – plandeel Oldebroek (Provincie Nijmegen-Ressen, Nijmegen (Archeologische Gelderland), Leiden/Amersfoort (ARCHOL- Berichten Nijmegen 15). rapport 142 / ADC-rapport 2627). 338: Ball, E. & P. van den Broeke, 2007: 265: Ruijters, M.H.P.M., 2014: Prehistorische Opgravingen op ’t Klumke te Nijmegen- vondsten op de vindplaats Susteren- Oosterhout. Boeren uit het midden-neolithicum, Aardenweg. Aardgastransportleidingtracé de ijzertijd en de Merovingische periode op een Hommelhof-Schinnen (A665), zandrug in de Oostelijke Betuwe, Nijmegen catalogusnummer 18, Gemeente Echt- (Archeologische Berichten Nijmegen 6). Susteren. Archeologisch onderzoek: 339: Nes, E. de & C. Tulp, 2012 : Leunen, opgraving, Weesp (RAAP-rapport 2368). Horsterweg 11. Gemeente Venray (Limburg). 272: Goossens T.A., 2009: Opgraving Archeologische begeleiding volgens protocol Hellevoetsluis-Ossenhoek. Een nederzeting van Opgraven, Zuidhorn (Steekproefrapport de Vlaardingen-groep op een kwelderrug in de 2012-01/04U). gemeente Hellevoetsluis, Leiden (ARCHOL- 340: ’t Hart. B. & E.I. Schuurman, 2013: Sporen rapport 87). aan het Hernense Meer te Leur. 281: Flokstra, L.M. & S.W. Jager, 2012: Plangebied Aardgastransportleidingtracé Beuningen- Flerdeweg III te Hernen, Gemeente Wijchen. Ravenstein (A-664), catalogusnummer 1, Archeologisch onderzoek: een archeologische gemeente Wychen, Archeologisch onderzoek: begeleiding, Weesp (RAAP-rapport 2577). opgraving, Weesp (RAAP-rapport 2351). 306: Ruijters, M.H.P.M., 2011: Plangebied 351: Prangsma, N.M., 2008: Lomm Hoogwatergeul WBL-leiding Iteren-Borgharen, gemeente fase 1 (gemeente Arcen en Velden). Een Maastricht. Een archeologische begeleiding, archeologische opgraving, Amersfoort Weesp (RAAP-rapport 2448). (ADC-rapport 1344). 308: Linde, C.M. van der, E. Heunks, A. Porreij- 352: Gerrets, D.A. & R. de Leeuwe, 2011: Rituelen Lyklema & I.M. van Wijk, 2011: Defnitieve aan de Maas. Een archeologische opgraving te Opgraving Iteren - Hoeve Haertelstein Lomm, Hoogwatergeul fase II, Amersfoort (vindplaats 3), Leiden (ARCHOL-rapport 152). (ADC-rapport 2333). 309: Porreij-Lyklema & I.M. van Wijk, 2014: 360: Campenhout, K. Van, 2013: De carréhoeves Sporen van de Michelsberg-cultuur en een van Wolder, Maastricht. Een archeologische laatprehistorische nederzeting bij Voulwames, opgraving, Amersfoort (ADC-rapport 3375). gemeente Meerssen, Leiden (ARCHOL-rapport 365: Wemerman, P.J.L., 2010: Archeologische 240). begeleiding De Borkeld te Markelo, gemeente 328: Brouwer, E.W., 2007: Archeologische Hof van Twente. Sporen van prehistorische begeleiding fetspaden Langelo, Assen jagers-verzamelaars en boeren aan de voet van (Arcadis-rapport 110312/ de Friezenberg, Doetinchem (Synthegra- NA7/032/000256/001). rapport S090338). 254 —

373: Jezeer, W. & L.P. Verniers, 2012: De Plantage: 438: Tulp, C., 2009: Partij, Van Pallantstraat, een nieuwe wijk, een rijk verleden. Een gemeente Gulpen-Witem (L.). Een archeologische opgraving op De Plantage in archeologische begeleiding onder protocol Meteren (gemeente Geldermalsen), Amersfoort opgraven, Zuidhorn (Steekproefrapport (ADC-rapport 2713. 2009-05/09. 377: Boer, R. den, 2012: Laat-middeleeuwse sporen 450: Broeke, P.W. van den, 2013: Laat- en vondsten uit de Steentijd op de vindplaats prehistorische steenstructuren in Nijmegen- Midwolda-Kolkepad; Aardgastransport- Noord. Archeologisch onderzoek op De Boel, leidingtracé Oude Statenzijl-Midwolda (A-677), Nijmegen (Archeologische Berichten catalogusnummer 7; gemeente Oldambt; Nijmegen 34). Archeologisch onderzoek: veldinspectie, Weesp 457: Glind, M. van de, 2010: Roermond plangebied (RAAP-rapport 2304). Keulsebaan, deelgebied A. Defnitief onderzoek, 381: Leeuwe, R. de, 2010: Opgraving van een Den Bosch/Deventer (BAAC-rapport nederzeting uit de periode bronstijd-vroege A-09.0070). ijzertijd te Geldrop Luchen (gemeente Geldrop- 462: Moree, J.M. & M. Sier, 2014: Twintig meter Mierlo), Leiden (ARCHOL-rapport 133). diep! Mesolithicum in de Yangtzehaven- 385: Jong, T. de, 2008: Archeologisch onderzoek Maasvlakte te Roterdam. Milheeze-Zuidrand (gemeente Gemert-Bakel). Landschapsontwikkeling en bewoning in het Omheinde nederzeting aan een splitsing van Vroeg Holoceen, Roterdam (BOOR-rapport wegen. Rapportage vlakdekkend onderzoek, 523). Eindhoven (Archeologisch Centrum 463a: Zijl, W., M.J.L.Th. NIekus, P.H.J.I. Ploegaert Eindhoven rapport zonder nummer). & J.M. Moree, 2011: Roterdam Beverwaard 392: Ilson, P.J., 2012: aardgastransportleidingtracé Tramremise. De opgraving van de top van een Swiferbant-Emmeloord (A683 en N-500-64). donk met sporen uit het Mesolithicum en Gemeenten Dronten en Noordoostpolder. Neolithicum (vindplaats 13-83), Roterdam Archeologisch onderzoek: begeleidingen op de (BOOR-rapport 439). vindplaatsen 1 (Elandweg, Swiferbant), 2 463b: Lelivelt, R.A., 2006: Roterdam Beverwaard (Elandtocht, Swiferbant) en 8 (Joh. Postocht, Tramremise. Een bureauonderzoek en een Nagele) en proefsleuvenonderzoek op vindplaats verkennend inventariserend veldonderzoek door 14 (Nagelerweg, Emmeloord), Weesp (RAAP- middel van grondboringen, Roterdam rapport 2542). (BOOR-rapport 315). 395: Roymans, J.A.M., 2013: Hert in de pot. 463c: Zijl, W. & D.E.A. Schiltmans, 2008: Herinrichting Tungelroyse Beek deeltrajecten C, D Roterdam Beverwaard Tramremise zone D. en E, gemeenten Nederweert en Leudal. Een bureauonderzoek en een verkennend Archeologisch onderzoek: een archeologische inventariserend veldonderzoek door middel van begeleiding van de graafwerkzaamheden, grondboringen, Roterdam (BOOR-rapport Weesp (RAAP-rapport 2650). 419). 398: Linden, B.A. van der & P. Wemerman, 2014: 467: Guiran, A.J. & O. Brinkkemper, 2007: Archeologische begeleiding conform KNA Roterdam-Randstadrail: Archeologisch protocol Opgraven, plangebied De Mendert Fase onderzoek 1, Emplacement Centraal Station. 2 en 3 te Neer, Doetinchem (Synthegra- Onderzoek van vindplaats 05-42 uit het Laat- rapport S090389). Mesolithicum gelegen op en bij een rivierduin 401: Dijk, X.C.C., 2013: Tussen twee wallen: een door middel van mechanische boringen, opgraving van prehistorische en middeleeuwse Roterdam (BOOR-rapport 318). bewoningssporen aan de Elsenewal in 468: Schiltmans, D.E.A., 2013: Roterdam Nieuwstadt, gemeente Echt-Susteren, Weesp Groenenhagen-Tuinenhoven De Zwanen- (RAAP-rapport 2600. Rietpark. Een opgraving van een vindplaats uit 407: Gerrets, D.A., M. Opbroek & G.L. Williams, het Neolithicum op een rivierduin op IJsselmonde 2012: Onvoltooid verleden. Sporen van meer dan (vindplaats 13-78), Roterdam (BOOR- 10.000 jaar menselijke activiteiten langs de rapport 491). Regge bij Nijverdal, Amersfoort (ADC-rapport 2800). 255 —

472: Ilson, P.J., 2013: Jagers en verzamelaars bij 504: Hoof, L.G.L. van, I.M. van Wijk & C.M. van Scheemderzwaag-Scheemda en –Opdiep. de Linde 2013: Zwervende erven op de löss? Aardgastransportleidingtracé Rysum-Scheemda Onderzoek van een nederzeting uit de vroege (A-660) en Midwolda-Tripscompagnie (A-666), ijzertijd en van sporen uit de Stein-groep te Hof KR-042 (A-660) & KR-006 (A-666), van Limburg (gemeente Sitard-Geleen), Leiden catalogusnummers 16 & 17, gemeente Oldambt. (ARCHOL-rapport 33). Archeologisch onderzoek: opgraving en 508: Rondags,E.J.N.,2013: Archeologisch begeleiding, Weesp (RAAP-rapport 2310). onderzoek aan de Bergerweg te Sitard, 482: Hoof, B.I., T.J. ten Anscher & P. van der aardgastransportleidingtracé Hommelhof- Krof, 2014: Plangebied olieveld Schoonebeek, Schinnen (A-665), catalogusnummer 33, gemeenten Emmen en Coevorden; Archeologisch gemeente Sitard-Geleen; Archeologisch onderzoek: een aanvullend inventariserend onderzoek: proefsleuven, Weesp (RAAP- onderzoek, begeleidingen, opgraving en overzicht rapport 2378). van al het archeologisch onderzoek in het kader 515: Jong, T. de, 2008: Archeologisch onderzoek Son van de herontwikkeling van het olieveld, Weesp en Breugel-plangebied Onstade op Ekkersrijt. (RAAP-rapport 2536). Sporen van bewoning uit de prehistorie en 483: Hoof, B.I. van, P. van der Krof & J.E.A. Jans Romeinse tijd, Eindhoven (Archeologisch 2013: Plangebied Olieveld Schoonebeek, gemeente Centrum Eindhoven rapport 17). Emmen; Archeologisch onderzoek: een opgraving 516: Jong, T. de & S. Beumer 2008: Prehistorische op vindplaats 14, Weesp (RAAP-rapport 2535). bewoning op Ekkersrijt-Onstade (gemeente Son 485: Hoof, B.I. van & P. van der Krof 2012: en Breugel). De opgraving, Eindhoven Plangebied Olieveld Schoonebeek, gemeente (Archeologisch Centrum Eindhoven Coevorden; Archeologisch onderzoek: rapport 18). waarderend proefsleuvenonderzoek en opgraving 517: Jong, T. de & S. Beumer 2012: Opgraving vindplaats 16, Weesp (RAAP-rapport 2436). knooppunt Ekkersrijt-IKEA, gemeente Son en 489: Krol, T.N. & B.I. van Hoof 2010: Plangebied Breugel. Deel II: Prehistorische bewoning in Olieveld Schoonebeek, Oliewinlocatie SCH2000, Ekkersrijt, Eindhoven (Archeologisch gemeente Emmen; Archeologisch vooronderzoek: Centrum Eindhoven rapport 52). een inventariserend veldonderzoek 519: Aalbersberg, G.,2012: Plangebied (proefsleuven), Weesp (RAAP-rapport 1983). Esdoornstraat te Stadskanaal, gemeente 492: Ruijters, M.H.P.M., 2014: Voetsporen, Stadskanaal. Archeologisch vooronderzoek: een dierenpoten en karrensporen op de vindplaats waarderend proefsleuvenonderzoek, Weesp Sevenum-Gelderdijk, Aardgastransportleidingtracé (RAAP-rapport 2507). Odiliapeel-Melick (A-665), catalogusnummer 48; 524: Obdam, T.J.,2014: Een mesolithische vindplaats gemeente Horst aan de Maas, Archeologisch in Wetering-West (Gemeente Steenwijkerland). onderzoek: proefsleuven, Weesp (RAAP-rapport Een archeologische begeleiding (conform protocol 2356). Opgraven), Amersfoort (ADC-rapport 3741). 495: Moonen, B.J. & X.C.C. van Dijk 2011: 528: Wijk, I.M. van, L. Meurkens & A. Porreij- Aardgastransportleidingtracé Odiliapeel-Melick Lyklema 2012: Opgraven langs de Rijksweg A2 te (A-665), gemeenten Sint Anthonis, Venray, Horst Stein-Heidekampweg, Stein-Steinerbos en Geleen- aan de Maas, Peel en Maas, Kessel, Venlo, Beesel, Chemelot. Een archeologisch onderzoek naar een Roermond en Roerdalen. Archeologisch bandkeramische nederzeting en erven uit de (late) vooronderzoek: een inventariserend veldonderzoek ijzertijd, Leiden (ARCHOL-rapport 150). (verkennende en deels karterende fase), Weesp 536: Ruijters, M.H.P.M.,2014: Vondsten uit de (RAAP-rapport 2279). Steentijd op vindplaats Swalmen-Blankwater. 501: Peters, S., 2010: Sint-Oedenrode Kerkstraat. Aardgastransportleidingtracé Odiliapeel-Melick Archeologisch onderzoek, ’s-Hertogenbosch (A-665), catalogusnummer 63, gemeente (BAAC-rapport A-05.0339). Roermond, Archeologisch onderzoek: 503: Mark, R. van der, 2013: “Naast de mote”. Een proefsleuven, Weesp (RAAP-rapport 2363). opgraving van prehistorische en middeleeuwse 537: Tichelman, G.,2006: Bandkeramiek en bewoningssporen in het centrum van Sitard (Het Middeleeuwen in het waterleidingtracé Domein), ’s-Hertogenbosch (BAAC-rapport Sweikhuizen-Geverik, gemeente Beek, A-09.0142). Amersfoort (ADC-rapport 590). 256 —

538: Hamburg, T., A. Müller & B. Quadfieg 2013: 567: Hakvoort, A. & L. van der Meij 2010: Urnen Mesolithisch Swiferbant. Mesolithisch gebruik onder de ploeg. Een opgraving van een van een duin ten zuiden van Swiferbant cultuurlandschap in de microregio ‘Floriade’ (8300-5000 v. Chr.). Een archeologische (gemeente Venlo), Amersfoort opgraving in het tracé van de N23/N307, (ADC-rapport,1204). provincie Flevoland, Leiden (ARCHOL- 572: Fermin, B.,2005: Spiekers biej laag water rapport,174). zeuk’n. Archeologisch onderzoek naar 542: Mooren, J.R., F. van Nuenen, S. de Jager, bewoningsgeschiedenis rond het huis P.A.M. Dijkstra & M. Hendriksen 2007: “De Revelhorst”, Zutphen (Zutphense Itervoort-Thorn, Tracé rioolwatertransportleiding, Archeologische Publicaties 16). Defnitief Archeologisch Onderzoek, 582: Janssens, M.,2013: Van een afvaldump en ’s-Hertogenbosch (BAAC-rapport 5.0289). nederzetingssporen uit de vroege ijzertijd tot het 543: Knippenberg, S. & M.A. Goddijn 2013: 15-eeuwse erf Ter Haer aan de Lochemseweg. Prehistorische resten en middeleeuwse bewoning Aardgastransportleidingtracé Esveld-Angerlo in het Wite Stadje. Een inventariserend en (A-662), catalogusnummer 26, gemeente defnitief archeologisch onderzoek aan de Zutphen, Archeologisch onderzoek: opgraving, Schoolstraat/Molenweg te Thorn, gemeente Weesp (RAAP-rapport 2334). Maasgouw, Leiden (ARCHOL-rapport 203). 592: Wal, A. ter & L.A. Tebbens 2012: Well-Aijen 544: Norde, E.H.L.D. & E.M.P. Verhelst 2012: De werkvak 1, Archeologische opgraving, start van Tiel. Resten uit het Neolithicum en de ’s-Hertogenbosch (BAAC-rapport Vroege Bronstijd in Tiel-De Start, archeologisch A09.0395). vooronderzoek in het plangebied Waterberging, 596: Aalbersberg, G., J.B. Hielkema & D. Mol een inventariserend veldonderzoek door middel 2013: Mammoetresten bij het Oranjekanaal. van proefsleuven, Weesp (RAAP-rapport 2524). Aardgastransportleidingtracé Scheemda-Ommen 548: Hesseling, I., S.M. Koeman, H. Kremer, (A-661), KR-077, catalogusnummer 35, gemeente J.S. Krist & E. Drenth 2013: Defnitieve Midden-Drenthe, Archeologisch onderzoek: opgraving, Tilligte-West te Tilligte, Doetinchem begeleiding, Weesp (RAAP-rapport 2320). (Synthegra-rapport S110246). 597: Krist, J.S., 2013: Defnitieve opgraving, 554: Ruijters, M.H.P.M., 2013: Plangebied cuneten Putenveld Enterweg, gemeente Wierden, te Maastricht Aachen Airport, gemeente Beek, Doetinchem (Synthegra-rapport S120352). archeologisch onderzoek: een opgraving, Weesp 599: Heirbaut, E.N.A., J. Hendriks & I. Hermsen (RAAP-rapport 2648). 2010: Onder een deken van zand. Archeologisch 548: Hielkema, J.B., 2013: Bewoningssporen op het onderzoek op de Meshallen en de Kraanvogel te noordelijke deel van de Varseneres, Wijchen, Nijmegen (Archeologische aardgastransportleidingtracé Scheemda-Ommen Berichten Wijchen rapport 10). (A-661), catalogusnummer 41, gemeente 600: Koning, M.W.A. de, 2010: Onder de rook van Ommen, archeologisch onderzoek: opgraving, Wijchen. Vondsten van de Vlaardingen-groep, de Weesp (RAAP-rapport 2321). Klokbekercultuur, sporen van ovens en resten van 562: Hielkema, J.B., 2014: Een grafveld en een mogelijk grafveldje uit de Late Bronstijd – nederzetingssporen op de Varseneres, Romeinse tijd, Nijmegen (Archeologische aardgastransportleidingtracé Scheemda-Ommen Berichten Wijchen rapport 11). (A-661), catalogusnummer 42, gemeente 615: Weerden, J.F. van der,2010: Veldhoven Ommen, archeologisch onderzoek: opgraving, Kruisstraat. Opgraving, ’s-Hertogenbosch Weesp (RAAP-rapport 2322). (BAAC- rapport A-08.0401). 563: Stokkel, P.J.A., 2008: Resten uit het Mesolithicum, 617: Deitch-van der Meulen, W., 2012: De Laakse IJzertijd/Romeinse Tijd en de Nieuwe Tijd: een Slenk. Slingerend door het Hulkesteinse bos, archeologische noodopgraving aan de Middegaal, Gemeente Zeewolde. Een Archeologische gemeente Veghel en gemeente Bernheze (N.-Br.), begeleiding, Amersfoort (ADC-rapport 3102). Groningen (ARC-publicaties 188). 619: Molthof, H.M. & F.J. Vermue 2013: Zeewolde, 566: Kooistra, M.J. & J.F. van der Weerden 2008: Rassenbeektocht. Een archeologische Veldhoven Habraken-Fase 2, opgraving, begeleiding, Amersfoort (ADC-rapport 2865). ’s-Hertogenbosch (BAAC-rapport A-08.0021). 257 —

627: Bakker, A.M., 2013: Archeologische begeleiding 733: Hermsen, I., M. van der Wal & H. Peeters (protocol opgraven), zes vuursteenvindplaatsen 2015: Afslag Olthof. Archeologisch onderzoek en drie essen in het aardgastransportleidingtracé naar de vroegprehistorische vindplaatsen op tussen winningslocatie NorgUGS (Langelo) en locaties Olthof-Noord en Olthof-Zuid in Epse- Overslag Sappemeer, Heerenveen Noord, Deventer (Rapportage Archeologie (Oranjewoud-rapport 2013/99). Deventer 34). 637: Straten, K. van & B. Fermin 2012: Het 736: Glind, A.M. van den, 2010: Roermond. urnenveld van Leesten-Meijerink. Archeologisch Plangebied Keulsebaan, deelgebied A, onderzoek naar een ritueel landschap uit de ’s-Hertogenbosch (BAAC-rapport ijzertijd en bewoningssporen uit de middeleeuwen A09.0070). en nieuwe tijd bij de erven Wekenstro en 737: Veldhuis, J.R. & J.B. Hielkema: 2010. Een Meijerink bij Zutphen, Zutphen (Zutphense mesolithische vuursteenvindplaats op de locatie Archeologische Publicaties 70). van het Winkelpark de Vosholen te Hoogezand, 645: Kleinen, J., s.d.: Op de enk van Het Asseler. Een gemeente Hoogezand-Sappemeer (Gr), archeologisch onderzoeksproject van de AWN, Groningen (ARC- publicaties 193). Vereniging van vrijwilligers in de Archeologie 738: Meirsman, E. & M.C. Dorst 2005: Roterdam Afdeling 18 Zuid-Salland–IJsselstreek–Oost- Groenenhagen-Tuinenhoven. Een waarderend Veluwezoom april – september 2011. inventariserend veldonderzoek door middel van 651: Schute, A.H. & M. Stiekema 2013: proefsleuven en grondboringen op de top van een Proefsleuvenonderzoek Bourgognestraat (Ong.) fossiel rivierduin, Roterdam (BOOR- te Beek in de Gemeente Beek, Swalmen rapporten 244). (Econsultancy Archeologisch Rapport 739: Meirsman, E. & F.J.C. Peters 2006: 12071584). Roterdam Groenenhagen-Tuinenhoven. 653: Boon, H., 2015: Archeologisch onderzoek Het documenterend archeologisch veldonderzoek Mandelanden bij Borger. Archeologische van neolithische bewoning op de top van een begeleiding, karterend booronderzoek, rivierduin: vindplaats 13-78, Roterdam oppervlaktekartering & fysiek behoud, Assen (BOOR-rapporten 284). (Grontmij Archeologische Rapporten 1317). 740: Diependaele, S. & E. Drenth: 2010. 686: Bongers, J.M.G., 2012: Hollandscheveld, Archeologisch onderzoek langs de rijksweg N11 Langedijk (Gemeente Hoogeveen, Dr.). Een (Spookverlaat) ten behoeve van de aanleg van Inventariserend Archeologisch Veldonderzoek, het windturbinepark Rijnwoude te Hazerswoude- Zuidhorn (Steekproefrapport 2012-10/04Z). Rijndijk (gem. Rijnwoude, prov. Zuid-Holland). 702: Zuidhof, F.S., J. Huizer & N. de Jonge 2012: Een Neolithische vindplaats langs de Oude Rijn, Gebiedsontwikkeling Ooijen, Wanssum, gemeente Nieuwerkerk aan den IJssel (ArcheoMedia- Horst a/d Maas en Venray. Inventariserend rapport A06-286-R en A06-359-R). archeologisch onderzoek ‘Verkenning Plus’ Project 741: Hielkema, J.B., 2006: Jagers-verzamelaars Maasvallei Deelgebieden Hoogwatergeul Ooijen, langs de Wâldwei. Een archeologisch onderzoek Kop van Ooijen, Hoogwatergeul Wanssum en van een vindplaats uit het Mesolithicum, het Wanssum haven, Amersfoort (ADC-rapport Midden-Neolithicum en de Late IJzertijd/ 3183). Romeinse Tijd bij Hempens, gemeente 725: Kimenai, P. & J.R. Mooren 2014: Steentijdsites Leeuwarden (Fr.), Groningen (ARC- langs de Maas Hoogwatergeul Well-Aijen, Werkvak publicaties 131). 2. Inventariserend veldonderzoek, 742: Harmsen, C., 2011: Werkplaatsen uit de ‘s-Hertogenbosch (BAAC-rapport A-10.0340). Steentijd. Een archeologische begeleiding en 731: Fermin, B., 2014: Een duik in het verleden. opgraving aan de Heumenseweg te Alverna – Sporen uit pre- en protohistorie rond het Gemeente Wijchen, Nijmegen (Archeologische voormalige erf Oyshamme op het terrein van het Berichten Wijchen rapport 15). nieuwe zwembad, Zutphen (Zutphense 743: Anscher, T.J. ten, 2013: Plangebied De Centrale Archeologische Publicaties 89). As, deelgebied 118 nabij Sumar. Gemeente 732: Louwe Kooijmans, L.P. & P.F.B. Jongste Tytsjerksteradiel. Arc-heologisch vooronderzoek: 2013: Schipluiden. A neolithic setlement on the een inventariserend veldonderzoek (waarderend Dutch North Sea coast c. 3500 cal BC, Leiden onderzoek), Weesp (RAAP-rapport 2584). (Analecta Prehistorica Leidensia 37/38). 258 —

744: Molthof, H.M. & G.H. de Boer 2014: 753: Huis in ’t Veld, J.Y., 2015: Opgravingen aan het Kabeltracé Emmeloord-Ens, vindplaatsen 1 en 4. Boterdiep te Groningen. De ontwikkeling van Gemeente Noordoostpolder. Archeologisch vuilstort tot voorstad, Groningen (Stadse vooronderzoek: een inventariserend veldonderzoek Fratsen 35). (proefsleuven), Weesp (RAAP-rapport 2825). 754: Krist, J.S. & B. Silkens 2005: Twee 745: Müller, A. & I. Devriendt 2015: Ede-Kernhem opgravingen met beperkingen en aanvullend Vlek B. Opgraving van een mesolithische archeologisch onderzoek te Holtum Noord II, vindplaats, Amersfoort (ADC-rapport 3757). gemeente Sitard-Geleen (L.), Groningen (ARC- 746: Alma, X.J.F. & R. Torremans 2010: Gouda, publicaties 133). Westergouwe – Deelgebied 3: Een vroege 755: Mousch, R.G. van, S. van Dalen, P.A.M. vindplaats op de crevasse. Een Inventariserend Dijkstra & L. Smit 2006: Rijssen-Holten Veldonderzoek in de vorm van proefsleuven, De Liesen. Inventariserend veldonderzoek door Amersfoort (ADC-rapport 2552). middel van proefsleuven, ‘s-Hertogenbosch 747: Opbroek, M. & E. Lohof 2012: Tijd in (BAAC-rapport 06.191). centimeters. Een kijkje in het landschap van een 756: Roller, G.J. de & J.R. Veldhuis 2005: Een dekzandrug te Almere. Een Inventariserend vuursteenvindplaats op het bedrijventerrein Veldonderzoek in de vorm van proefsleuven en ‘De Munt II’ te Emmeloord, gemeente een hoogwaardig booronderzoek, Amersfoort Noordoostpolder (Fl.); een archeologisch (ADC-rapport 2662). inventariserend veldonderzoek (IVO) door middel 748: Brijker, J.M., J.A.A. Bos, D.A. Gerrets & van twee proefsleuven, Groningen M. Opbroek 2013: Tunnel door verdronken (ARC-publicaties 127). landschappen. Kanaalkruising Sluiskil – Zone 3; 757: Wieringa, A. & H. Buitenhuis 2006: Terneuzen. Een Inventariserend Veldonderzoek in Een archeologisch waarderend booronderzoek de vorm van zeefvakken, Amersfoort (IVO) door middel van megaboringen langs de (ADC-rapport 3445). Ter Borghlaan te Eelderwolde, gemeente Tynaarlo 749: Hamburg, T. & S. Knippenberg 2006: (Dr.), Groningen (ARC- rapporten 2006-76). Steentijd op het Spoor. Proefsleuven op drie 758: Tops, B., 2008: Een goed bewaarde locaties binnen het tracé van de Hanzelijn mesolithische vindplaats in het plangebied ‘Oude Land’, Leiden (ARCHOL-rapport 54). Kernhem-vlek B, gemeente Ede. Eerste fase 750: Fens, R.L., J.P. Mendelts & W. Prummel Defnitief Archeologisch Onderzoek, 2013: De trechterbekernederzeting Helpermaar. Amsterdam (Zuidnederlandse De systematische opgraving van een neolithische Archeologische Notities 161). scatervindplaats aan de westzijde van de 759: Wieringa, A.R., 2012: Graven aan het Hondsrug in Groningen-Zuid, Groningen Damsterdiep te Groningen, Groningen (Stadse (Stadse Fratsen 33). Fratsen 30). 751: Heeringen, R.M. van, W.A.M. Hessing, 760: Hoof, B.I. van, 2014: Plangebied Kloosterveen L.I. Kooistra, S. Lange, B.I. Quadfieg, III. Gemeente Assen. Archeologisch R. Schrijvers & W. Weerheijm 2014: vooronderzoek: waarderend onderzoek Archeologisch landschapsonderzoek in het kader (proefsleuven) vindplaatsen I en II, Weesp van het project Kwaliteitsverbetering Koterbos (RAAP-rapport 2813). (locatie Natuurboulevard) in de gemeente 761: Kruining, M.E. van, 2008: Groenlo, Voormalig Lelystad, provincie Flevoland. Menselijke activiteit Grolschterrein (Gem. Oost Gelre). Een in nate landschappen in de Steentijd en de Inventariserend Veldonderzoek in de vorm van (Vroeg-) Romeinse tijd, Amersfoort (Vestigia- proefsleuven, Amersfoort (ADC-Rapport rapport V1132). 1270). 752: Fens, R.L. & J.P. Mendelts 2015: Een 762: Corver, B.A., 2008: Een mesolithisch jachtkamp steentijdcomplex op het Europapark te langs de Hanzeweg N-307, gemeente Dronten. Groningen. Resten uit het laat-mesolithicum en Een Inventariserend Veldonderzoek in de vorm de trechterbekercultuur op de oostelijke fank van van een proefsleuf, Amersfoort (ADC-rapport de Hondsrug, Groningen (Stadse Fratsen 36). 1297). 259 —

763: Veldhuis, J.R., 2010: Een archeologische 772: Molthof, H.M. & T.J. ten Anscher 2015: opgraving van meerdere mesolithische Uitbreiding begraafplaats Siegerswoude, kampementen op de locatie van de toekomstige Gemeente Opsterland. Archeologisch onderzoek: Brede School te Oostwold, gemeente Oldambt een opgraving, Weesp (RAAP-rapport 2828). (Gr. ), Groningen ( ARC--Rapporten 773: Malssen, N. van, J.R. Veldhuis & A.R. 2010-19). Wieringa 2010: Een archeologische begeleiding 770: Alma X.J.F., S. Bloo, R. Machiels, J. de Moor langs het persriooltracé tussen Siddeburen en & S. Ostkamp 2008: Neer, de Mendert Fase 2 Foxhol, gemeentes Slochteren en Hoogezand- en 3, een Laat Paleo-/vroeg mesolithische Sappemeer (Gr), Groningen (ARC- rapporten vuursteenconcentratie. Een inventariserend 2010-231). veldonderzoek in de vorm van proefsleuven, 774: Graaf, W.-S. van de & A. Loonen 2013: Amersfoort (ADC-rapport 1494). Archeologisch onderzoek in het plangebied 771: Hoof, B.I. van & H.W. Veenstra 2012: Daalderveld-Pasestraat te Borgharen, Zevenaar Plangebied Sanjesfild te Oentsjerk. Gemeente (Archeodienst-rapport 137). Tytsjerksteradiel. Archeologisch vooronderzoek: een inventariserend veldonderzoek en waarderend onderzoek (proefsleuven), Weesp (RAAP-rapport 2511). In the last ten to ffeen years development-led archaeology has boosted the number of prospections and subsequent excavations in the Netherlands. Despite the number of excavations and the availability of much data litle is known with regard to the actual gain of knowledge in connection to the wide range of research questions and topics of the National Archaeological Research Agenda. This book reports on a synthetic analysis of reports produced in the context of development-led projects over the past 10-15 years and concerning the early prehistory of the Netherlands. The degree to which development-led work permits to take major steps forward is highly variable. A limited number of topics has profted from the generated data, and in some cases development-led research has shown to be on the forefront of innovative approaches. A larger number of topics and research questions, however, appear to remain unanswerable; data have been generated, but remain of an anecdotal nature. This volume discusses various issues with regard to the gain of knowledge regarding the early prehistory of the Netherlands, and identifes problems, yet also provides possible solutions.

This scientifc report is aimed at archaeologists and other professionals occupied with Archaeology.

With knowledge and advice the Cultural Heritage Agency of the Netherlands gives the future a past.