STUDIES ON STONE AGE HOUSEPITS IN FENNOSCANDIA ( CAL BC)

Teemu Mökkönen STUDIES ON STONE AGE HOUSEPITS IN FENNOSCANDIA (4000 2000 CAL BC) Changes in ground plan, site locaon, and degree of sedensm Academic dissertation to be publicly discussed, by due permission of the Faculty of Arts at the University of Helsinki in auditorium XV, on the 17 th of June, 2011 at 10 o clock. Helsinki 2011

Copyright Teemu Mökkönen ISBN 978-952-92-9059-8 (book) ISBN 978-952-10-6979-6 (PDF) PDF version available at: http://ethesis.helsinki./ Printed in Unigraa Helsinki, 2011 Cover design and layout: Teemu Mökkönen

TABLE OF CONTENTS Preface...i List of Papers...iv Abstract...1 1 INTRODUCTION 1.1 Research questions...3 1.2 The structure of the dissertation...5 2 A BRIEF OUTLINE OF THE RESEARCH AREAS 2.1 Geographical location and watercourses...9 2.2 Research history...11 2.2.1 Previous archaeological research...11 2.2.2 Stone Age cultures in the research area...15 3 HOUSEPITS ARCHAEOLOGICAL REMAINS OF PITHOUSES 3.1 About terminology...19 3.2 Pithouses in ethnographic sources...20 3.3 Archaeological research on housepits in Fennoscandia...23 3.3.1 Finland...24 3.3.2 Sweden...29 3.3.3 Norway...32 3.3.4 The Republic of Karelia, Russia...34 3.4 Housepits and sedentism in archaeological research...34 3.4.1 Finland...35 3.4.2 Sweden...38 3.4.3 Norway...40 3.5 Summation and discussion of housepit research in Fennoscandia...41 4 HOUSING VIS-À-VIS THE ENVIRONMENT 4.1 The relation between subsistence and settlement pattern...45 4.2 Cultural and technological aspects of housing...47 4.3 Environmental studies in Finnish Stone Age research...48 5 CULTURAL CHANGE IN THE STONE AGE, 4000 2000 CAL BC 5.1 The burden of research history: Pottery types as a synonym for culture...51 5.2 Shoreline displacement: A tool for observing cultural change...54 5.3 Changes in houses...55 5.5 Middle Neolithic sedentism...61 5.6 At the northern limit of Corded Ware distribution...62 5.7 After Typical Comb Ware: Population decline or merely cultural change?...63 6 CONCLUSIONS...67 REFERENCES...70

Preface I became concretely involved with housepits in the summer of 1996. At the time, as a young student, I was more interested in Late Iron Age and Medieval archaeology. However, I ended up working as a eld assistant at the excavations of two large housepit sites, Kaustinen Kangas and Yli-Ii Kuuselankangas. These sites truly fascinated me. I believe it was not only the sites but also the enthusiastic atmosphere pervading Finnish housepit research in those days that affected me. There were many new things to explore, and the topics were actively discussed at the excavations. Especially the time in Yli-Ii, with two excavation teams working at adjacent sites and spending their free time together, was perfect from this point of view. Later, I worked as eld assistant on another important housepit site, Kärmelahti in Puumala parish. As a consequence, I consider the excavation leaders I worked with during the late 1990 s, Petri Halinen and Kaarlo Katiskoski, partly responsible for the current state of affairs, with which I am not displeased at all. The mid 1990 s saw an intense period of housepit research in Finland. There was, of course, an active period of housepit excavations in the Kemijoki River area already in the late 1970 s and 1980 s but, as I see it, the 1990 s research was stimulated to a greater degree by studies carried out in Sweden (e.g. Loefer & Westfal 1985; Halén 1994) and Norway, and by the Nordic-Russian joint excavations in the Republic of Karelia, Russia (Karjalainen 1996a). Several projects in Finland resulted in the discovery of large numbers of new housepits, some of which were excavated. The Finnish Ministry of Labour nanced most of the excavations in the Lake Saimaa area as well as in Yli-Ii, Northern Ostrobothnia. At the time, government employment programmes sponsored archaeological excavations in order to provide work for people with records of long-term unemployment caused by the economic depression of the early 1990 s. In practice, the unemployed were forced to work as diggers if they wished to ensure the continuation of their unemployment benets. The archaeological boom in housepit research declined in the beginning of the 21st century. The results of the active research of the 1990 s were published in the Huts and Houses volume (Ranta (ed.) 2002), compiled from papers presented at a seminar on Stone Age and Early Metal Period building remains. At the time, the rst Stone Age timber-frame structures that were brought to light by excavations in the late 1990 s were drawing much attention (Katiskoski 2002, Leskinen 2002). This dissertation is not quite what I thought it would be in the beginning. The rst article (Paper I) was my starting point for this research. It brings up the question of chronological changes in housepits and housepit sites that did not take place simultaneously with changes in material culture, especially in pottery styles. According to my initial research plan, I was supposed to work mainly with the data gathered from the Ancient Lake Saimaa area located in the Vuoksi River catchment. I spent over six months in the archives of the Finnish National Board of Antiquities, making notes and building up a database of Stone Age sites located in that region. However, an unpleasant chain of events, including public false allegations, obliged me to open my research to a competitor, after which I lost my motivation to work on i

that data. Consequently, I ended up working with other material I acquired through other projects. Papers II and IV deal with housepits located on the shores of Ancient Lake Ladoga on the Karelian Isthmus, Russia. I became acquainted with this material in connection with the Kaukola Räisälä Project (2004 2005). Another article on multi-room houses (Paper III) resulted from the discovery of the Meskäärtty site during a survey project I carried out with students in Virolahti parish in 2007 2008. A great many persons have contributed to my work. Firstly, I want to thank my supervisors, Professor Mika Lavento (University of Helsinki) and PhD Jarmo Kankaanpää. I am especially indebted to Jarmo who in addition to most helpful comments revised my English in most of the papers (Papers II V) and in this synthesis paper. I owe many thanks to a number of archaeologists who worked with me at the Department of Archaeology at the University of Helsinki. Discussions, feedback, and developing and generating ideas with Vesa-Pekka Herva, Antti Lahelma, Mikael A. Manninen, Kerkko Nordqvist, Oula Seitsonen, Miikka Tallavaara, Satu Koivisto, Petri Halinen, Eeva-Maria Viitanen, Janne Ikäheimo, Anna Wessman (née Wickholm), and Kristiina Mannermaa were of great importance both for this work and for generally coping with life at the university. On the practical level, ofce secretary Tuovi Laire deserves many thanks for taking care of various matters over the years. There are also numerous other people who have contributed to my work. I owe thanks to Dr Peter Jordan (University of Aberdeen) for the encouraging and valuable comments he made on the early draft of Paper IV. Special thanks to palynologist Teija Alenius (University of Helsinki) for discussions and for her comments regarding the paper on early cereal cultivation (Paper V). I would also like to thank Dr Volker Heyd (University of Bristol) for his valuable comments in discussions during his term as visiting scholar in Helsinki. I wish to thank the students taking part in the surveys in Virolahti parish, southeastern Finland, in 2007 2008 (listed in Paper III). Originally, these surveys were not meant to play any part in my dissertation work, but fate decided otherwise. One of the discovered sites, the Meskäärtty housepit site (Paper III), was documented in extreme conditions in November 2007 (it was freezing and the rst snow fell). I thank Kerkko Norqvist, Heidi Nordqvist, and Wesa Perttola for assisting in generation of the surface model of the housepit. If one wishes to do eldwork abroad, cooperation with local archaeologists is essential. Two papers in this work deal with material from the Karelian Isthmus, Russia. Archaeologists Dmitriy Gerasimov and Stanislav Belskij facilitated the work in Russia. In addition to taking care of much of the practical arrangements during the eld season and preparing the Russian reports, they also offered their great hospitality in St. Petersburg when I was analysing the nds there. Therefore, I owe special thanks to them and to the others participating in the eldwork of the Kaukola Räisälä Project (Listed in Papers II and IV). I would also like to thank the people with whom I have had various, more or less scientically relevant, discussions every now and then: Taisto Karjalainen, Petro Pesonen, Esa Hertell, Timo Jussila, Aivar Kriiska, Vadim Adel, and Santeri Vanhanen. Although the topics were often far from the core of this work, those discussions have ii

been important believe it or not! In addition, I wish to thank the personnel of the Finnish National Board of Antiquities (Fi. Museovirasto) who were of great help during the archival research I carried out (but which was not used in this study): Sanna Saunaluoma, Tanja Tenhunen, Leena Söyrinki-Harmo, and Leena Ruonavaara. Special thanks for the excellent and refreshing coffee breaks. In addition, I thank Helena Ranta (National Board of Antiquities) as well as Rami Urrio and Kimmo Oikarinen (Gummerus Printing House, Jyväskylä) for their efforts to provide me with a printable.pdf le of Paper I. Last but not least, I would like to thank my family my wife Liisa Lohtander and my sons Aarni and Kauko. Liisa has served as proofreader for some of the papers. Otherwise, my family deserves special thanks for giving me many very un-academic matters to think about (often related in some way or other to Lego building or diapers). Finally, I offer my gratitude to the foundations and institutions that have made it nancially possible to write this dissertation. I have received generous support from the Finnish Graduate School in Archaeology at which I worked during 2007 2009. Before and after graduate school, my work was nanced by grants awarded by the Finnish Cultural Foundation (Fi. Suomen kulttuurirahasto) and a minor grant provided by the Oskar Öunds Stiftelse. The eldwork on the Karelian Isthmus, Russia, carried out in connection with the Kaukola Räisälä Project (2004 2005), was nanced by a Chancellor s research grant (University of Helsinki, Finland). The eldwork in Virolahti and the radiocarbon datings were nanced by The Centenary Foundation of Kymi Corporation (Fi. Kymin osakeyhtiön 100-vuotissäätiö). To all of these institutions, I am thankful for the support that made this dissertation possible. iii

List of papers This dissertation is composed of a synthesis paper and ve previously published papers. Papers II V are peer-reviewed. In this synthesis, the papers are referred to by their Roman numerals. Paper I Mökkönen, T, 2002. Chronological variation in the locations of hunter-gatherer occupation sites vis-à -vis the environment. In H. Ranta (ed.). Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 53 64. [references p. 240 254] Paper II Mökkönen, T., Nordqvist, K. & Belskij, S. 2007. The Rupunkangas 1A site in the archipelago of ancient Lake Ladoga: a housepit with several rebuilding phases. Fennoscandia archaeologica XXIV, 3 28. Paper III Mökkönen, T. 2008. A Review of Neolithic multi-room housepits as seen from the Meskäärtty site in Virolahti parish, extreme south-eastern Finland. Estonian Journal of Archaeology, Vol. 12, No. 2, 114 151. Paper IV Mökkönen, T. 2009. Neolithic housepits in the River Vuoksi Valley, Karelian Isthmus, Russia chronological changes in size and location. Fennoscandia Archaeologica XXVI, 133 161. [Errata & Corrigenda. Fennoscandia Archaeologica XXVII, 108 109.] * Paper V Mökkönen, T. 2010. Kivikautinen maanviljely Suomessa (with English summary Neolithic cereal cultivation in Finland ). Suomen Museo 2009, 5 38. * The original article (Fennoscandia Archaeologica XXVI: 133 161) contained errors in Figures 8 and 9. Those gures were interchanged and, added to this, the gure texts contained further errors. The Errata is published in Fennoscandia Archaeologica XXVII (2010). For the sake of clarity, a version of the whole article with corrected gures is published here. iv

Teemu Mökkönen STUDIES ON STONE AGE HOUSEPITS IN FENNOSCANDIA (4000 2000 CAL BC) Changes in ground plan, site locaon, and degree of sedensm Abstract Housepits have a remarkably short research history as compared to Fennoscandian archaeological research on the Stone Age in general. The current understanding of the numbers and the distribution of Stone Age housepits in the Nordic countries has, for the most part, been shaped by archaeological studies carried out over the last twenty to thirty years. The main subjects of this research are Neolithic housepits, which are archaeological remains of semi-subterranean pithouses. This dissertation consists of ve peer-reviewed articles and a synthesis paper. The articles deal with the development of housepits as seen in the data gathered from Finland (the Lake Saimaa area and south-eastern Finland) and Russia (the Karelian Isthmus). This synthesis expands the discussion of the changes observed in the Papers to include Fennoscandian housepit research as a whole. Certain changes in the size, shape, environmental location, and clustering of housepits extended into various cultures and ecological zones in northern Fennoscandia. Previously, the evolution of housepits has been interpreted to have been caused by the adaptation of Neolithic societies to prevailing environmental circumstances or to re-organization following contacts with the agrarian Corded Ware/Battle Axe Cultures spreading to North. This dissertation argues for two waves of change in the pithouse building tradition. Both waves brought with them certain changes in the pithouses themselves and in the practices of locating the dwellings in the environment/landscape. The changes in housepits do not go hand in hand with other changes in material culture, nor are the changes restricted to certain ecological environments. Based on current information, it appears that the changes relate primarily to the spread of new concepts of housing and possibly to new technology, as opposed to representing merely a local response to environmental factors. This development commenced already before the birth of the Corded Ware/Battle Axe Cultures. Therefore, the changes are argued to have resulted from the spreading of new ideas through the same networks that actively distributed commodities, exotic goods, and raw materials over vast areas between the southern Baltic Sea, the north-west Russian forest zone, and Fennoscandia.. 1

2

1 INTRODUCTION 1.1 Research quesons Finnish archaeological research on housepits peaked in the 1990 s (see Pesonen 2002). At the time, research was largely focused on chronological and chorological questions, the material culture associated with housepits, and structural details. The settlement pattern was expected to be either semi-sedentary or fully sedentary, and the argumentation concerning the seasonal duration of the occupation was based on the species detected in osteological analyses (e.g. Pesonen 1996b; Halinen et al. 1998; Karjalainen 1999). Most of the studies were based on a single excavated housepit or housepit site. They may be characterized as descriptions of excavated sites rather than as comprehensive studies on housepits. As in Finland, housepits are a fairly recently discovered subject on the Karelian Isthmus (Russia). The rst housepits in this area were found in 1999 (Lavento et al. 2001), and surveys in 2004 and 2005 brought the number of housepits in the Kaukola Räsiälä area up to over eighty (see Paper IV). During the rst decade of the 21st century, housepits were also discovered in other areas of the Karelian Isthmus (Nordqvist & Seitsonen 2008; Seitsonen & Gerasimov 2008; Seitsonen et al. 2009). Based on the research carried out in the 1990 s, it was known that the main building phase of housepits in Finland dated to the Middle Neolithic Period (4000 2300 cal BC) and that most of the housepits were associated with Typical Comb Ware, Late Comb Ware, and asbestos-tempered Kierikki and Pöljä(-Jysmä) wares (e.g. Kotivuori 1993; Karjalainen 1996b; 1999; 2002; Nuñez & Uino 1997; Pesonen 2002). Actually, this association with ceramics was already indicated in Meinander s article from 1976, which was based on his presentation at the Nordiska arkeologimötet (Nordic Archaeological Congress) in Helsinki in 1967. During the 1990 s, it was also known that the largest structures dated to the late Middle Neolithic were often associated with asbestos-tempered Kierikki and Pöljä Wares (e.g. Nuñez & Uino 1997; Núñez & Okkonen 1999; Halinen et al. 2002; Pesonen 2002; Paper I). Only during the last decade, however, have a number of more comprehensive studies sought to nd chronological variation in the housepit building tradition with the help of chronological frameworks offered by shoreline displacement (Mökkönen 2000; Paper I; Okkonen 2003; Norberg 2008; Vaneeckhout 2008a b; 2009a c, 2010). These frameworks, in which the changes are studied with the help of shoreline displacement, avoid the pitfalls attached to chronological typologies in ceramics (see Mökkönen 2009). Constant shoreline displacement, a legacy of the last Ice Age, is in a global sense an exceptional natural phenomenon. In most parts of Finland, it al- 3

lows the arranging of shore-bound archaeological sites into chronological order and, at the same time, allows the creation of parallel chronologies on various themes that are independent of the chronological frameworks of material culture (i.e. artefact typologies). This facilitates the relative dating of unexcavated sites without any direct dating methods such as radiocarbon dating (see more in Chapter 5.2). This study focuses on chronological changes in the construction and design of pithouses as well as in the environmental locations of housepit sites (Papers I, II, III and IV). Other cultural aspects such as interaction between groups with a different material culture (Paper III), degree of sedentism (Papers I, II, and IV) and early cereal cultivation (Paper V, see also Paper IV) are discussed through the housepits and material associated with them. This thesis seeks answers to three main questions: (1) What kind of variation is there in Neolithic housepits (size, shape, clustering) and in housepit site locations (in relation to the immediate environment as well as to larger environmental zones), and how is this variation chronologically and regionally distributed? (Papers I, III and IV) (2) What sort of settlement system do the changes in housepits indicate? (Papers I, II, and IV) (3) What were the driving forces behind the changes? (Papers I, III, IV and V) This dissertation deals with housepits dating mainly from ca. 4500 to 2000 cal BC. The material derives from two case study areas, the Ancient Lake Saimaa area, Finland (Paper I), and the Karelian Isthmus, Russia (Papers III and IV). Paper III introduces a recently discovered three-room housepit from Virolahti parish, extreme south-eastern Finland, and discusses the emergence, distribution, and cultural context of large multi-room housepits in Finland and in the Republic of Karelia in Russia. Papers I IV are based on survey data, in the gathering of which I was also involved through three research projects. The observations made in connection with the case study in the Ancient Lake Saimaa area (Paper I) form the starting point on which the present study builds. In that data, the chronological changes in the ground plan and clustering of housepits, as well as the radical change in the pithouses placement in the landscape, did not occur concurrently with the changes seen in the material culture, especially in ceramics. In the Lake Saimaa area, the shoreline displacement caused by isostatic land uplift rendered it possible to observe the chronological changes in housepits and to study those changes in relation to ceramics found at the sites. The rst paper sets down the observations but makes no suggestions towards a wider cultural understanding of the observed phenomena. Although housepits form the core of this study, other subjects are also discussed at length. Paper III deals with general cultural evolution during the late Middle Neolithic period as much as it does with multi-room housepits. The last article (Paper V) concentrates on the latest results concerning early cereal cultivation. It is included here because the most obvious changes in the pithouse building tradition appear approximately simultaneously with the anticipated initial cereal cultivation in the 4

spheres of both Lake Ladoga and Lake Saimaa. Papers III V I seek to reach a broader understanding of the cultural developments to which the changes seen in housepits are chronologically related. 1.2 The structure of the dissertaon This dissertation consists of this synthesis paper and ve papers published between 2002 and 2010. The primary archaeological data used in the papers originates from three areas (Fig. 1). Papers I IV are in English and the last one (Paper V) is in Finnish with an English summary. The papers are included here in chronological order according to the publishing date. The main theses of the papers can be summarized as follows: Paper I. This paper presents the chronological changes in housepits and housepit sites in the Lake Saimaa case study area. The paper argues that a fairly radical change as regards the sites immediate environment takes place at the time when Typical Comb Ware was in use. During the said period, the pithouse sites shifted from sheltered locations at the heads of bays to unsheltered islands and peninsulas. This shift was accompanied by a village-like clustering of pithouses. The period with villagelike concentrations seems to have been quite short-lived. Based on the concurrent emergence of village-like pithouse sites and the drop in the number of other dwellings sites, this paper submits that the degree of sedentism increased already during the late Typical Comb Ware period. This article is based on the data I used for my Master s thesis (Mökkönen 2000). Paper IV continues this theme with data from the Karelian Isthmus, Russia. Paper II. This paper concentrates on housepits discovered on a large former island that was once located in the archipelago of Ancient Lake Ladoga but is now part of the Karelian Isthmus, Russia. The paper focuses on a partly excavated housepit, Rupunkangas 1 A, which has a notably long reuse history that covers the whole Ancient Lake Ladoga period (ca. 7800 1350 cal BC). The scarcity of nds and the unsheltered site location in the archipelago, clearly outside the heartland of the Stone Age occupation, led me to interpret the site as a hunting station used on a recurring seasonal basis. The formation of distinctly thick cultural layers on the site is assumed to be connected with the abandonment processes rather than with the actual occupation phases. The last part of this paper discusses other housepits located on the former island. Paper III. This paper introduces a large three-room housepit from the Meskäärtty site, extreme south-eastern Finland. This site is exceptional in several respects: previously, such large Stone Age structures were not known from the Gulf of Finland and the nds are atypical as compared to both the material usually found in Finland and to the material typically associated with housepits. The nds include a heterogeneous pottery assemblage best described as Late Comb Ware and Late Corded Ware, and a sharp-butted axe, both of which have similarities with material found on the southern shore of the Gulf of Finland in Estonia. This paper discusses the late Middle and Late Neolithic (ca. 3500 2000 cal BC) phenomena relating to cultural contacts seen in the 5

material culture and in multi-room housepits (dating and chorology). In conclusion, this paper suggests that contact with the Corded Ware Culture had a profound impact on local cultures. This is clearly visible in the appearance of elongated multi-room houses that express ideas similar to longhouses but were built following the local building tradition in which structures were, as a rule, semi-subterranean. However, as will be noted later in this synthesis paper, the idea presented in article that the hybridlike pottery is connected to Corded Ware contacts is not applicable in the case of the oldest pottery, the dating of which is a bit too old considering the current dating of the Corded Ware Culture (vide Chapter 2.2.2). Paper IV. Since the discovery of the rst housepits on the Karelian Isthmus, Russia, in 1999, a number new of housepits have been found. The housepits discovered in the study area of the Kaukola Räisälä project (2004 2006) are published in this article. The article focuses on the chronological changes in housepit size, clustering, and location, and presents suggestions as to the causes behind the observed developments. The most distinctive change in the pithouse building tradition occurred during the time when Typical Comb Ware (ca. 4000 3400 cal BC) was the dominant pottery style. At that time, several changes appeared: the number of pithouses per site increased, the size of individual pithouses grew, and the placement of pithouse sites changed from highly sheltered to windy locations in the archipelago. The last of these developments obviously favours other than winter-only habitation. This paper argues for a growing degree of sedentism during the Typical Comb Ware period. Although it is still poorly arguable, the larger and more oblong housepits dating to the late Middle Neolithic are seen as probable multi-family houses representing fairly stationary settlement. Paper V. The last paper deals with the beginning of cereal cultivation around the eastern shores of the Baltic. During the last decade, pollen studies have shown that cereal cultivation appeared in the Baltic States already during the Early Neolithic. As far as the northern areas are concerned, the most interesting results come from Estonia and the eastern shore of Lake Onega, Russia, where cereal cultivation was practised during the early 4th millennium cal BC. This early cultivation was practised among archaeologically detectable cultures, such as Typical Comb Ware and Comb-Pit Ware, which are either the same as or closely related to the archaeological complexes found in the study area (eastern and south-eastern Finland and the Karelian Isthmus, Russia). The lack of evidence (i.e., unquestionable cereal pollens or macrofossils) of such early cereal cultivation in the study area may be due to the paucity of research as well as to the wrong kinds of assumptions deriving from Central European palynological research on early cultivation. This paper calls for new conventions for studies on northern early cultivation, which should be oriented towards nding low-intensity cereal cultivation unlike that practised among Central European Early Neolithic cultures. This paper points out that the prevailing model of the spreading of agriculture in the eastern Baltic Sea region synchronously with the spreading of the Corded Ware Culture is no longer valid. Instead, the presence of cereal cultivation among local (sub)neolithic cultures over a thousand years prior to the appearance of the Corded 6

Ware Culture raises important questions: What is the mechanism through which early cereal cultivation spread to North European Subneolithic cultures, and is the term Subneolithic still acceptable? The reason for including this paper in this dissertation is the fact that the suggested beginning of initial cereal cultivation and the most distinctive changes in housepits (which are interpreted to be connected with a growing degree of sedentism or even fully stationary settlement, see Papers I and IV) both take place almost simultaneously during the Typical Comb Ware Phase. ***** All the papers deal with housepits and the archaeological cultures that built these structures. The papers dealing mainly with housepits pose new questions and interpretations through the observations made on the material. Paper I presents the chronological changes in housepits and, for the rst time, brings up the problem of a changing settlement pattern during one cultural phase dened by its ceramic style. This idea is carried on throughout the work. This synthesis paper is a further exploration of the most interesting themes presented in the Papers and expands the discussion to cover the corpus of research on Fennoscandian housepits. After the introduction, this synthesis continues with a general presentation of the main research areas (Chapter 2). I then dene the used terminology (Chapter 3.1) and present a review of ethnographic research on pithouses (Chapter 3.2). Next, I outline the archaeological research on Fennoscandian housepits (Chapter 3.3). This chapter traces the changes in housepits as observed in the papers. This is followed by a chapter dealing with pithouses and sedentism (Chapter 3.4), which focuses on archaeological interpretations of sedentism among pithouse dwellers in Fennoscandia. Chapter 4 examines the relationships between the physical environment, culture, and housing. The last chapter before the conclusions, Cultural change in the Stone Age 4000 2000 cal BC (Chapter 5), combines and discusses further the observations made in the papers and in this synthesis paper. Both the evolution of housepits and the ways of dening an archaeological culture are at the centre of the discussion. In the last, concluding chapter, I sum up the main points of this work. One ambition of this dissertation was to maintain an approach in which the questions are not answered only through individual sites. Here, the sites are integrated into wider chronological, environmental, and cultural contexts. In my opinion, it is necessary to maintain the connection between different cultural aspects. I believe that an understanding of a culture cannot be attained without a holistic view on various cultural phenomena. This dialogue should also be attached to the broader discourse on the subject. It is not an easy task to keep up this dialogue, but it is still worth trying. 7

8

2 A BRIEF OUTLINE OF THE RESEARCH AREAS 2.1 Geographic locaon and watercourses The primary archaeological data used in this study originates from the Lake Saimaa area in the Finnish inland lake region, the River Vuoksi Valley on the Karelian Isthmus, Russia, and Virolahti Parish in extreme south-eastern Finland (Fig. 1). Although these three areas are associated with different water systems, they still lie within a radius of less than 250 kilometres. Virolahti parish is located on the Gulf of Finland, while the two case study areas are located by two major inland lake systems, Lake Saimaa and Lake Ladoga, which are connected to each other by the present outow channel of Lake Saimaa, the Vuoksi River. During most of the Stone Age, Ancient Lake Saimaa was a transgressive basin. Because the rate of land uplift is highest in the north-western part of the basin and lowest at the south-eastern corner, the outow channel has slowly migrated from the north to the west and onwards to the south-eastern corner of the basin (Fig 2). After the formation of the present outow channel, the Vuoksi River, in the south-eastern corner ca. 4000 cal BC, the shoreline has been regressive over the whole lake basin (Saarnisto 1970; Jussila 1999). The change from a transgressive to a regressive shoreline also changed the environment. Due to the regressive shoreline, new soil, more ne-grained and fertile than before (Vikkula 1994), was exposed, and this probably had some impact on the vegetation (Mökkönen 2000: 129 131). In the case study area around Lake Saimaa (Paper I), the highest water level was reached between 4500 and 4000 cal BC, that is, just before the formation of the present outow channel, the Vuoksi River. Nowadays, this area is referred to as the Vuoksi River catchment. Despite the fact that the whole Vuoksi River catchment area is larger than the maximum extent of Ancient Lake Saimaa, these two names are often used as synonyms. In the region of Ancient Lake Ladoga (ca. 7800 1350 cal BC), the rate of land uplift is much slower than in the Lake Saimaa area. Actually, the present zero isobase for land uplift runs a few kilometres north of St. Petersburg, whence it continues as a roughly SW. NE. oriented line through Lake Ladoga. During the Ancient Lake Ladoga Phase (ca. 7800 1350 cal BC), the outow channel was located in the northern part of the Karelian Isthmus. At that time, the land uplift isobase on which the threshold was located split the ancient lake into a northern regressive area with a higher land uplift rate and a southern transgressive area with lower rate of land uplift. The case study area in the present Vuoksi River Valley (Papers II and IV) was located approximately on the same land uplift isobase as the Stone Age threshold. 9

Consequently, the shoreline was nearly stable before the formation of the present outow channel, the River Neva (ca. 1350 cal BC) (see Saarnisto 2008). From ca. 4000 cal BC onwards, Lake Saimaa and Lake Ladoga have been connected by the Vuoksi River (Saarnisto 1970; Saarnisto & Siiriäinen 1970). Although the new inow from Lake Saimaa accelerated the transgression in Lake Ladoga, the uctuation in the research area (the Kaukola Räisälä area) has been limited to one to two metres (Saarnisto 2008). Lake Saimaa and Lake Ladoga are today lakes of considerable size. During the Stone Age, their surface areas were much wider still than today. Although the water was not salty and species were not as abundant as they were in the Baltic at the time, they can still be considered inland seas due to their impressive size, with their own seal populations that were trapped in the basins after the Ice Age. In contrast to the lake areas, the whole Neolithic Stone Age in the Virolahti area on the Gulf of Finland was a period of continuous and constant shoreline regression (Miettinen, A. 2002). During the Neolithic, the easternmost part of the Finnish South Coast was topographically very similar to the Vuoksi River Valley on the Fig. 1. The research areas. The Kerimäki case study area (Paper I) is located in the Eastern Lake Saimaa area, Finland. The Kaukola Räisälä case study area (Papers II and IV) is located on the Karelian Isthmus, Russia. The Meskäärtty site (Paper III) is located in extreme south-eastern Finland. The area with darker grey shows the extent of the River Vuoksi catchment, which corresponds quite closely to the Ancient Lake Saimaa area (see Fig. 2). 10

Fig. 2. The River Vuoksi catchment and the highest shoreline of Ancient Lake Saimaa. The Kerimäki case study area is marked with a square. The opening dates of the outow channels are the following: (1) Pielavesi ca. 7500 cal BC, (2) Matkuslampi ca. 4900, (3) Kärenlampi 4400 cal BC, and (4) the Vuoksi River ca. 4000 cal BC. Before the formation of the Vuoksi River, the lake was divided into two parts, one with transgressive and the other with regressive shores, depending on the location of the active outow channels. As a consequence of the shift of the outow channels from the NW. to the SE. corner of basin, i.e., from the area of rapid land uplift to the one with a slower rate, the highest shoreline is metachronous in character. It is oldest in the north-western part and youngest in the south-eastern part. Therefore, the maximum extent of Ancient Lake Saimaa illustrated in the map has never existed at one time. The isobases show the elevation of the highest shorelines illustrated in the map. (Isobases after Saarnisto 1970: App. VIII; The highest shoreline after GTK 1996). Karelian Isthmus, Russia. Deep and narrow bays stretching far into the interior and, in contrast, long capes reaching far out to sea characterized both areas (Mökkönen & Seitsonen 2007; Halinen & Mökkönen 2009). With respect to topography, southeastern Finland and the Vuoksi River Valley are more clear-cut than the mosaic-like topography of the Ancient Lake Saimaa area. 2.2 Research history 2.2.1 Previous archaeological research The primary data used in the Papers originate from areas where comprehensive archaeological surveys are of quite recent date. These surveys have revealed the true archaeological potential of the areas. This holds true especially in the case of housepits, which Finnish archaeologists in general learned to indentify only in the late 11

1980 s and early 1990 s. 1 Accordingly, the corpus of housepit data from these areas has been built up during the last twenty-odd years. In addition, the two case study areas one in Kerimäki parish and another in the Kaukola-Räisälä area were both intensively surveyed for the case studies (see below). During the period from the 1940 s to the 1950 s, a small number of excavations in the Ancient Lake Saimaa area produced asbestos-tempered pottery. Based on studies of the Pöljä site in Siilinjärvi Parish in the northern part of the Vuoksi River catchment, C. F. Meinander (1954a) dened the Pöljä type of asbestos ceramics, i.e., Pöljä Ware. Ten years later, Torsten Edgren (1964) described a Late Stone Age asbestostempered pottery type, Jysmä Ware, which was considered younger than Pöljä Ware. The latest typological and chronological study of asbestos-tempered ceramics deals with Early Asbestos Ware (Pesonen 1995; 1996a; 2001), which is a pottery type known to have existed from the early 20 th century BC onwards (see also Meinander 1954a; Edgren 1966). Looking at the studies of ceramics, one could say that the Stone Age research history of the Ancient Lake Saimaa area is entwined with asbestostempered potteries. Before the 1980 s, Ancient Lake Saimaa was a poorly-studied area as far as archaeology is concerned. During the 1980 s, this status began to change. In 1987, the Savonlinna Provincial Museum launched a survey project in Southern Savo Province (Fi. Etelä-Savon muinasjäännösinventointi 2000 -projekti), which revealed the research potential of the Lake Saimaa area (Fig. 3; Jussila et al. 1992; Lehtinen & Sepänmaa 1995). The results of this project were mainly published in a series of the Savonlinna Provincial Museum called Sihti (numbers 1 4). The rst extensive survey project was followed by the Ancient Lake Saimaa Project (Fi. Muinais-Saimaa -projekti, 1992 1996), which was conducted as a co-operation between the University of Helsinki, the provincial museums of Savonlinna and Kuopio, and the National Board of Antiquities (see Siiriäinen 1996; Vikkula 1995; Lavento 2008a). The project had a major impact on the knowledge of the prehistory of the area (Fig. 4). Even thought the studies yielded a mass of new sites and materials of various kinds, the use of this data in academic studies is still rather limited. Later on, this project was followed up by the Saimaa-Ladoga Project (Fi. Saimaan Vuoksi projekti, 1998 2003), in which the focus of the study was shifted from the Ancient Lake Saimaa area to the lower part of the Vuoksi River, i.e., to the Karelian Isthmus, Russia (Lavento 2008b). The case study area in Kerimäki municipality, eastern Finland, was surveyed for the rst time in 1991 (Jussila et al. 1992; Sepänmaa 1991, see also Sepänmaa 1995). For the case study (Mökkönen 2000; Paper I), the area was intensively re-surveyed in 1998 (Mökkönen 1999). The studies in Kerimäki parish were carried out in connection with the Martinniemi Project (1998 1999), which can be dened as an autonomous project intermediate between the Ancient Lake Saimaa Project and the Saimaa- Ladoga Project (Lavento 2008b) or as a sub-project within the latter. 1 It must be noted, however, that during late 1970 s and the 1980 s numerous housepits were found and excavated in Tervola Parish, southern Lapland, (see Kotivuori 1993; 1998; 2002). Before the 1990 s, several housepits were known and some were excavated also in other parts of Finland (Meinander 1976; Matiskainen & Jussila 1984; Miettinen 1982; 1983; Hiekkanen 1984). 12

Fig. 3. In 1987, the Savonlinna provincial museum initiated a survey project in Southern Savo Province, in the Lake Saimaa area. The front cover of a questionnaire booklet for local inhabitants asks: Are there any ancient sites in Jäppilä? (Lehtinen 1987). The extreme south-eastern part of Finland, i.e., the south-eastern corner of Finland bounded by the Gulf of Finland and the Finnish-Russian border, has long been a weakly studied area. Until a few years ago, it lacked even the most elementary archaeological surveys. Consequently, this area has been poorly represented in the distribution maps of housepits (Miettinen, T. 1998; Pesonen 2002). Over the last few years, new archaeological surveys have revealed several new housepit sites. The Meskäärtty housepit (Paper III) is no longer the only large multi-room housepit by the Gulf of Finland. Another two-room housepit measuring 28 x 20 metres in size at the Karpankangas site in Virolahti parish was discovered in 2009 (Vuoristo 2009a b). In addition, there are other large-sized housepits at the Karpankangas site, the largest measuring ca. 37 x 20 metres in size. Also a small number of other large housepits (Enqvist 2007; see also Nordqvist & Seitsonen 2008) as well as smaller ones (Enqvist 2006) have been reported over the last few years from the eastern Gulf of Finland coast. The recent surveys have revealed that the Stone Age coastal zone in south-eastern Finland has been underrepresented in the data used in previous chorological studies on housepits (Pesonen 1999a; 2002). Although the latest surveys are recent, the time available for eldwork was limited when compared to the large area of the surveyed municipalities. There are thus reasons to suspect that a large number of unknown housepit sites still remain to be discovered. While intensive archaeological research in the Lake Saimaa area and in extreme south-eastern Finland is of recent age, the situation on the Karelian Isthmus is quite the opposite. Archaeological material collected and excavated in the Kaukola Räisälä case study area located on the Karelian Isthmus (Russia) has played an essential part in early Stone Age research in Finland. The rst Stone Age excavation ever conducted by a Finnish archaeologist took place in Räisälä in 1892, when Theodor Schvindt excavated the Teperinaho site (Uino 2003: 127). The most intensive period of Stone Age research was a bit later, during the rst two decades of the 20 th century (Lavento et al. 2001; Huurre 2003: 154 157; Uino 2003). The early research is well documented in publications and articles both by the archaeologists who conducted the research (Ailio 1909; Pälsi 1915; 1918) as well as by current archaeologists (Uino 1997; 2003; Lavento et al. 2001; Huurre 2003; Lavento 2008a b; Nordqvist et al. 13

40 35 30 25 20 15 10 5 0 Excavaons Excavaons Trial excavaons and rescue excavaons Inspecons 140 133 120 109 100 80 60 40 20 0 4 2 2 6 0 0 19 12 12 5 9 9 14 12 6 5 60 11 24 4644 31 5 9 Inspecons Surveys Fig. 4. The number of archaeological reports from the Ancient Lake Saimaa area up to the summer of 2005. The peak in research activities in the Ancient Lake Saimaa area associated with the projects of the late 1980 s to the late 1990 s is obvious. The sum total of archaeological reports consisted of 163 excavation reports, 189 survey reports, and 426 inspection reports. 14

2009). The material from the Kaukola Räisälä area has been frequently used in many later studies concerning the Stone Age and Early Metal Period (e.g. Carpelan 1965; 1979; Huurre 2003; Lavento 2001; Meinander 1954a b; Vikkula 1988). In the Treaty of Paris (1947), the research area in Kaukola and Räisälä was ceded by Finland to the Soviet Union along with the rest of the Karelian Isthmus. This was followed by a quiet period in Stone Age research, and only a few surveys and excavations were carried out there before the 1990 s. The archaeological material from the early 20 th century is catalogued at the Finnish National Board of Antiquities, and consequently Russian archaeologists were not aware of the existence of sites in the area. From 1993 to 2002, a Russian archaeologist, V. I. Timofeev, conducted small-scale excavations at some of the sites around Lake Riukjärvi in Kaukola. These sites were originally discovered and excavated by Finnish archaeologists during the early 20 th century. (Uino 1997: Append. 1; 2003: 141 142; Lavento et al. 2001: 6 8; Mökkönen & Nordqvist 2006; Nordqvist et al. 2008). A second intensive research phase began in 1999, when Finnish Russian co-operation commenced with a survey of the Kaukola Räisälä area (Lavento et al. 2001; Nordqvist & Lavento 2008). The co-operation continued in 2002 with an excavation of a housepit site, Juoksemajärvi Westend, in Räisälä municipality (Seitsonen 2005a; Halinen et al. 2008; Timofeev et al. 2004; see also Halinen & Mökkönen 2009). The research continued in the form of the Kaukola Räisälä Project (2004 2006), with intensive surveys and small-scale excavations (Lavento et al. 2006; Mökkönen et al. 2006; Halinen & Mökkönen 2009). Since 1999, the total number of known Stone Age and Early Metal period sites increased from 24 to 145. Similarly, after the rst housepits on the Karelian Isthmus were discovered in 1999 (Lavento et al. 2001), the number of housepits in the Kaukola Räisälä area has grown to over eighty (Papers II and IV). 2.2.2 Stone Age cultures in the research area In Finnish archaeology, the typology of ceramics has played a central role. From the establishment of Äyräpää s (1930) typology of Comb Ware up to the early 21 st century, pottery types were thought to form chronological periods that followed one another quite rigidly. This lead to the ceramic types being treated as chronological sequences, and these in turn were transformed into cultures or cultural phases in a similar fashion as in Russian archaeology (e.g. Lavento 2002). Finnish research has traditionally concentrated on artefact typologies, and in Neolithic and Early Metal Period contexts the ceramic types are considered cultural/ethnic markers. However, thanks to AMS radiocarbon dates on charred crust adhering to ceramics, it is now known that there is much more overlap between pottery types than has been previously thought (Fig. 5). For example, Typical Comb Ware and Late Comb Ware (referred to henceforth respectively as CW2 and CW3), which were thought to represent two successive chronological sequences, actually have a longer period of co-occurrence than the span of these ceramic types as the sole type of Comb Ware in use (Pesonen 2004; Pesonen & Leskinen 2009). 15

Fig. 5. Suggestion for a chronological diagram of the pottery styles in Finland and the Republic of Karelia, Russia. Stone Age ceramics in grey; Early Metal period/bronze Age Early Roman Iron Age ceramics in white. A lighter grey colour towards the end of the column indicates a paucity of dates. The diagram is adapted from Paper IV with slight modication. On right, the periodization of the Finnish Neolithic Stone Age after Carpelan (2002). Abbreviations: Finland (after Asplund 1997; 2004; Carpelan 1999; Edgren 1999; Lavento 2001; Pesonen 2004; The beginning of Corded Ware is relocated, see text) CW1 = Early Comb Ware (a.k.a. Sperrings), SÄR1 = Säräisniemi 1 Ware, EAW = Early Asbestos Ware, JÄK = Jäkärlä Ware, CW2 = Typical Comb Ware, CW3 = Late Comb Ware, KIE = Kierikki Ware, PÖL JYS = Pöljä and Jysmä Wares, PYH = Pyheensilta Ware, Corded W = Corded Ware, MZC = Middle Zone Ceramics/Pottery, KIU = Kiukainen ware, TXT = Textile pottery, PAI = Paimio ware, MORBY = Morby Ware, SÄR2 = Sär[äisniemi]2 ceramics. Republic of Karelia Left columns after Kosmenko (2004): Sperrings = Early Comb Ware, PIT-COMB = Pit- Comb Ware, COMB-PIT = Comb-Pit Ceramics, RHOMB-PIT = Rhombic Pit Ceramics, CLASSIC = Classic ceramics (organic and asbestos tempered), TXT = Net ( Textile ) Ware, EIA = Early Iron Age, Net Ware-Ananino mixed type ceramics Right columns after Zhul nikov (1999): C-P & R-P = Comb-Pit and Rhomb-Pit Ceramics, VOI = Voinavolok XXVII ceramics, TXT = Textile pottery In the Ancient Lake Saimaa area, the Karelian Isthmus, and south-eastern Finland, the rst part of the Neolithic Stone Age is considered to have been dominated by people using Comb Wares. During the few centuries before and after 4000 cal BC, Early Asbestos Ware was used alongside Early Comb Ware and CW2 (Pesonen 1996a; 2001). The centre of the distribution of Early Asbestos Ware lies in the Ancient Lake Saimaa area, but it is also found on the Karelian Isthmus and in south-eastern Finland, albeit to a lesser degree (Pesonen 1999). At the time when CW2 was in use, the material culture of the area was fairly homogenous. There were, of course, some regional characteristics in ceramics, such as the above-mentioned Early Asbestos Ware and the occasional use of asbestos temper in CW2 in the Lake Saimaa area. Another regional characteristic is the appearance of more eastern-oriented pottery types i.e., Pit-Comb Ware, Comb-Pit Ware, and Rhomb-Pit Ware in the eastern Vuoksi River catchment and on the Karelian Isthmus (e.g. Luho 1948: 48; Huurre 1990: 28 32; Edgren 1993: 44 47, 104 107; Pesonen 1999b). However, CW2 dominates the overall picture of ceramics (Carpelan 1999). The regional differences in ceramics emerge most strikingly after the middle of the Middle Neolithic (ca. 3500 3400 cal BC), when the Lake Saimaa area came to be dominated by asbestos tempered wares (Kierikki Pöljä Jysmä). A bit later, the differences in material culture were further amplied when the Corded Ware Culture spread to the shores of the Gulf of Finland and to the Karelian Isthmus. In south-western Finland, Pyheensilta Ware continued the Comb Ware tradition alongside Corded 16

Ware (e.g. Carpelan 1999). In fact, the heterogeneity of ceramics during latter part of Middle Neolithic is well represented even at individual sites, examples including the material found at the Puumala Kärmelahti (Katiskoski 2002) and Meskäärtty (Paper III) dwellings sites. The beginning of the Corded Ware Culture is traditionally dated in Finland to ca. 3200 cal BC (see Carpelan 1999; 2004; Carpelan & Parpola 2001). In light of current Central European research this Finnish date, based on conventional radiocarbon dates on charcoal (see Carpelan 2004), is far too early. 2 In Central Europe, the oldest reliable radiocarbon dates date to 3000 2900 cal BC, while the majority date to after 2800 cal BC (Furholt 2003; 2008; Wodarczak 2009). The oldest radiocarbon date of Corded Ware on crust adhering to pottery available from the research area dates 2 Not until the nal stage of this work I did understand the weakness of the Finnish chronology concerning the beginning of the Corded Ware Culture. The early beginning date of 3200 cal BC is used in the articles, but in the chronological scheme (Fig. 5) published here, I have corrected the beginning to be more in line with the oldest reliable radiocarbon dates obtained from Central Europe. I thank Dr. Volker Heyn for drawing my attention to this peculiarity in Finnish Stone Age chronology 17

to 2900 2600 cal BC (Hela 468, 4130±60 BP, Saarnisto 2003). This means that the Corded Ware culture was present at the Kaukola Lavamäki site on the Karelian Isthmus, Russia, already during the very early phase. In the papers, I have used the traditional, i.e., incorrect, dating of the Corded Ware Culture. This changes some of the interpretations (especially those in Paper III), but viewing the whole chronological schema, I do not consider this to be a major problem. The number of available recent AMS radiocarbon dates on crust adhering to ceramics or bones is so diminutive that the chronology is still for the most part anchored to the results of other available dating methods. In this, CW2 is an exception (see Pesonen 2004). The majority of given dates are based on conventional radiocarbon dating and shoreline displacement dating, which are not as accurate as modern radiocarbon dates. In this respect, the dates of different events in the archaeological material used in this study are to be seen as suggestive rather than absolute. The accuracy of different chronologies is discussed below in Chapters 5.1 5.2. During the later part of the Neolithic (ca. 3400 1800 cal BC), the ceramics prole of the research area is diverse. It can be said that the material culture of the late Middle Neolithic and Late Neolithic is generally not very well known. This is especially true of material dating from 2500 to 2000 cal BC. The heterogeneity and regional differentiation of ceramic styles renders typological studies very challenging. One of the main difculties in late Middle Neolithic and Late Neolithic pottery studies is derives from the research history. The non-asbestos-tempered ceramic types in the Finnish Baltic coastal zone, i.e., Pyheensilta Ware and Kiukainen Ware, are dened on the basis of the material of a single or a small number of dwelling sites located in southwestern Finland (Meinander 1940; 1954a). Later, attempts were made to use these pottery types in a similar way as the older ceramics types. However, the ceramics dating to the latter part of the Neolithic found in other areas, e.g., south-eastern Finland and the Karelian Isthmus, do not t very well into the mould created in south-western Finland (Paper III, see more in Chapter 5.1, see also Carpelan 1999: 260). In the Ancient Lake Saimaa area, the later part of the Neolithic or simply the period after CW2 is dominated by asbestos and organic-tempered wares, of which the Pöljä-Jysmä ceramic tradition is thought to persist until the end of the Stone Age. Such wares are also found to a minor extent on the Gulf of Finland coast as well as on the Karelian Isthmus, but there the material of the later part of the Neolithic is poorly known archaeologically. In south-eastern Finland, the latter half of the Neolithic is characterized by cultural inuences from every direction (see Miettinen, T. 1998; Mökkönen & Seitsonen 2007; Paper III). On the Karelian Isthmus, the period after CW2 is also poorly known (Paper IV). The asbestos wares dominating the material in the Lake Saimaa area did not spread intensively to the Karelian Isthmus (Huurre 2003: 198). In addition, it seems possible that CW3 ceramics remained in use longer on the Karelian Isthmus and in coastal south-eastern Finland than it did in other parts of Finland (Paper IV). Likewise, in the same area, the CW3 tradition and the Corded Ware tradition presumably interacted more intensively and earlier than in south-western Finland (Carpelan et al. 2008, Paper III). 18

3 HOUSEPITS ARCHAEOLOGICAL REMAINS OF PITHOUSES 3.1 About terminology A housepit represents the remains of a dwelling structure, the oor of which was dug below the ground surface. In Finnish archaeological studies, the Finnish term for housepit asumuspainanne (Fi.) became established during the 1990 s. The previously used terminology varied, and the following Finnish terms were frequently used: Madeneva-tyypin kodanpohja Engl. Madeneva-type hut oor (Meinander 1976; see also Karjalainen 1996a), asuinpainanne Engl. dwelling depression (Karjalainen 1996a; Kotivuori 1996a), asumuksenpohja Engl. base of a dwelling (e.g. Moisanen 1991) and kodansija/kodanpohja Engl. hut site/hut oor (Kotivuori 1993; 1996b; Koivunen 1996). A standard English terminology has yet to be agreed upon. Texts written in English by Finnish archaeologists have used terms such as semisubterranean/semi-subterranean house (remain)/dwelling (Nuñez & Uino 1997; Núñez & Okkonen 1999; 2005; Núñez 2004; 2009a; Pesonen 2002), dwelling depression (Katiskoski 2002; Costopoulos 2005; Vaneeckhout 2008a 2009b), pithouse (Karjalainen 1996b, Pesonen 2006), pit house (Karjalainen 1999), house pit (Paper I; Halinen et al. 2002; Kankaanpää 2002; Ojanlatva & Alakärppä 2002; Rankama 2002), and house depression (Leskinen 2002). In other Nordic countries, the terminology is also variable. In Sweden, terms like Stone Age house remain (Lundberg 1985) and pit house (Loefer & Westfal 1985) have been used. Some archaeologists have used paraphrases such as pit construction with more or less pronounced embankments (Liedgren 1998). In recent dissertations, the use of semi-subterranean house (remain) seems to be well established (Lundberg 1997; Norberg 2008). In studies on Norwegian housepits written in English, the most frequently-used terms have been pit house/pit-house (Engelstad 1988; 1990; Hood 1995) and simply house (Helskog 1984; Renouf 1984; Hesjedal et al. 1996; Hodgetts 2010; Niemi 2010). In recent papers, however, the term semi-subterranean house has been preferred (Bjerck 1991; Engelstad 1991; Ramstad et al. 2005; Skandfer 2009; Hood & Helama 2010). In this work, I have used the simple term housepit (excluding Paper I, in which the term is written separately), which I have adopted from North American archaeological literature. According to Canadian archaeologist Brian Hayden (1997), the term pithouse refers to a whole pithouse and its standing wooden structures. When the structures have collapsed and only a pit remains, the preferred term is housepit. I 19

liked the simplicity of this terminology. Accordingly, in the terminology used in this work a housepit is an archaeological remain of a pithouse. The appearance of housepits is rather polymorphic. In the Finnish archaeological record of housepits, the depth of the depressions visible on the surface varies usually between 20 and 60 cm, although deeper examples do exist. The shapes of the depressions vary from round to elliptical. Some are clearly rectangular. There is great variation in the size of the housepits. The smallest ones are less than 4 metres in diameter while the largest are tens of metres in length. The longest multi-room houses, also referred to as terrace houses, measure over 50 metres from end to end (Núñez & Okkonen 2005, see Paper III). The primary data used in this study originates from an area where no other Stone Age semi-subterranean dwelling structures than housepits exist. However, housepits are not the only Stone Age semi-subterranean dwellings/dwelling-like structures in Finland. The others are known as settlement embankments (Fi. asuinpaikkavallit) and Giant s Churches (Fi. jätinkirkot). Although these are usually larger than housepits, there is still some overlap in size. In addition to size, the grouping of semi-subterranean structures is based on the soil type on which the structures have been erected. Housepits lie on sandy soils, settlement embankments on very stony moraines or boulder elds, and Giant s Churches mainly on boulder elds (Okkonen 2003: 28 30, 101 103; Pesonen 2002: 13 14; see also Schultz 2009). As I see it, this grouping is partly articial, and all the structures are manifestations of a single tradition of building semi-subterranean structures (Paper III). 3 3.2 Pithouses in enthnographic sources This section presents some general correlations between different aspects of housing and mobility. Semi-subterranean dwelling structures known from archaeological or ethnographical sources have a distribution that largely coincides with the arcticsubarctic climate zone and includes boreal coniferous forest, temperate mixed forest, and regions such as mountains and deserts, where the climate is cold at least part of the year (Fig. 6). 3 Traditionally, Giant s Churches have been considered to be structures used for other purposes than human habitation. A wide range of propositions includes natural formations (Ailio 1923), hunter s campsites (Forss 1981; Edgren 1993: 105), storage structures (Okkonen & Ikäheimo 1993; Koivunen 1997), fortresses (Sipilä & Lahelma 2006) and ceremonial structures (Okkonen 2003; see also Okkonen & Ridderstad 2009). A more comprehensive list of suggested intepretations has been published, for example, in Okkonen (2003), Núñez & Okkonen (2005), Sipilä & Lahelma (2006) and Núñez (2009b). The great variety of interpretations is obviously due to the minute number of excavations. Recently, excavations at the Giant s Church at the Honkobackaharju site, Middle Ostrobothnia, revealed well develped cultural layers and high phosphate values inside the structure, as well as marks of wooden structures and at least two building stages (Schulz 2009). The number of nds including lithic artefacts of quartz and other rocks, and asbestos tempered pottery resembling Pöljä Ware is low as compared to material typically found in contemporary dwelling sites (ibid.). At present, it seems that Giant s Churces are to be interpreted as structures with various possible functions, including a function as a dwelling (Okkonen 2009; Schulz 2009). 20

Fig. 6. The worldwide distribution of pithouses in archaeological and ethnographic sources. The original map published by Gilman (1987: Figure 1.) depicts the North Fennoscandian distribution incorrectly. It is slightly modied here. The map also lacks the housepits from the North American Central and Eastern Arctic and Subarctic as well as from the Northern Woodland on the East Coast known from ethnographic and archaeological records (e.g. Fitzhugh 1972; Trigger 1978; Helm 1981; Damas 1984; Hood 1995). In the Ethnographic Atlas compiled by Georg P. Murdock (1967), 84 groups have subterranean pit structures as their primary or secondary dwellings. According to the data, three conditions are nearly always connected with pithouse habitation: (1) a non-tropical, i.e., cool or cold climate during the season of use, (2) reliance on stored food while the dwelling is inhabited, and (3) permanent winter sites or a more sedentary settlement pattern. (Gilman 1987.) Lewis R. Binford (1990) has studied the relations between mobility, housing, and subsistence on the basis of ethnographic sources. The Ethnographic Atlas forms the core of his data, although he has modied it by adding new variables. According to Binford, there is a fairly strong correlation between rectangular ground plans and sedentary habitation (Fig. 7). Likewise, the distribution of semi-subterranean structures is emphasized among semi-sedentary and sedentary hunter-gatherers. In the ethnographic sources, pithouses are used primarily as winter houses (Gilman 1987). This means that ethnographic data point to the presence of alternative summer housing, either at separate summer sites or at the same year-round dwelling site where the pithouses were built (see Binford 1990; Gilman 1987). There are only a few known examples of summer habitation in pithouses around the North Pacic Ocean, where the climate is cool to cold all year round. They include the Koryak (Kamchatka Peninsula, Russia) as well as the Aleut and the Alutiiq (Prince William Sound, the Alaska and Kenai peninsulas, Kodiak, Afognak, and the Aleutian Islands, 21

100 % 90 % 80 % 70 % 60 % 50 % 40 % 30 % 20 % 10 % 0 % Sedentary Semi-sedentary Semi-nomadic Nomadic 100 % 90 % 80 % 70 % 60 % 50 % 40 % 30 % 20 % 10 % 0 % Semi-subterranean (n 55) Ground surface (n 119) Fig. 7. The relations between hunter-gatherer mobility and primary housing in ethnographic data according to Binford (1990). A (above) the relation between ground plan and mobility. B the distribution of ground surface and semi-subterranean dwelling structures in relation to mobility. Complex ground plans are polygonal or quadrangular and may include an interior court. Alaska). The Koryak have distinct winter and summer sites, the former on the coast and the latter upriver inland (Jochelson 1975). In the Aleutian region, the large communal pithouses ( barabaras ) were mainly used as winter dwellings, although they were, at least during late 18 th century, occasionally also used during the summer autumn (Hoffman 1999). I believe that prehistoric pithouses were primarily winter houses. However, the ethnographically based notion of pithouses as nearly exclusively winter houses does not necessarily hold equally well for all prehistoric cases. As concerns the Stone Age housepits considered in this study, it is probable that the conditions preventing summer habitation were not valid in the latter part of the Middle Neolithic (Paper IV). At that time, pithouses were erected in locations that suggest other than winter-only 22

Fig. 8. Map of Fennoscandian areas cited in text. occupation. Therefore, I believe that there is no need to categorically deny the possibility that some Stone Age housepits are the remains of dwellings inhabited yearround (vide Chapter 3.5). 3.3 Archaeological research on housepits in Fennoscandia This chapter summarizes the research history and chronological trends relating to Neolithic housepits in Fennoscandia (Fig. 8). In contrast to Finland, large numbers of Mesolithic housepits are known from the other Fennoscandian countries. Still, over all of Fennoscandia the most active period of pithouse building was the Neolithic Stone Age, the epoch with which this study is mainly concerned. This section focuses on chronological changes in size, shape, clustering, site location, and materials associated with housepits. Housepits on the Karelian Isthmus, Russia, are not discussed in this section. They are presented in Papers II and IV. 23

3.3.1 Finland Housepits are a rather recently discovered subject in Finnish archaeology. Prior to the 1980 s, other kinds of structures dominated the picture of Stone Age housing. The Räisälä hut, a conical hut with an upright frame structure supporting the roof beams, was excavated and interpreted by Sakari Pälsi (1918). It became the prototype of a Finnish Stone Age dwelling for decades. Recently, Pälsi s interpretation has been questioned, and based on a re-evaluation of Pälsi s eld documentation the remains have been re-interpreted as half of a pithouse associated with asbestos tempered Pöljä Ware (Seitsonen 2006). During 1950 s and 1960 s, several remains of semi-subterranean dwellings were excavated in different parts of Finland. The results were introduced in the Nordisk arkeologimötet conference in Helsinki in 1967. In a paper based on his presentation, Meinander (1976) introduced a new Stone Age dwelling type referred to as the Madeneva-type hut oor. Nowadays this term is no longer used and the Madenevatype huts are counted as housepits. The 1990 s was the decade when Finnish housepits were discovered on a large scale. In the municipality of Tervola in Southern Lapland, a number of housepits were excavated already during late 1970 s and 1980 s but the results were never published, though they have since been utilized by Hannu Kotivuori (1993; see also 2002). During 1990 s, large numbers of new housepit sites were found in surveys in the Finnish inland region, especially in the Lake Saimaa area, and on Stone Age shorelines of the Gulf of Bothnia in western northwestern Finland. I assume that the housepit distribution maps published by Pesonen (2002a) are still quite well up to date. During the past ve years, the largest number of new housepits (as compared to previously known numbers) has been found in south-eastern Finland, which used to be an area with a sparse housepit distribution (Fig. 9, see also Chapter 2.2.1). As I see it, the boom in housepit research declined after the two-day seminar held at the Finnish National Museum in 2000 and the book Huts and Houses, Stone Age and Early Metal Period buildings in Finland (Ranta (ed.) 2002) in which the presentations were published. To be sure, the decline was not due to the seminar but to changes in the nancial arrangements that had sustained the earlier research. 4 In consequence, the book became an unintentional intermediate summary of the most active period of housepit research so far. With a few exceptions, the articles in the book are still focused on empirical description concerning structural details, dating, and materials associated with housepits. Two articles in the book that described the rst concrete remains of Stone Age timber frame constructions in Finland (Katiskoski 2002; Leskinen 2002) closed the debate on the possibility of timber constructions in Finnish housepits (see Meinander 1976; Karjalainen 1996a). Most of the papers in the publication deal with individual housepits or housepit sites representing the nds and structural remains. At the time, research still largely concentrated on the basics of understanding the dating and the variability of materials associated with housepits. 4 The decline in housepit research was mainly due to the withdrawal of the government employment programmes, which were the main source of nancial support for the housepit excavations in the Lake Saimaa area and Northern Ostrobothnia during the 1990 s. 24

Fig. 9. The distribution of housepit sites in Finland up to the year 2002. The main difference as compared to the present situation is represented by the new housepit sites discovered in coastal south-eastern Finland, which was totally blank prior to 2006. (Picture from Pesonen 2002: Fig. 1). By the year 2000, there were about 3500 known housepits distributed over approximately 650 dwelling sites. At that time, there were 117 housepits of which at least one quarter had been excavated. The greatest numbers of housepits were found on Stone Age shorelines in Ostrobothnia and southern Lapland, where several large villagelike housepit clusters with tens of housepits are located. In contrast, in the Finnish inland region there are less than twenty housepit sites exceeding ten housepits per site. (Pesonen 2002.) Most of the housepits date to the Middle Neolithic period (ca. 4000 2300 cal BC) and are associated with CW2, CW3 and asbestos tempered wares of the Kierikki and Pöljä type. A notably smaller number of housepits are associated with organic-tempered, poorly studied Stone Age ceramics (see Paper III) and Early Metal Period asbestos-tempered wares (Sär 2 group) (Paper I; Pesonen 2002). In Ostrobothnia, the largest number of housepits and housepit sites, including village-like settlements sometimes consisting of clusters of over a hundred housepits, date to a period extending less than a thousand years before and after 3000 cal BC (Núñez & Okkonen 1999; Pesonen 2002; Okkonen 2003: 169 172; Vaneeckhout 2009a b). In coastal Ostrobothnia, a rapid increase in the number of housepit sites began after 3500 cal BC, and an even slightly more drastic decrease occurred after 2500 cal BC (Okkonen 2003: 169 172). Parallel to the appearance of village-like settlements ran a shift towards larger houses. In Finland, the larger and deeper housepits with surrounding embankments and often with entrances or antechambers visible on the surface predominantly date from ca. 3500 cal BC to the beginning of the Bronze Age ca. 1800 cal BC (Papers I, 25

III and IV; Halinen et al. 2002; 2003; Katiskoski 2002; Ojanlatva & Alakärppä 2002; Núñez & Okkonen 2005). The larger housepits, including multi-room examples, date predominantly later than 3300 cal BC (Okkonen 2003; Núñez & Okkonen 2005; Paper IV). In the Yli-Ii area, Northern Ostrobothnia, a nearly exponential growth of the largest housepits began during the late CW2 period and peaked about 1500 years later, ca. 2200 2000 cal BC (Vaneeckhout 2008a, b; 2009a, c). The growth of the largest housepits is frequently more to do with an increase in length than in width (Núñez & Okkonen 2005; Paper IV). Despite the fact that a similar increasing trend in housepit sizes is observed also in the Lake Saimaa area (Paper I) and on the Karelian Isthmus, Russia (Paper IV), the largest housepits in these areas are not as numerous nor as large as in Ostrobothnia. There seems to be a general trend in Finland that the largest housepits date from the last quarter of the 4 th millennium onwards. In some areas, however, this notion does not hold true. On northernmost shore of the Gulf of Bothnia, in Tervola municipality, the largest housepits are associated with CW2 while the younger housepits with Pöljä Ware are of smaller size (Kotivuori 2002: 165). The correlation between the housepit and the original pithouse is not completely straightforward. In the opening paper of Huts and Houses, Pesonen (2002a: 29 31) states that all the excavated housepits have proven to be quadrangular, even though the ultimate shape of the depression itself was roundish. He seems to assume that all housepits from CW2 onwards have been rectangular. Pesonen argues that in the cases when the locations of nds have been pinpointed exactly, the foundation of the house has turned out to be rectangular. According to Pesonen, not one of the excavated housepits has turned out to be round or oval, only quadrangular. In a table summarizing the excavated housepits, Pesonen (2002: Table 5) divided the ground plans into rectangular, quadrangular, and unidentied. I read this as a binary division into housepits with (the rectangular and square examples) and without observed corners. The examples with corners are predominantly associated with Kierikki-Pöljä asbestos wares and with material dating to the Early Metal Period. The excavated housepits with an unidentied ground plan (i.e., the cornerless ones ) are predominantly associated with CW2 and CW3, although some cornered examples of similar age are also known. I believe that there are certain links between the visible shape of a housepit and the original pithouse. The chronological difference in housepits with or without corners cannot be merely a function of the excavator s ability to observe the shape of the dwelling. Nor do I suppose that that the presence or absence of corners derives simply from differences in post-depositional processes. In my view, it is quite possible that a real change in the way pithouses were constructed took place during the CW2 period (4000 3400 cal BC), the cornerless types dominating the older phase and the ones with corners becoming predominant in the younger phase after CW2. It is true that observations made on the surface do not automatically correspond to the shape of the original pithouse as revealed by excavations. However, measurements using the housepits dip points as datums have turned out to correspond well with the original wall lines (e.g. Helskog 1984: 63 64; Pesonen 2002: 27). Therefore, 26

Fig. 10. Reconstruction of the Kärmelahti pithouse. The dwelling 8 x 7-7.5 metres in size dates to 3200 2800 cal BC. Drawing by Mikko Rautala from Katiskoski (2002). the size of the original pithouse can be approximately measured on the surface of the housepit. Regarding the relation between surface observations of a housepit and the original shape of the pithouse, I am not as pessimistic as Pesonen (2002: 29). In some cases, the shape of the housepit is clearly rectangular/quadrangular with slightly rounded corners, and in some other cases it is clearly round or oval. I am convinced that there is a reasonable correlation between surface observations of housepits and the original shape of the pithouse, and that the correlation has a temporal dimension as well. The largest housepits and those with a clearly rectangular or oblong shape in contrast to older ones of more elliptical or roundish shape date largely later than CW2. The size of the housepit began to increase already during the CW2 phase, mainly in length and occasionally also in depth (Paper I; Núñez & Okkonen 2005; Vaneeckhout 2008a b; 2009 a, c; Paper IV). There is currently no concrete evidence of log frames earlier than ca. 3400 3000 cal BC, the two oldest examples being Puumala Kärmelahti (Fig. 10, Katiskoski 2002) and Yli-Ii Purkajansuo/Korvala (Fig. 11, Shultz 2000). I believe it is too early to conclude that all the housepits that date later than the CW2 phase (4000 3400 cal BC) have been rectangular and constructed on a log frame foundation. It is noteworthy that a rectangular shape does not automatically indicate a log house (Vaara 2000; see also Chapter 3.3.2), and other types of structures are still fully conceivable. As an example, the structure of the Martinniemi housepit, located in Kerimäki parish in the Lake Saimaa area and associated with Kierikki and Pöljä wares, is interpreted to be composed only of a ridged roof constructed of poles leaning against each other (Halinen et al. 2002). This type of house or hut does not have any form of upright wall. So far, only few studies have dealt with the environmental location of housepit sites. In the Lake Saimaa case study area (Paper I) the location of housepits changed from highly sheltered heads of bays to windy islands and capes during CW2 phase. Some of the new unsheltered housepit sites were used repeatedly over long periods 27

by people using CW2, CW3, Kierikki Ware, and Pöljä Ware. This led to the formation of village-like concentrations of housepits, the degree of contemporaneity of which, however, is not currently known. In the Lake Saimaa area (Paper I), an aquatic orientation coupled with changes in site location is not particularly clearly pronounced. Although the change in terms of larger environmental zones is very small, the shift of site locations to capes and islands during the latter part of CW2 is conspicuous. These sites are aquatically oriented according to the specic parameters that apply to Lake Saimaa. In other words, an aquatic orientation cannot be as clearly distinguishable in a topographically mosaic-like inland lake system as on the seacoast or in an area with large-scale topographical features. In the Kaukola Räisälä case study area on the Karelian Isthmus, Russia, a similar change one could say a colonization of the archipelago by pithouses is also as- Fig. 11. A reconstruction of the Purkajansuo/Korvala terrace house at the Kierikki Stone Age Centre (Fi Kierikkikeskus). The dwelling is dated to 3360 2930 cal BC based on a radiocarbon date on charred crust adhering to Pöljä Ware found inside the dwelling (Schultz 2000). For more information on the reconstruction, see Vaara (2000). Drawing by Antti Heikkilä/Kierikki Centre, photos by the author. 28

sociated with the CW2 phase (Paper IV). Comparable changes in the environmental location of housepit sites during the CW2 phase have not been observed in other Finnish studies. However, the number of studies concerning the environmental location of housepits is very limited, and therefore it is too early to evaluate whether changes similar to those in the Ancient Lake Saimaa and Ancient Lake Ladoga areas have taken place also in other parts of Finland, in the inland lake district or the coastal regions. It is also possible, if not even probable, that there were different kinds of adaptations taking place in different environments. In most of the studies carried out in coastal Ostrobothnia, where housepits often occur in village-like concentrations, such concentrations have been located close to river mouths (e.g. Halinen 1997; Núñez & Okkonen 1999; Ikäheimo 2002; Vaneeckhout 2008a b; 2009). There are rare examples of housepit sites located in the archipelago. In the estuary of the Oulujoki River, the housepit sites dating to ca. 2500 1800 cal BC are located on an island outside the river mouth (Ikäheimo 2002). As an exception, the large stone enclosures called Giant s Churches, which often form complexes accompanied by housepits, cairns, and heaps of re-cracked stones, are usually located by the sea on islands, necks of land, and capes (Forss 1996; Koivunen 1997). The sites with Giant s churches (ca. 3000 2000 cal BC according to Okkonen 2003: 123) are located in decidedly marine environments as compared to housepit clusters located in the river estuaries. In the northernmost part of the Gulf of Bothnia, the housepits in the Stone Age estuary of the River Kemijoki in Tervola municipality in southern Lapland have been located both on large islands and on the shores of an inlet. The sites located on the large bay outside the estuary were directly exposed to open water. There appear to have been no changes regarding site location during the most active occupation phase with pithouses ca. 4000 2500 cal BC (Kotivuori 1993; 2002). 3.3.2 Sweden In Sweden, housepits are found mainly in Norrland. There, they are distributed in both inland and coastal areas (see Lundberg 1997; Liedgren 1998). The rst housepits (Swe. boplatsvall or skärvstenvall) were recognized and studied already in the 1920 s, but the actual housepit boom followed surveys carried out during the 1980 s (Liedgren 1998; Norberg 2008: 16-18). In order to learn how to identify a housepit, the Swedish archaeologist responsible for the surveys embarked on a eld trip to Finnish housepit sites by the River Kemijoki in the Tervola area (Norberg 2008: 47). Swedish housepits date from the Mesolithic Period to the end of the Stone Age. In Norrbotten, the northernmost part of Upper Norrland, the number of housepits increases for the rst time during the Late Mesolithic (at sites around 90 m a.s.l.). However, the most impressive increase occurred during the Middle and Late Neolithic periods, when the number of housepits peaked at 60 45 m a.s.l., corresponding roughly to 3000 2300 cal BC. This increase in numbers is accompanied by a concurrent increase in housepit size. During this period, the average size of a housepit was 29

28.6 m 2, which is remarkably high as compared to the average sizes of Mesolithic housepits (less than 10 m 2 ) and Early and late Neolithic/Early Bronze Age housepits (around 15 m 2 ). (Norberg 2008.) During the peak period between 60 and 40 m a.s.l., housepit sizes exhibit a growing trend with the largest housepits being found at the lowest elevations. Average housepit size at 50 45 m a.s.l (ca. 2800 2300 cal BC) is over 30 m 2, and at 45 40 m a.s.l. (ca. 2300 1800 cal BC) it is even larger, over 40 m 2 (Norberg 2008: 56 57). It should be noted, however, that by the time that housepit size peaked, the total number of housepits had already drastically decreased. In Sweden, the Neolithic pithouses are thought to have been constructed over a framework made of horizontal logs with or without central posts (Halén 1994: 96 97; Norberg 2008: 87 88). Another proposed construction is that of merely a simple roof structure made of slanting poles resting against each other at the ridge (Lundberg 1997: 107 108). The great variability in house building traditions is testied to by two recently excavated sites in Middle Norrland, namely Bjästamon and Kornsjövägen in Örnsköldsvik municipality. The distance between the sites is only 400 metres. The Bjästamon site, with an overall date of 2800 2100 cal BC, shows that dwellings with round and oblong oor plans as well as housepits and surface dwellings were in use concurrently (Holback 2007; Runeson 2007). Excavations at the site have revealed an interesting two-room house. The northern part of the house is interpreted to have been constructed on a horizontal timber framework, while the wall of the southern room consisted of an upright post-and-wattle structure that was covered with earth on the outside (Fig. 12). The northern room with four hearths is interpreted as the residential part and the southern room with a grinding stone as storage space. Between the rooms there was a hallway, where a 5 x 2 x 0.6 metre storage pit was located beneath the oor level. The house measured ca. 24 x 6.5 metres from wall to wall and dated to 2400 2200 cal BC. (Holback 2007.) The second site, Kornsjövägen, indicates that rounded and oblong structures utilizing upright posts were also being built around 3000 2300 cal BC. Here, the housepit-like dwellings (Houses 1, 3 and 4) were partly dug into the slope. In the two largest houses at the site, House 1 (15 x 4.5 metres) and House 4 (13 x 5 metres), a number of postholes indicated a frame structure based on upright posts. In House 1, a line consisting of a trench and postholes shows the location of the wall (or possibly of two non-contemporaneous walls) in the northern long façade and at the ends. The trench shows that the house was not rectangular but rounded at the corners. House 1 was surrounded by 1 1.5 metre wide embankments, whereas House 4 did not have any clear embankment. (Lindqvist 2007.) These two sites show that different building traditions co-existed in Middle Norrland during the Neolithic: one with housepits founded on sandy moraines, a second one with housepit-like dwellings built on stony moraine hillsides, and a third one with surface dwellings accompanying the former two. In addition, both sites, Bjästamon and Kornsjövägen, yielded similar pottery referred to as the 3 rd group (Swe. tredje gruppen ). 3 rd group pottery possesses traits originating predominantly from 30

Fig. 12. Interpretation of a pithouse structure at the Bjästamon site. The storage area (left) and the residential part (right) have been connected by a hallway with underground storage. Drawing by Franciska Sieurin-Lönnqvist from Holback 2007: Figur 123. Corded Ware, moderately from (East Swedish) Pitted Ware, and slightly also from the Comb Ware tradition (i.e., Pyheensilta Ware) and Kiukainen Ware (Lindholm et al. 2007). Other ceramics found at the sites include Fagervik III/IV type pottery and some unidentied asbestos tempered pottery from one of the housepits at the Bjästamon site. In Norrbotten, Upper Norrland, there is a change in the way housepits were located in the landscape. During the Mesolithic Period, housepit sites were located in the bays of river estuaries. During the Neolithic Period, which begins in the area ca. 4200 cal BC, a shift occurred in the way the pithouses were located. At elevations between 59 40 m asl., which corresponds to ca. 3000 1800 cal BC, the location of pithouses was not tightly bound to river estuaries and pithouses were erected in a more maritime environment, often on islands, capes or large bays in the outer archipelago (Norberg 2008: 161 165, 173). At the same time, housepits were erected in locations that were exposed to the east, i.e., towards the sea (Norberg 2008: 61, 127). The shift in site location was simultaneous with other changes, namely the increase of housepit size combined with more solid evidence of the sedentary nature of the occupation, and, in the nal Neolithic, with a decrease in the number of sites (see Chapter 3.4.2). In Middle Norrland, the Bjästamon and Kornsjövägen sites exhibit a maritime orientation, manifested in their location by larger open-water areas and in a rich osteological fauna dominated by seals (Lindholm & Runeson 2007; Lundberg 2007; Olson et al. 2007). In this region, a maritime orientation is clearly visible in sites dating to 2800 2100 cal BC (Gustafsson 2007: 320). Similar trends have been noticed further south, in southern Norrland and northern Uppland, where the focus of Neolithic settlement rst shifted from the river mouths towards the outer archipelago between 3400 3000 cal BC, and then moved entirely to the archipelago between 3000 2300 31

cal BC (Björk 2003; see also Hallgren 2010). The rst period was associated with an increase both in the size and in the number of dwelling sites, and the latter period with a contrary development. It seems that a tendency to settle in more marine areas commenced in the area ca. 3400/3000 cal BC. 3.3.3 Norway Norwegian archaeology has the longest history of research on housepits in the Nordic countries. Although the original excavations of the eponymous sites for the housepit types Karlebotn, Nyelv, Gressbakken, and Mortensnes took place already between the 1930 s and 1960 s, the chronological typology of housepits was rst published somewhat later by Simonsen (1976). The classication of housepits, however, was generated already during the 1960 s on the basis of extensive surveys and excavations (Engelstad 1988: 71). There are thousands of housepits in coastal northern Norway in the area between Nordland Troms in the west and Finnmark in the east (Olsen 1994: 38 39; 1998). The oldest housepits date to the Mesolithic period. They are shallow depressions of moderate size, usually ca. 3 4 metres in diameter (Olsen 1998). Most of them are possibly foundations for tent-like structures, but in the Varangerfjord area there are also dwellings of similar size that were built of sod (Olsen 1994: 39). Besides northernmost Norway, Mesolithic housepits are also known from Nordland, south of Troms (Bjerck 1991). In Finnmark, northern Norway, the Younger or Late Stone Age is divided into three phases: Phase I 4500 3700 cal BC, Phase II 3700 3000 cal BC, and Phase III 3000 1800 cal BC (Olsen 1994: 52 59). During the Younger Stone Age, which nearly corresponds to the Finnish Neolithic Stone Age (5100 1800 cal BC), the sizes and shapes of housepits go through developments similar to those observed in other parts of Fennoscandia. During Phase I, the slightly larger so-called Karlebotn house type appeared. In the beginning, the larger rectangular houses with a ca. 12 20 m 2 oor area were quite uncommon, but during the 4 th millennium cal BC, they became predominant (Olsen 1994: 65 68). During Phase III, evolution led towards larger pithouses. In the Varangerfjord area, the somewhat larger Nyelv house type with two or three central hearths and a rectangular ground plan was in use during the transition between Phases II and III (3200 2500 cal BC). During 3200 3000 cal BC, both the Karlebotn and the Nyelv house types were concurrently in use (Olsen 1994: 58 59, 71). The trend toward larger houses culminated in Finnmark with the introduction of the Gressbakken house type, which was both larger and deeper than the previous pithouses (Olsen 1994: 72 82). At this time, the number of houses was at its highest (Helskog 1984: 53). In Gressbakken houses, starting from ca. 2500 cal BC, the rectangular ground plan was typically around 30 m 2 in size, although larger examples are also known (Helskog 1984: 63 65; Engelstad 1988: 72; Olsen 1998: 186). The period with larger pithouses was rather short. According to Schanche (1995), most of the Gressbakken houses date to 2200 1800 cal BC. 32

Following the decline of the Gressbakken house type, the size and numbers of dwellings declined as well. The next new house type, with a smaller square ground plan and a foundation dug deep into ground, is called the Mortensnes house. These date to ca. 1700 1200 cal BC. No pithouses are known from the following Kjelmøy Phase, corresponding to the 1st millennium BC (Olsen 1994: 113 116). In Norway, the debate on the number of pithouses in use contemporaneously has been lively. During the early period of research, the large housepit clusters were thought to represent large villages of contemporaneous pithouses (Gjessing 1942: 501). In the early studies of Gutorm Gjessing and Povl Simonsen, not all the housepits in large dwelling sites were thought to have been simultaneously in use (see Schanche 1995: 176). The number of supposed contemporaneous pithouses was, however, quite high. Simonsen (1979: 397 402) proposed, that 20 30 out of the over 80 housepits at the Gressbakken Nedre Vest site were in use at the same time. In his studies on North Norwegian housepit villages, Knut Helskog has claimed that Povl Simonsen merely repeated the assumptions made in earlier studies (Helskog 1984: 39). Based on radiocarbon dates, the housepits elevation in relation to shoreline displacement chronology, and a suggested lifespan of a pithouse made of sod, Helskog argues that there were no large villages in the Varangerfjord area. He suggests that there were only one to six contemporaneous pithouses in use at individual sites (Helskog 1984: 51 52). During the 1990 s, the discussion concerning the number of simultaneously occupied housepits continued. In her study on coastal housepit sites in Finnmark, Kjersti Schanche (1995) suggested that 10 20 houses at the largest dwelling sites with Gressbakken houses were occupied contemporaneously. By analysing the radiocarbon dates, she concluded that during the limited use period of Gressbakken houses around the Varangerfjord (2200 1800 cal BC), both smaller and larger housepit sites were in use contemporaneously. Schanche considered that the dwelling sites belonging to one society were divided in a hierarchical manner. Bryan Hood has also taken part in discussion. He seems to think that the correct number of simultaneously occupied houses is somewhere between the minimum (Helskog 1984) and the maximum (Schanche 1995) calculations (Hood 1995: 83). Although there are certain chronological trends to be seen in housepits, the comprehensive validity of the chronological typology of pithouses has been questioned. In a correspondence analysis of North Norwegian housepits, Erica Engelstad (1988) has shown that Karlebotn and Gressbakken are not typologically denable house types. According to Engelstad, there is great variation in housepits in North Norway and the traditional house types are to be found, as they are described, only in Varangerfjord area. A number of changes in the maritime orientation of the region s population took place during the Younger Stone Age. During Phase II (3700 3000 cal BC), there occurs a stronger shift towards maritime resources and an increase in the number of housepits on the coast as compared to Phase I (Olsen 1994: 68, 70 71). During Phase III (3000 1800 cal BC) the maritime orientation continues, the settlement pattern connected with Gressbakken houses is more sedentary (fully or semi-sedentary), and 33

during the end of the Phase III there is also more material from the inland (Olsen 1994: 73 75, 85; Hood 1995: 85). 3.3.4 The Republic of Karelia, Russia In the Republic of Karelia, which lies on the Russian side of the Finnish-Russian border between the Karelian Isthmus and the Kola Peninsula, the oldest housepits date to the Mesolithic Period. There, the history of research on housepits goes back a fair stretch, starting from the rst half of 20 th century (Zhul nikov 2003: Tab. 2). Currently, nearly 900 housepits are known from the area, and over 170 have been excavated (Zhul nikov 2003: 101). During the late Mesolithic and early Neolithic, there was a period when pithouses were not built in the Republic of Karelia. In the Lake Onega area, the rst rectangular pithouses with a frame of horizontal logs date approximately to 4500 cal BC. The pottery associated with the housepits is Pit-Comb Ware. During 5000 4000 cal BC, the groups using Pit-Comb Ware built winter houses with both round and rectangular ground plans. Summer houses built on the ground surface and inter-connected pithouses are also known from the same period (Zhul nikov 2003: 101 2). In the Republic of Karelia, as in Finland, pithouses reached their peak in numbers and size during the latter part of the Middle Neolithic. The increase in the number of housepits began after 4000 cal BC. In the course of the Eneolithic period (ca. 3300 2000/1800 cal BC), which is a transitional period between the Neolithic Stone Age and the Bronze Age, both the number and the size of housepits rst peaked (ca. 3300 2500 cal BC) and then decreased (ca. 2500 2000/1800 cal BC). During the latter part of Eneolithic, preferences for locating pithouses in the terrain changed and became more variable. The changes are thought to represent inuences that originate from contacts with southern societies practicing animal husbandry (i.e. Baltic Corded Ware and Fatyanovo). Before the beginning of the Bronze Age (from ca. 2000/1800 cal BC), housepits vanished from the archaeological record. (Zhul nikov 1999; 2003: 101 2). 3.4 Housepits and sedensm in archaeological research One of the most essential questions concerning Stone Age housepit research concerns the degree of sedentism, i.e., the length and seasons of occupation. Do housepits represent winter-only occupation, or were they occupied permanently year-round? Here, I follow a commonly used denition of sedentism that focuses on year-round habitation at one dwelling site by at least part of the population (Rafferty 1985: 115; Kelly 1992). It is possible to distinguish several degrees of sedentism that are more detailed than the traditional categorization of mobility using a four-step division ranging from fully sedentary to fully mobile nomadic hunter-gatherers. This means that there are grounds for suggesting that the data indicates a growing degree of sedentism at a certain point, although it cannot be stated whether the outcome is fully sedentary or something better described as semi-sedentary (see Kelly 1992). 34

There is no such thing as a single or denite indicator of sedentism in the archaeological record. Usually, arguments advocating the presence of sedentism are based on the co-occurrence of several factors supporting the interpretation. The most essential elements in the identication of sedentism are the nature of dwellings and dwelling sites (Rafferty 1985: 128; Marshall 2006: 157). Changes in settlement pattern, substantial houses, the presence of pottery, cemeteries, ceremonial structures (i.e. monumental structures), storage, heavy artefacts, luxury goods, agriculture, and osteological or palaeobotanical material indicating year-round habitation are typically included in the palette used for justifying an interpretation of sedentary settlement (Rafferty 1985; Kelly 1992; 2007: 152, 160; Marshall 2006). This section summarises the arguments presented for and against the sedentary nature of pithouse occupations in archaeological research. The focus is on Nordic research. 3.4.1 Finland The degree of sedentism has proven to be a difcult matter to deal with. In Finnish studies, housepits are most often associated with a sedentary or semi-sedentary settlement pattern. It should be noted that comments on the question of sedentism among Stone Age pithouse dwellers are usually imprecise. Typically, this question is bypassed by general statements about the nature of structures and nds that suggest a certain degree of sedentary occupation (e.g. Núñez & Okkonen 1999; Leskinen 2002; Ojanlatva & Alakärppä 2002; Okkonen 2003: 171). The statements usually do not specify what sort of settlement pattern is envisaged. As a basic default, housepit sites are considered to have been used, at least, for winter habitation. In the early phase of housepit research, the Middle Neolithic housepits located in the River Kemijoki estuary in Tervola, southern Lapland, were suggested to represent winter sites in the yearly cycle of semi-sedentary groups. The housepit sites on the seaside were interpreted as winter sites used for seal hunting, while the inland sites by the river and lakes were used during warmer seasons (Kotivuori 1993; 1996a b; 1998; for a similar settlement pattern see Siiriäinen 1981; 1987). In his licentiate thesis, Hannu Kotivuori (2002) articulates on behalf of a yearly cycle along the River Kemijoki. His interpretation of coastal housepit clusters as winter sites is based on several observations. In addition to semi-subterranean structures themselves, osteological data distinctly dominated by seals, the distribution of nds concentrated inside the structures, and the presence of thin, weakly developed cultural layers all point to winter-only/early spring habitation. In contrast, the inland sites extending nearly one hundred kilometres upstream are mostly devoid of housepits, and the diverse osteological data suggests habitation during the warmer seasons. However, such a strict dualism of seasonal sites is not evident in the archaeological material. The housepit sites and their surroundings have also produced nds that indicate late spring summer occupation, and therefore Kotivuori (2002: 167) suggests that only the most active members of the society left the winter base camps during warmer seasons in order to exploit the hunting and shing grounds located 35

upriver in the inland regions. Hence, Kotivuori s (2002) interpretation of an active yearly cycle does not actually mean a semi-sedentary but rather close to a fully sedentary settlement pattern. Other archaeologists have put forth similar opinions about the season of habitation as Kotivuori. In studies based on the Kauvonkangas housepit site in Tervola, Jarmo Kankaanpää (2002; 2003) also argues on behalf of seasonal habitation (most probably winter habitation) by a group that hunted seals and shed by the sea just outside the River Kemijoki estuary. He develops his argumentation using the same features as Kotivuori, and adds new ones. Kankaanpää (2002) points out that according to ethnographic parallels of housepit habitation in arctic and subarctic environments, such structures were usually winter dwellings. He notes that the site location by the sea is suitable for sealing but not for salmon shing. This points in the same direction as the other evidence, i.e., towards winter habitation. In studies dealing with other parts of Finland, housepits are thought to represent a more sedentary type of settlement than winter-only. It has been suggested that a relatively high degree of sedentism is reected by the occurrence of several indicators: (1) Housepits are substantial and massive structures, the building of which imposes a notable workload (Núñez & Okkonen 1999; 2005; Leskinen 2002; Núñez 2004). (2) Rectangular or quadrangular oor plans suggest a high degree of sedentariness (Karjalainen 1999). (3) The presence of large pots at the housepit sites is interpreted to mark summer habitation. This is based on the ideas that large pots were not suitable for transportation and that pottery making was a summer activity (Halinen 1997; Núñez & Okkonen 1999; 2005; Núñez 2004; see also Karjalainen 1999). Other suggested indicators of summer habitation are a high proportion of sh in the bone assemblage (Halinen 1997, Halinen et al. 1998, probably also Pesonen 1996b cf. Pesonen 2006), the presence of shing equipment (Halinen 1997), and the presence of red-ochre graves at the sites (Halinen 1997). In coastal Northern Ostrobothnia, the idea of the sedentary nature of the occupation is based on diverse arguments. In addition to phenomena implying sedentism, such as non-portable pottery vessels and labour investments on houses appearing in large clusters (e.g. Núñez & Okkonen 1999; 2005; Okkonen 2003: 171), there are plenty of other signals of sedentism. The other phenomena presumed to substantiate, or to suggest, an increasing rate of sedentism are the following: a resource rich environment maintained by an abnormally high rate of land uplift (Núñez & Okkonen 1999), territoriality expressed by large megastructures called Giant s churches (Núñez & Okkonen 1999), the presence of long-distance trade (Núñez & Okkonen 1999; 2005), intensied hunting (Vaneeckhout 2009c), and the shortening of the coastal shoreline due to land uplift. The latter has been recently put forth as a factor working together with the abundant resource base, which eventually increases the rate of sedentism among the groups packed together in the river estuary (Vaneeckhout 2008b; 2009a, c; 2010). Two papers of this study discuss the question of sedentism through changes in housepit sizes/shapes and site locations. Paper I argues for an increasing rate of sedentism in the Lake Saimaa area during the CW2 period, when larger housepit clusters 36

start to appear on windy locations such as capes and small islands and the numbers of other kinds of dwelling sites simultaneously decreases. Paper IV presents a similar argument concerning the succession of housepit sites during the CW2 period on the Karelian Isthmus, but due to the low temporal resolution of the available data, the decrease of dwelling sites without housepits cannot be observed in this area. In Paper IV, I have argued that housepits sites located in windy, unsheltered locations in an aquatic environment present several aspects that would support longer than winteronly, i.e., possibly year-round, habitation. Although most of the studies on housepits imprecisely refer to a certain degree of sedentary occupation without specifying what is really meant, there are also other lines of interpretation. In two quite recent studies, inland housepit sites are interpreted as winter-only dwelling sites. This interpretation is based on the osteological fauna and the restricted distribution of both nds and cultural layers either exclusively inside the housepit or in its immediate vicinity (Katiskoski 2002; Pesonen 2006). Interestingly, the interpretation of the Late Stone Age Kuorikkikangas site in Posio, Southern Lapland, has changed from year-round (Pesonen 1996b) to winteronly habitation (Pesonen 2006). Obviously, the change in interpretation results at least in part from a re-evaluation of sh bones as an indicator of habitation during the non-freezing seasons. In the Lake Saimaa area, the overall osteological material associated with Neolithic (ca. 4000 2000 cal BC) housepit sites (sample size: 26) differs from the assemblages of other dwelling sites (sample size: 68) (Mökkönen 2000: 52 58; 2001). In the housepit sites, the proportions of mammals and birds, both terrestrial and aquatic species, are slightly higher than on the other sites. However, the largest part of identied bone fragments consists of sh, which constitute 80% at housepit sites and 93% at other sites of all bones by numbers. If the bones not identied to species are also counted, the proportion of sh is smaller: 71% at housepit sites and 91% at the others. The most notable distinction in individual species was in the number of pike (Esox lucius), which constituted 74% of all sh and 38% of all identied bones at the housepit sites, and 19% and 17% respectively at the other sites. One cannot be sure, of course, but I would suggest that the high proportion of pike might be connected with the consumption of dried sh during the winter. Pike is a good sh for drying and was traditionally caught and dried during the spawning season in the spring/early summer and consumed later during cold seasons (Itkonen 1948: 283). In the Finnish inland, the year-round occupation of housepits has also been argued on the basis of archaeobotanical studies. According to Pirjo Jussila (1992; 1994), the charred plant remains found at two housepit sites the Naarajärvi site with CW2 (Matiskainen & Jussila 1984) and the Tahinniemi site with Pöljä Ware, both located in Pieksämäki parish in the inland lake area indicate year-round habitation. Especially the presence of plants gathered and used during the summer, such as strawberry and raspberry, is interpreted to indicate summer occupation. Only two of the charred seeds (both bearberry) from the Tahinniemi site have been dated. One was found to be contemporaneous with the Stone Age occupation while the other dated to the Early Metal Period. 37

It would appear that the emphasis on a sedentary interpretation of housepits has increased in recent papers. There is still a certain ambivalence in the interpretations. The most obvious weakness is the lack of studies in which all dwelling sites are included. Were the coastal housepit sites really occupied year-round, or have the settlements dispersed to summer sites on inland lakes and riverbanks? In his article dealing with sedentism and Stone Age houses in the Lake Saimaa area, Taisto Karjalainen (1999: 189) asks a reasonable question about the relation between housepit sites and other sites: on what grounds do we attach more signicance to the housepit structures than to the other archaeological material thick cultural layers, rich stone tool assemblages, and the presence of pottery when dening the permanence of a settlement? I cannot be certain, but I presume Karjalainen had the CW2 sites in mind, since they often yield large numbers of nds scattered over vast areas. Studies that include all the dwellings give different answers to the prevailing settlement pattern. In the area of the Kemijoki River, a semi-sedentary annual round based on coastal winter sites with housepits and inland summer sites mostly without housepits has been suggested by Kotivuori (1993; 2002). Following the denition of sedentism applied here, the suggestion that part of the population spent the whole year on the coastal housepit sites (Kotivuori 2002: 167) actually refers to fully sedentary settlement. In both the Lake Saimaa area and the Vuoksi River area, Ancient Lake Ladoga, a change in the settlement pattern is suggested to have taken place during the CW2 phase (ca. 4000 3400 cal BC) when dwelling sites with a higher number of housepits than before were established in highly unsheltered places (Papers I and IV). In the Lake Saimaa area, this change was accompanied by a reduction in the number of other types of dwelling sites. If similar changes in site location and in ratios between different site types could be observed in coastal areas adjacent inland areas, then we would have better grounds for arguing for sedentariness at the coastal sites. 3.4.2 Sweden In Sweden, there are also different interpretations concerning the degree of sedentism at the housepit sites. In inland Norrland, the housepit sites dating to 4500 2500 cal BC have been interpreted as winter villages (Lundberg 1997). In contrast, the coastal housepits are interpreted to have been inhabited on a more permanent basis. Arguments concerning a winter-only or longer occupation period at housepit sites are often based on osteological material. In the inland regions, the dominance of elk bones in housepit sites and the location of sites close to large pitfall systems are interpreted to indicate winter-only habitation (Lundberg 1997: 136). In coastal housepit sites, arguments concerning the degree of sedentism are also based on osteological material. In Överkalix, Upper Norrland, the Lillberget site with CW2 pottery dates to 3900 3500 cal BC. It is thought to have been a sedentary site because the optimal hunting seasons of the species in the osteological material cover all the seasons (Halén 1994: 164 167, see also Norberg 2008: tabell 4.6.). The previously mentioned Bjästamon housepit site in Middle Norrland (Chapter 3.3.2) is interpreted to have been occupied year-round based on the osteological material (Olsson et al. 38

2007; Runeson 2007). Similarly, the Neolithic housepit sites in coastal Norrbotten, Upper Norrland, yielded species in the osteological material that indicate longer than winter-only habitation, but the sites are nevertheless interpreted not to have been fully sedentary (Norberg 2008: 177, see also tables 4.5 4.9). On the coast, other factors also support the interpretation of sedentary habitation. In his dissertation on the Lillberget housepit site with CW2, Ove Halén presents several arguments supporting sedentary year-round habitation at the site. According to Halén (1994: 177 178) the presence of pottery, the rectangular house form, site location in a productive environment, great variation in stone tool types, and increased exchange relations (i.e., the presence of amber and copper in the nds) all speak for fully sedentary settlement. He considers Lillberget to be an extraordinary pioneering site that has been permanently settled at a time when most other contemporary sites have still been occupied on a less permanent basis (for a similar opinion see Norberg 2008: 174). Although the pithouse was primarily a winter dwelling, the Neolithic coastal housepit sites are thought to have been occupied for longer periods. This interpretation is accompanied, at least in Norberg (2008: 177 179), by the assumed presence of surface dwellings obviously to be used as alternative summerhouses at the housepit sites. At the Bjästamon site, year-round occupation is argued in addition to osteological material and the presence of the dwellings themselves on the basis of the archaeobotanical material (charred grains of barley Hordeum vulgare, one grain of wheat Triticum compactum, seeds of raspberry and hazelnuts) and the presence of burials (Olson et al. 2007, Runeson 2007). As concerns the osteological material associated with coastal housepits, summer occupation has for the most part been deduced from the presence of the bones of the harp seal (Phoca groenlandica), which was hunted on the Norrland coast during the summer. The shing and hunting of other species, excluding certain shes such as ounder (Platichthys esus) and bullhead (Cottidae sp.), is not limited to a certain season, and the species can be caught at almost any time of the year. (Olsen et al. 2007.) In coastal Norrbotten, Upper Norrland, the osteological material from Neolithic housepit sites yields species that could support an interpretation of sedentary habitation at the sites. In addition, the growing size of the housepits and sites, together with a concurrent decrease in the number of sites and the increased exploitation of marine resources (as seen in the bones and in site location), is interpreted to mark a growing degree of sedentism during the Neolithic period (Norberg 2008: 161 164). Even so, Erik Norberg (2008: 177) does not consider Neolithic coastal housepit sites to be fully sedentary sites but rather sites with winter houses with a longer than winter-only yearly occupation phase. In contrast, the Bjästamon site in Middle Norrland is interpreted as a sedentary site with year-round occupation (Runeson 2007). 39

3.4.3 Norway In the early studies on North Norwegian Stone Age housepits, the structures were presumed to go together with a semi-nomadic settlement pattern (Gjessing 1941: 38; 1945: 170 172). From the 1960 s to the late 1970 s, the large coastal agglomerations of housepits were thought to represent the winter early spring component of an annual round that consisted of seasonal sites extending from the outer archipelago to the inland mountain region (Gjessing 1975; Simonsen 1975; 1979). During the 1980 s, several new studies suggested that settlement has been more sedentary than was previously thought. The archaeological material, that is, osteological data and lithics, reects an extensive degree of variation during the Late Stone Age. Both semi-sedentary and year-round, permanently settled sites as well as sites used for shorter periods have been identied. According to Erica Engelstad, this variation results from different organizational levels of the population, and on the other hand, from exible adaptation systems that adapted to uctuating resource abundance (Engelstad 1984; see also Engelstad 1990). Other archaeologists have used the osteological material to argue that at least some of the Late Stone Age pithouses were occupied all year round (Helskog 1983; Renouf 1984: 24; 1986; 1988). During the two rst phases of the Younger Stone Age (4500 3000 cal BC), settlement patterns have obviously been based on seasonal sites (Olsen 1994: 65 66, 69; Hesjedal et al. 1996: 206 211). Alongside the trend towards larger-sized pithouses and the occurrence of Gressbakken-type pithouses during Phase III (3000 1800 cal BC), the archaeological material reects a greater degree of sedentism than before. This has even been interpreted as evidence of year-round habitation (Olsen 1994: 71, 82; Schanche 1995; Hesjedal et al. 1996: 210 211). Olsen (1994: 73 74) argues that the variation in both osteological and archaeological materials, the construction of permanent houses, the high number of settlement phases on the sites, and the increased number of burials at sites all support a hypothesis of sedentism during Phase III. He considers the coastal settlement to have been sedentary or semi-sedentary (Olsen 1994: 76, 85). Recently, Lisa Hodgetts (2010) has pointed out the differences in mammal fauna at the Gressbakken type housepit sites in the Varangerfjord area. According to Hodgetts, there is a clear difference between the inner fjord sites with high proportions of whale and dolphin at Gressbakken and a high proportion of reindeer at Karlebotn, and the southern fjord sites with an absolute dominance of seal. She suggests that the communities that settled each site had a restricted hunting territory, which also dened the exploitable species and thereby the seasons the sites were settled. Judging by the species, all of the sites have evidence of spring occupation (seals), while the high number of reindeer at Karlebotn speaks for intensive autumn occupation and the notable proportion of cetaceans (whales and dolphins) at Gressbakken points to intensive summer occupation. Hodgetts argues that the variability in refuse fauna reects the differences of locally available resources within hunting territories and, hence, also, the variation in the seasonal intensity of settlement at each site. Another recent study on the Karlebotnbakken housepit site in the Varangerfjord (Hood & Helama 2010) has shed a shadow on the interpretations based on osteo- 40

logical data referred above. In contrast to previous studies, new radiocarbon dates have shown that the housepit and the surrounding middens are of different ages. The material in midden dates to 3360 2870 cal BC and the house of Gressbakken type to 2210 1530 cal BC. The new dates displace the whole midden material from the Gressbakken phase. The archaeological materials found in the midden deposits are of notably older age than the last habitation on the site during the Gressbakken phase. The thick midden deposits include a few pottery shards related to CW3 and Kierikki Ware, the context dating of which now makes sense. Interestingly, the copper dagger/ point previously dated to the beginning of the Bronze Age is now dated following the context in the midden to ca. 3200 2800 cal BC. Another recent study also shares the view of a high degree of sedentism among the coastal Late Stone Age societies in the Finnmark area. The settlement at Sundfæra, Melkøya Island, is interpreted to be fairly sedentary during the most intensive occupation period (4000 3300 cal BC). Thick cultural layers, huge amounts of recracked stones, remains of substantial pithouses, and other structures together with a rich artefact inventory reecting a maritime orientation are thought to indicate repeated and long-lasting occupation at least during the cold part of the year (Niemi 2010). This can be referred to as a semi-sedentary settlement pattern. In the beginning of the Early Metal Period (1800 cal BC), the size and number of housepits decline. At this time, the settlement pattern is thought to have become more mobile again, with seasonal settlement sites (Olsen 1998: 187). 3.5 Summaon and discussion of housepit research in Fennoscandia In this section, I summarize the main trends in the Neolithic pithouse building tradition of Fennoscandia (Chapter 3.3). These trends include a large number of concurrent or nearly concurrent changes in size, shape and number of housepits as well as in site location in relation to the environment. After a chorological and chronological summation of the pithouse tradition, I will summarize and briey discuss the bases on which the interpretations concerning the degree of sedentism (Chapter 3.4) are built. At the moment, Finland, with only a few Mesolithic housepits, is an anomalous area as compared to neighbouring countries at the same latitude. Otherwise, in the course of the Neolithic Stone Age, there is much common development in housepits (see Norberg 2008: 159 160). In Scandinavian countries on average, the highest number of housepits appear ca. 3000 2500 cal BC, while the largest ones come along slightly later, ca. 2400 1800 cal BC. Simultaneously with the occurrence of the largest housepits comes a decreasing trend in the total number of housepits. There are, however, some regional distinctions (see Fig. 13 and further discussion in Chapter 5.3). As a subtle temporal discrepancy, the peak period concerning both size and number in the Republic of Karelia, Russia, dates slightly earlier than in other areas of Fennoscandia. The peak in the Republic of Karelia dates to ca. 3300 2500 cal BC, and similarly, the decrease in size and number took place earlier 41

than elsewhere in Fennoscandia, ca. 2500 2000/1800 cal BC. As distinct from the rest of Fennoscandia, there are no known Bronze Age/Early Metal Period housepits in the Republic of Karelia. I cannot say for sure, but I presume that the development of housepits in the case study areas on the Karelian Isthmus and in the Lake Saimaa area was quite similar to what is seen in the Republic of Karelia, at least as concerns the decrease in the number of housepits. In the Lake Saimaa area, there are only a few housepits with Early Metal Period material (1800 BC AD 300), and no Stone Age housepits dating unquestionably to 2500 1800 cal BC (Karjalainen 1999; 2002). 5 By the Gulf of Finland in south-eastern Finland lie a few large housepit sites that are associated with both CW2 and later Middle Neolithic ceramics. The only solitary large multi-room housepit in this area with datable material and radiocarbon dates, the Meskäärtty site, is proof that the largest housepits were in use at least during the latter part of the 3 rd millennium cal BC and possibly even as early as the late 4 th millennium (Paper III). On the Ostrobothnian coast in north-western Finland, the increase in size of the largest housepits began already during the late CW2 period, beginning ca. 3500 cal BC, although the fastest growth rate was reached slightly later (Vaneeckhout 2009a, c; 2010). In Norrbotten, Sweden, the most rapid change in housepit size measured as average oor area took place after 2800 cal BC (Norberg 2008: 57). In Finland, the shift towards larger pithouses took place earlier than in Sweden and Norway. In the Republic of Karelia, Russia, the increase in size seems to be fairly concurrent with the development seen in Finland, but in Russia pithouse size peaked slightly earlier than in Finland. The preferred type of location for building pithouses changed in the course of the Neolithic. On the Karelian Isthmus (Russia), in the Lake Saimaa area (Finland), and on the coastal areas around the Gulf of Bothnia (at least in Sweden), the sites shifted during the Neolithic to more aquatically oriented locations. Interestingly, this change appeared at different times in different areas. In the Vuoksi River Valley, on the Karelian Isthmus, housepits colonized the archipelago during the period when CW2 was in use (ca. 4000 3400 cal BC). 6 There, the chronological division of the sites is based on the ceramics associated with the housepits, and therefore it is difcult to give specic dates. However, it is probable that the largest oblong housepits in this area date younger than 3300 cal BC. They appeared as solitary structures by strategic water routes in a more terrestrial environment than the older housepits located in the archipelago (Paper IV). In the Lake Saimaa case study area, the change in housepit site locations also occurred during the CW2 period (Paper I). In this area, the very same unsheltered sites on capes and islands where the housepits were located during the latter part of 5 At least there are no housepits dated to this period by radiocarbon. This statement is also based on my own database on Stone Age and Early Metal Period sites in the River Vuoksi catchment, updated to the summer of 2006. Reports led at the Archives of the Department of Archaeology, National Board of Antiquities (Helsinki, Finland) after July 2006 are excluded. 6 On a large former island in the outer archipelago of Ancient Lake Ladoga, there were pithouses already during the Mesolithic Period (Paper III, Halinen & Mökkönen 2009). 42

CW2 phase were often utilized later by groups that used asbestos tempered wares (Kierikki Pöljä Jysmä). In northern Upland southern Norrland (Sweden), settlement shifted to a maritime environment in two stages: rst, colonizing the archipelago (3400 3000 cal BC), and later shifting the whole focus of settlement to the archipelago (3000 2300 cal BC). In more northern areas of Norrland, the location of housepit sites shifted towards more maritime environments ca. 3000 1800 cal BC. There, the maritime orientation is also apparent in the osteological material found at the sites. Parallel developments in housepits and housepit sites occurred earlier in the southeast and later in the north-west. The changes appear to represent a chain of events, but the absolute timing as well as the duration of the development phases varies from place to place. The increase in numbers and the aquatic/maritime orientation of housepit sites seem to emerge together on the Karelian Isthmus (ca. 4000 3400 cal BC, Paper IV), in the Lake Saimaa area (ca. 4000 3400 cal BC, Paper I), and in coastal Norrbotten, Sweden (3400/3000 2300/1800 cal BC). The development in the Stone Age estuary of the Kemijoki River seems to be exceptional with respect to other areas. There, the development towards larger housepits after the CW2 phase is missing, and unlike elsewhere, the largest housepits in the area are associated with CW2. The changes have their own temporal cycles and occurrences. The temporal distance between (1) the growing number of housepits and the concurrent relocation of sites, and (2) the growth in housepit size is longer in the Lake Saimaa area and on the Karelian Isthmus than in Norrbotten, where all three changes (the increase in number and size, and aquatic orientation) appeared during a notably shorter time span. In coastal Central and Northern Sweden, the relocation of dwelling sites is associated with the rst signs of cereal cultivation (see also Paper V). I have examined here only those trends in housepits that I have found interesting in my articles. Therefore, it is probable that the relocation of housepit sites into more maritime environments is not the whole story. Some societies focussed on utilizing rich resources that occurred only in a restricted area. Thus, it is not surprising that pithouse villages that controlled salmon shing did not go through a relocation phase but continued their life on the riverside near the weirs. This was probably the case at, e.g., the housepit sites by the Iijoki River in Yli-Ii, Finland (Kankaanpää 2002: 73). No relocation of housepit sites to a more maritime environment occurred at this area (Vaneeckhout 2009a; c; 2010). It is obviously not an easy task to present a well-founded argument concerning the degree of sedentism. The osteological material, which would offer the clearest evidence of particular seasons of occupation, is quite limited in Finland, especially in the inland lake areas. Species with a short and intensive exploitation period are the most useful ones for specifying the season. Unfortunately, such species are rare during the summer, and thus the identication of summer habitation is challenging. On the shores of Gulf of Bothnia, the only species in the osteological material clearly associated with hunting and shing during summer are the harp seal (Phoca groenlandica) and the ounder (Platichthys esus) (Olsen et al. 2007). If there are 43

only two species in the northern Baltic exclusively connected to summer subsistence activities, in the Finnish inland lake district there are none. It is always possible to emphasize the season with the easiest catch (sh spawning etc.), but nearly all species, excluding migratory birds, are also available for shing or hunting during the rest of the year. In some studies, the distributions of nds and cultural layers have been used to distinguish winter-only habitation. Nevertheless, one may ask whether it is possible to interpret a nd distribution that is closely concentrated in a housepit as a signal of winter-only occupation. In Finland, the idea of the connection between such a nd distribution and winter-only occupation has been put forward in two rather recent articles on totally excavated housepits, one dating to 3200 2800 cal BC (Katiskoski 2002) and another to 2900 2300 cal BC (Pesonen 2006). There are, however, a few aspects challenging that idea. First, the excavation areas were restricted to the housepits and the surrounding areas were excavated only with a coarse grid of test pits. Second, the excavations in Middle Norrland, Sweden, in which large areas adjacent to year-round inhabited housepits were excavated, revealed that starting from c. 2800 cal BC, the nd material clearly concentrated inside the dwellings and the activities in the yard were limited and well organised following a cultural schema of acting in space (Gustafsson 2007; Holback 2007; Runeson 2007). The excavations in Middle Norrland, Sweden, have shown that from ca. 2800 cal BC onwards, activities were both more well-ordered and more concentrated inside the dwellings than on older sites (Gustafsson 2007). Although comparable extensive excavations on which to base such arguments have not been executed in Finland, a similar development towards an increased concentration of activities inside dwellings have taken place also in Finland. Therefore, I suggest that simultaneously with the increase in housepit size, which occurred in Finland mainly after CW2 (4000 3400 cal BC) and culminated around the mid 3 rd millennium cal BC, the activities performed at the dwelling sites may have increasingly been concentrated inside the dwellings. Consequently, a nd distribution concentrated inside a housepit may not necessarily be considered a self-evident marker of winter-only habitation. Another aspect to highlight here is that the richness or poorness of nds inside or outside a housepit cannot be measured between two non-contemporaneous sites. It is not reasonable to measure the richness of Neolithic sites based on experiences gained with CW2 sites, where the rich material is usually spread over vast areas. The CW2 presuppositions are not automatically valid in other contexts during the Neolithic Stone Age (see Chapter 5.7). 44

4 HOUSING vis-à-vis ENVIRONMENT In this chapter about housing and environment, I rst describe the relations between the subsistence base and the settlement pattern and then dene the cultural aspects related to housing. The third section is concerned with environmental studies carried out in connection with Finnish Stone Age research. Here, I discuss research carried out both in natural environments and within the concept of landscape. 4.1 The relaon between subsistence and selement paern the environment gures prominently in how hunter-gatherers decide what to eat, whether to move or stay, to share or to hoard, to let someone into their territory or not.. (Kelly 2007:64). The environment denes the stage and marks the borders in which man nds his ways to adapt. The density, predictability, and nature of resources all are directly linked to the applicable subsistence strategies and thus also to settlement patterns. Subsistence is the primary variable affecting mobility (Binford 1980; Kelly 1992: 46). Other factors that affect the selection of certain subsistence strategies and settlement patterns are cultural concepts and general knowledge, for example, of the environment and technological innovations (e.g. in tools, mass-capture devices, storage, and transport) (Kelly 1992; 2007: 108 110; Ames 2002; 2003: 24, 32). According to Binford (1990), there is a very straightforward pattern between effective temperature (ET, measuring the overall warmth and length of growing season) and hunter-gatherer mobility: the more severe the winters, the less mobile the huntergatherers. During the Neolithic Stone Age, the effective temperature in Finland is estimated to have been 12.1 11.5 degrees Centigrade (Hertell 2009). According to Binford (1990: Table 8), ethnographically documented hunter-gatherers in such circumstances (ET 12) practise residential mobility in the following proportions: fully nomadic 4.5 %, semi-nomadic 59.1 %, semi-sedentary 31.8 %, and fully sedentary 4.5 %. However, it must be noted that in this result the subsistence base of the communities has not been taken into account. The major resources in the subsistence base have an effect on mobility. According to Binford (1990), there is a difference between hunter-gatherers favouring terrestrial resources and those favouring aquatic resources. The groups with a high dependence on aquatic resources, especially sh, are usually less mobile than the ones oriented towards terrestrial hunting. They tend to have smaller territories and make 45

fewer residential moves than their terrestrially oriented fellows. Binford suspects that the groups with logistical mobility strategies ( collectors ) are a consequence of an aquatic resource revolution and the perfection of water transport. Globally, the dependence on aquatic resources increases towards higher latitudes (Binford 1990: 137). One can say that Finnish Neolithic cultures are aquatically oriented: in the coastal areas both shing and sealing have played an important role in subsistence (Siiriäinen 1981; 1982; Hertell 2009; Núñez 2009) and in inland lake areas shing has been the base of livelihood (Ukkonen 1996; Mökkönen 2000: 42 52; 2001). Therefore, it might be expected that the aquatically oriented Neolithic huntergatherers who built non-portable pithouses were rather less than highly mobile in Finland as well. The settlement pattern is not dened only by the quality and abundance of certain resources but also by the way the resources are distributed. In Fennoscandia, the movements of animals and sh are regulated by the seasons. In certain areas, there are clear hotspots in the abundance of resources. In the Baltic Sea area, locations close to river mouths are the number one hotspots. There, particularly salmon shing is a highly productive activity that is concentrated in a few exceptionally suitable locations. The seasonality and the volume are both, basically, higher by the sea than in the inland lakes, where the modest resources are spread more evenly. It is clear than if hotspots are present, there is not much room for variation in site location, at least in the case of the communities holding the rights to exploit the hotspot. Although the environment sets the limits to subsistence and mobility strategies, there is still room for variation. A good ethnographic example comes from the Varangerfjord area in northern Norway, were two historically documented 18th century Sami groups had different settlement patterns in use at the same time within a restricted area. According to Engelstad (1990: 29 31), both groups had seasonal base camps, one for summer and another for winter use. One group, which settled the fjord area year-round, had winter settlements in the northern interior part of the fjord and summer settlements further out in the fjord towards the sea. The second group settled the interior part of the fjord during summer and had winter settlements inland, located a few kilometres from the head of the fjord. Quite similar observations have been made about Stone Age communities practising different subsistence strategies and settlement patterns in the same area (Hodgetts 2010, see Chapter 3.4.3). During the Neolithic Stone Age, housepit sites were found in different environmental settings. After, or maybe already during, the CW2 phase, dwelling sites with semi-subterranean dwelling structures were located by the rivers, at the river mouths, and in the archipelago along the coasts of the Bothnian Bay. Although I cannot prove this view with the help of osteological material, I presume that during the Middle Neolithic each community may have had its own subsistence niche. According to this idea, it may be expected that the communities that lived in different environmental zones focussed on the utilization of certain available species. A simplied version might run something like this: communities by the rivers based their subsistence on salmon shing while the others in the coastal areas specialized in seal hunting. This is only a coarse generalization, but I believe that there were different branches of 46

resource utilization and settlement pattern within communities with fairly a uniform material culture. So far, the only data supporting this view is the varying location of the Middle Neolithic housepit sites (see Chapter 3.3.1). 4.2 Cultural and technological aspects of housing The relations between hunter-gatherers and the environment are not dened only through subsistence practices. In research that emphasizes the experience of place/ landscape, the way prehistoric man has acted in his environment is seen as a culturally constructed phenomenon (e.g., Tilley 1994: 15 17, 67; Ingold 2000: 153, 171; Thomas 2001). Although different approaches have their own emphases on studying the physical world as a natural environment or culturally dened place/landscape, these approaches do not need to be in conict with one another. The ecological approaches establish the broader framework for resource utilization, while approaches concerned with place/landscape introduce the culture-related aspects of resource selection and the sites chosen to be inhabited (Jones 2002: 87). According to Ingold (2000: 60), there can be no radical break between human social and ecological relations with the environment. He considers the social/cultural aspects to constitute a subset of the ecological relations. Building a dwelling is a combination of technological aspects (such as available material, technological knowledge of construction) and culturally dened housing needs (Sanders 1990: 44). Usually, there is a positive correlation between the degrees of segmented architecture, segmented use of space, and segmentation of culture (Kent 1990b). Just as in the case of available resources (vs. actually exploited resources), the existence of required knowledge is not enough to cause it to be utilized in practise. A new technology and architecture can be adopted into a culture only if it is compatible with prevailing cultural convention (see Kent 1990a). I accept both environmental factors and cultural aspects as determinants for the selection of a dwelling site. In research, these opposites can be maintained as different points of departure, but I do not consider it reasonable to separate these aspects. I have tried to understand site location in relation to both broader environmental zones (Mökkönen 2000; Paper IV) and the immediate environment (Papers I and IV). I propose that a site s immediate environment, i.e., the surroundings of a site observable by human senses at the site, tell us about the specic circumstances that were of major signicance when choosing the particular location for habitation. In an environment where larger ecological and geographical zones are clearly separable, as in the Stone Age Vuoksi River Valley (Paper IV), the changes in demands on the sites immediate environment resulted in clear changes also in the geographical distribution of the sites. In a mosaic-like environment such the Lake Saimaa area (Paper I), the physical distance between two dwelling sites that display very different aspects regarding the sites immediate environment can be very short. In other words, the essence of the natural environment affects the physical distance between housepit sites that exhibit different concepts of housing/dwelling. Here, it must be remembered that 47

both areas discussed above do not include any particular resource concentrations of high productivity, and that there is therefore much room for cultural variation not dictated by the resources. On the Ostrobothnian coast, the building of Giants churches (also simply called megastructures), settlement embankments, and cairns is seen as the beginnings of an articially modied landscape. Following the dating of various large-scale structures and village-like concentrations of housepits, the beginning of a conscious shaping of the landscape in Ostrobothnia is dated to 3500 2200/2000 cal BC (Okkonen 2001; 2003; Núñez 2004; Núñez & Okkonen 2005). Although the changes are not as clearly visible elsewhere as they are around the Gulf of Bothnia, a trend towards larger-size pithouses appeared in most of Fennoscandia during the late 4 th and the 3 rd millennium cal BC (see Chapter 5.3, Fig. 13). Looking at the variety of structures and site locations, it can be said that a change in attitudes towards dwelling and subsistence practises occurred during the CW2 phase (see also Okkonen 2003; 2009; Vaneeckhout 2010). During the 3 rd millenium cal BC, the increasing variety in structures and site locations may refer to specialised resource utilization. As concerns the practise of cereal cultivation (Paper V), it ts well into this background. 4.3 Environmental studies in Finnish Stone Age research Studies focusing on the relations between Stone Age dwelling sites and the environment can be divided into three categories: (1) studies utilizing the ideas of site catchment analysis, including both non-computer-aided and computer based GIS-studies, (2) studies concerned with the location of dwelling sites in relation to ecological zones, and (3) studies focusing on the immediate environment of the dwelling sites. Other studies that can be labelled environmental, such as zooarchaeological (e.g. Ukkonen 1993; Nurminen 2007; Mannermaa 2008; Seitsonen, S. 2008) and ecological studies (e.g. Siiriäinen 1982; Hertell 2009) are excluded because they are not focused on dwelling site environment relations. Environmental reconstructions have long been based on shoreline displacement (see e.g. Ailio 1915; Pälsi 1915) but the rst environmental studies in Finnish Stone Age research, in the sense meant here, were carried out by Marek Zvelebil (1981). Zvelebil used site catchment analysis and von Thünen s rings in dening the resources most easily exploitable from the sites. It took a while before Finnish archaeologist produced the rst studies relating to the ideas of site catchment analysis. The rst studies were manual (Saukkonen 1990; 1994); later ones were carried out in a GIS-environment and studied Stone Age settlement in relation to soil type (Vikkula 1994b; Kylli 2000; 2001) or in the absence of accurate soil maps to modern vegetation, which has certain links to Stone Age vegetation (Mökkönen 2000). The 1980 s and 1990 s also produced other studies comparable to site catchment analysis. In these studies, the locations of dwelling sites were studied in relation to ecological zones on a coast inland axis (Siiriäinen 1981; 1987; Matiskainen 1989, Sartes 1991; 1994; see also Halinen & Mökkönen 2009). The authors dened the pre- 48

vailing settlement patterns of different ages based on the distribution of different kind of sites in ecological zones with certain resources available during certain seasons. GIS-based environmental studies have been carried out also in order to understand environmental change caused by land uplift. Milton Núñez and Jari Okkonen (1999) explored the temporal changes in the interplay between land uplift and local topography. They found that the orescence period of the North Ostrobothnian Neolithic Stone Age coincided quite closely with an abnormal increase of new land, which produced extensive wetlands. Ever since that study, this environmental phenomenon has been considered the main factor behind both the rise of local cultural complexity that began during the CW2 period and the decrease in signs of prosperity and complexity that occurred simultaneously with the termination of the wetland period around 2500 cal BC (Núñez & Okkonen 1999; 2005; Núñez 2004; 2009; cf. Paper III and Chapter 5.3). In addition to these studies, carried out, as it were, from a satellite perspective and in a more or less environmentally deterministic mode, some studies have concentrated on the immediate environment of Stone Age dwelling sites (Vikkula 1994a; Wilhelms 1995; Pesonen 1996a; Mökkönen 2000; Nordqvist & Lavento 2008; Seitsonen & Gerasimov 2008; Halinen & Mökkönen 2009; Papers I and IV). These studies describe the dwelling sites immediate physical surroundings through several variables (shoreline type, amount of open water and direction into which the dwelling site faces) rst described by Vikkula (1994a) and later slightly modied by others. In general, the aim of these studies was to understand the variation in the dwelling sites topographic location. Some of the studies have focused purely on the degree of protection from winds (Halinen & Mökkönen 2009; Paper IV). My MA thesis on the Lake Saimaa area did not produce any remarkable results in the section employing the ideas of site catchment in a GIS-environment there were no noteworthy changes in site location in relation to larger environmental variables (Mökkönen 2000). Instead, the part of the study dealing with changes in the sites immediate environment did produce highly interesting results (Mökkönen 2000, Paper I). In the Lake Saimaa case study area, where the environment is a ne-grained mosaic of land and water and where the resources are evenly distributed, the radical change in the way the landscape was occupied was visible only in the change in the housepit sites immediate environment (see Paper I). In the case of Lake Saimaa, the change in site location occurred within a very limited area. The distance between the older well-sheltered housepit sites and the younger ones in unsheltered windy locations was often less than one kilometre. It appears that the change in settlement pattern was not related to change in the resource base exploitable from the site but rather a cultural change in the way of settling the landscape. This is supported by the osteological analyses, which do not show any great disruptions in resource utilization during the Neolithic Stone Age (Mökkönen 2000; 2001). 49

50

5 CULTURAL CHANGE IN THE STONE AGE, 4000 2000 CAL BC In this chapter, I discuss further aspects more or less explicitly touched upon in this study. First, I summarize my views concerning the late Middle Neolithic cultures as dened with the help of ceramics. I then briey review the use of shoreline displacement as a useful tool for detecting cultural changes irrespective of periodizations based on artefacts. This discussion is of vital importance for two reasons. First, the changes in the way landscape has been inhabited cannot be observed in a reliable way if time is incorporated into the study as blocks dened by distinct pottery styles (Paper I, see also Mökkönen 2009), and second, the late Middle Neolithic cannot be outlined on the basis of distinct ceramic styles simply because ceramics are too heterogeneous for reliable typologization. After making my point concerning the possibilities of outlining late Neolithic cultures purely on the basis of ceramics typologies, the changes in housepits are introduced as the focus of discussion. The changes in housepits in Fennoscandia are discussed, and the cause of the changes is proposed to lie in cultural and technological innovations rather than in environmental changes. I then discuss aspects of Middle Neolithic sedentism and the parallel development seen in the contact zone at the northern limit of the Corded Ware Culture in Fennoscandia. Lastly, I present my reasons for scepticism concerning the proposed population decline during the late Neolithic and the beginning of the Early Metal Period, which I believe has more to do with cultural changes affecting the visibility of archaeological material than with an actual large-scale population decline. 5.1 The burden of research history: Poery types as a synonym for culture The tradition of typological studies on ceramics remains strong in Finnish archaeology. In the Neolithic Stone Age, cultural phases are still referred to through pottery types (see Lavento 2006), some of which were created over fty years ago based on the materials of a single or just few dwelling sites. Neolithic cultures dened on the basis of typologies of ceramics have been seen to follow one another with only minor temporal overlap. Today, there are good reasons to question the traditional ceramics-based view of cultural development in the Finnish Neolithic Stone Age, and particularly during its latter part. The fragility of pottery typology as a chronological tool is revealed by recent AMS radiocarbon dates on organic crust adhering to clay vessels. New dates have 51

shown that the use periods of Neolithic pottery types overlapped more than was previously thought (Leskinen 2003; Pesonen 2004; Pesonen & Leskinen 2009; see Fig. 5). Hence, pottery types are not chronological sequences directly following one another. This has an enormous impact on Stone Age research. Previously, the chronological framework of ceramics typology was used in nearly all studies on the Neolithic Stone Age, and thereby all the studied topics were based on that chronology (see also Mökkönen 2009). Now that the traditional chronology has turned out to be poorly based, it appears that a number of studies utilizing the ceramics-based chronological framework have treated contemporaneous materials as though they derived from different chronological levels. This has had an impact, for example, on studies of settlement patterns. Another problem of the chronological framework based on traditional pottery typology is the way it forces us to handle time. The basic problem is the perception of time as blocks. Each cultural block is identied with a certain type of ceramics lasting some ve hundred years. Since the chronological framework denes the accuracy of the results, it is not difcult to foresee the consequences. A chronology with xed long periods tends to have a self-fullling character. This means that it is difcult to observe change within a cultural block, while the interfaces between the successive blocks automatically become illusory periods of change. In my view, late Middle and Late Neolithic potteries in the southern part of Finland may be too varied in shape, size, decoration, and temper to be easily dealt with using traditional pottery typology. I assume that the strong belief in the accuracy of existing ceramics typologies has limited the studies in two ways: by forcing some ceramics into the current typology (e.g. Kiukainen Ware in south-eastern Finland, Paper III), and, on the other hand, by excluding some ceramics from typologies (e.g. organic tempered variants of Pöljä Jysmä wares). As an example of the latter, the denition of Oravnavalok XVI type pottery in the Republic of Karelia, Russia (Zhul nikov 1999), which is a parallel of the Finnish Pöljä Jysmä wares, includes both asbestos and organic tempered variants. This differs from the denitions of Pöljä Jysmä wares, where asbestos temper is highlighted as a dominant characteristic of the type (Meinander 1954a: 162, cf. Huurre 1984; 1986). The existing pottery typology has directed archaeological thinking to a marked degree. Since the denition of Kiukainen Ware (see Meinander 1954a), the relations between Corded Ware and the pottery traditions that previously existed in Finland have been conceptualized through the ideas incorporated into Kiukainen Ware, which is seen as an amalgamation of two distinct pottery traditions, namely the Comb Ware and Corded Ware traditions. Consequently, Middle Zone Pottery, which is distributed inland at the northern limit of the Corded Ware Culture, was seen as an inland parallel to Kiukainen Ware (Carpelan 1979:15). However, a number of recent studies challenge the traditional view of the cultural development after the arrival of the Corded Ware Culture (Edgren 1997; Carpelan et al. 2008; Paper III). These studies argue that the contacts between the local traditions and the Corded Ware tradition started long before the birth of Kiukainen Ware. It now appears that the whole concept of late Middle and Late Neolithic cultural development might be turned upside 52

down. The mutual contacts between Corded Ware and local pottery traditions probably took place rst in the northern contact zone, and, thereby, Kiukainen Ware may not belong to the beginning but rather to the last phase of the amalgamation. Middle Zone Pottery 7 and Kiukainen Ware might display parallel ideas, but I believe that they are not chronological parallels. Middle Zone Pottery may actually be older than is currently believed. 8 For example, the stratigraphy of the Kouvola Huhdasjärvi housepit site suggests that Middle Zone Pottery was in use at the site earlier than Pöljä Ware (Miettinen, T. 2004). In Finland, the main contact zone between Corded Ware and the local pottery traditions was denitely located along the northern limit of the distribution of Corded Ware. There, the contacts may be observed in coastal Southern Ostrobothnia in the west (Edgren 1997; see also Miettinen, M. 1998) and in extreme south-eastern Finland (Paper III). In Russia, on the Karelian Isthmus and in the Republic of Karelia, the inuence of the Corded Ware culture reached farther north, towards Lake Onega, than it did in south-eastern Finland (Huurre 1998: 235 236). In the Republic of Karelia, Palaiguba II type ceramics (ca. 2500 1500 cal BC) have been suggested to display inuences from southern cultures already practising agriculture, i.e., from the Corded Ware Culture or Fatyanovo (Zhul nikov 1999). Such inuences are seen in vessels shapes, i.e., in the proling and the at bottom. In addition to Kiukainen Ware, Corded Ware inuence may possibly be observed also in other Finnish ceramics. As an example, both proling and at bottoms are present in Pöljä Ware found in a housepit at the Outokumpu Laavussuo site (2700 2400 cal BC) 9 (O Caellacháin 2008). The inuence of Corded Ware may be presumed because proling and the at bottom are not known from the local pottery traditions, Pyheensilta Ware being an exception. It is reasonable to suggest that the interaction between Corded Ware and the local traditions commenced immediately after the arrival of the Corded Ware Culture (see Carpelan 2004). Archaeology encompasses various kinds of studies. In this chapter, I have tried to underline the problems involved in studies where the chronological framework is based on ceramics typologies. In such studies, the main problems are an inability to observe change within long pottery sequences and the ephemeral nature of poorly based typologies. The chronological framework of ceramic cultures is certainly a practical means of chronological perception (see Paper IV), but ceramics types should be only a medium for discussion, not the master. In the repertoire of Finnish archae- 7 In his article on asbestos tempered ceramics in Fennoscandia, Carpelan (1979) placed Middle Zone Pottery / Middle Zone Ceramics a poorly studied late Middle Late Neolithic organic tempered pottery type that displays similarities with Late Corded Ware found in the Northern Baltic region chronologically in the same period with Kiukainen Ware. However, I would characterise the Middle Zone Pottery as a possible outcome of similar contacts between the local and Corded Ware traditions as observed in Southern Ostrobothnia (Edgren 1997) and in south-eastern Finland (Paper III). 8 As far as I know, there are no radiocarbon dates from the context of Middle Zone Pottery. The current dating of that poorly studied material is based on reasoning. 9 The radiocarbon date (4010 ±60 BP, Hela-153) was obtained from a wad of chewing resin found inside the house. Other conventional radiocarbon dates on charcoal and birch bark from the house structures date to a slightly broader period ca. 2900 2350 cal BC (see Karjalainen 1999). 53

ology, there are much better vehicles for measuring the tempo of cultural change. From a global perspective, the still continuing phenomenon of land uplift, combined with archaeological material, offers an almost unique possibility to observe cultural changes by creating parallel independent levels describing different aspects of cultural change. This is discussed further in the next chapter. 5.2 Shoreline displacement: A tool for observing cultural change During the last Ice Age in Fennoscandia, the Earth s crust sank under the pressure of a huge mass of ice. After the Ice Age, the crust began to rebound slowly, and this process still continues. This phenomenon is called land uplift, and its effect vis-á-vis water systems is called shoreline displacement (see e.g. Miettinen, A. 2002). Shoreline displacement is basic knowledge for every archaeologist in Finland, whatever period is studied. In Finnish archaeology, shoreline displacement is used for dating and reconstructing the past environment. As a chronological tool, shoreline displacement can be used for relative dating, and especially after the introduction of radiocarbon dating, for a kind of absolute dating as well (see Siiriäinen 1969; 1970; 1971; 1972; 1974; 1978). In my opinion, the utilization of shoreline displacement in archaeological research after the early 1970 s has mostly been concerned with absolute dating. During the period of conventional radiocarbon dating, the rate of land uplift, which was anchored to calendrical time with the help of radiocarbon-dated limnological samples, facilitated the relatively accurate dating of shorebound dwelling sites. From the current point of view, shoreline displacement chronologies cannot reach the same level of accuracy as modern AMS radiocarbon dates. The imprecise nature of dating based on shoreline displacement, however, does not mean that the whole method belongs to the past. It seems to me that the basic idea and strength of shoreline displacement as a tool for relative dating was, in a sense, forgotten at the time the focus was on absolute dating. Considering the great potential of relative dating based on shoreline displacement, there are surprisingly few recent studies that utilize it. Within the last ten years, however, a number of housepit studies employing relative shoreline dating have been carried out in Finland (Mökkönen 2000; Paper I; Okkonen 2003; Vaneeckhout 2008a b; 2009a c; 2010) and in Sweden (Norberg 2008). The possibility of dating archaeological sites relatively with the help of elevation renders it possible to arrange a large number of sites in chronological order. In a normal situation, in an area without land uplift, such a chronological arrangement of a multitude of sites would require a massive amount of radiocarbon dates. It has been proposed that Stone Age sites followed the regressive shoreline towards lower elevations. This might be the rule on the seashore, but not really in riverine environments. In Yli-Ii, Northern Ostrobothnia, it has been suggested that the settlement was relocated once in every 20 40 years following the rate of land uplift (Núñez 2009). However, the archaeological material and radiocarbon dates from the most extensively studied village-like housepit site in Finland present contradicting 54

evidence. The housepits at the Kuuselankangas site, all located within ca. 1.5 metres in elevation, date to ca. 3700 2900 cal BC (Vaneeckhout 2009c; Mökkönen 2010). This means that the site remained inhabited irrespective of the regressive shoreline. This was most probably due to good shing conditions at the Kierikinkoski rapids, which were located next to the site before the building of the current hydroelectric power plants. Although it is clear that not all the sites followed the regressive shoreline, the relative dating based on shoreline displacement facilitates the observation of the main trends. From my point of view, the most fundamental characteristic of shoreline displacement chronology is that it provides an opportunity to study temporal changes without recourse to ceramic typologies (Mökkönen 2009), and thus to recognize the changes that have taken place within a certain period of time or within an archaeological culture manifested in a certain material. In Paper I, quite a radical change in the method of placing pithouses in the environment was observed to take place during the CW2 period. 10 This example shows that the changes in material culture and in other aspects of culture do not go hand in hand. Therefore, it is necessary to study different cultural aspects without interdependence, as contiguous levels of culture. The dialogue between the levels should be carried out, whenever possible, without any subordinate relationships. 5.3 Changes in houses The parallel development of the shape and size of pithouses all over northern Fennoscandia over a rather short period is an interesting phenomenon. There are some regional differences in the dating and duration of the development. In this chapter, the changes in housepits regarding size, shape, clustering, and site location are discussed further. In this study, I have demonstrated that the period of rapid growth in size began at different times in different areas, and also that the duration of the growth period differed from place to place. Likewise, the periods that represent an increase in the number of housepits vary. It seems to me that the different events growth in size, increase in number, decrease in number, and the period with the largest housepits appeared rst in south-eastern and later in north northwestern Fennoscandia (see Chapters 3.3 3.5). In gure 13, I have tried to illustrate the temporal changes in housepits in Fennoscandia based on the data presented in Chapters 3.3 3.4 and Papers I and IV. The absolute dating of certain developments in each of the areas is difcult due to the different nature of the data. However, it is possible to arrange different developmental stages roughly in chronological order. Obviously, the details, such as the size and number of housepits, vary between regions, but the main point here is to demonstrate the dating of the main trends in pithouse building. 10 In Paper I, however, the chronological horizons based on ceramics types are included alongside the chronological changes observed with help of shoreline displacement and elevation of the sites. At that time, it was still thought that ceramics types could really be applied as chronological horizons that did not overlap extensively 55

The dates are based for the most part on shoreline displacement studies (Paper I) combined with a number of radiocarbon dates (the developments around Gulf of Bothnia). In the Republic of Karelia, the chronology is based on conventional radiocarbon dates. On the Karelian Isthmus (Paper IV), the chronology is based for lack of anything better on associated ceramics types, a few radiocarbon dates, and general trends in pithouse building traditions in nearby areas. As it is with the beginning of the Corded Ware Culture in Finland (see Chapter 2.2.2), the chronological difculties relating to the old wood effect and conventional radiocarbon dates are present here, too. In addition, the settlement at some of the sites dating to the 3 rd millennium cal BC were no longer tightly shorebound and might therefore be dated too old by shoreline displacement chronology (see Chapter 5.5; Paper III). Figure 13 shows how the eastern part of the Gulf of Bothnia (Finland and Russia) differs from the western part (Sweden). It seems that the different trends in pithouse building appeared as a series of events following each other. Despite the discrepancies in the bases of the chronologies of the different areas, I believe that the main trend in which certain developments take place slightly earlier in south southeastern than in north northeastern Fennoscandia reects the correct direction. The only exception to the rule is the Stone Age estuary of the Kemijoki River in Tervola municipality, southern Finnish Lapland, where the actual growth in the size of housepits (which occurred elsewhere after CW2) is not observed. As compared to other changes affecting housepits, the shift to windy and aquatically oriented locations falls within the same period as the beginning of the peak periods in the number of housepits, northern Norway being an exception (Fig. 13). This is seen in case study areas around Lake Saimaa (Paper I), on the Karelian Isthmus (Paper IV), and in Norrbotten in Sweden. In coastal Norrland, the settlement concentrated in the archipelago around 3400/3000 2300 cal BC in the southern area and around 3000 1800 cal BC in northern Norrbotten. The trend was accompanied by a growing utilization of seals (Norberg 2008: 169-170). On the Finnish side of the Gulf of Bothnia, the period with an increase in the number of housepits is also linked to the intensive exploitation of seals (Halinen et al. 1998; Hertell 2009), even though a similar shift in site location towards the seaside as seen in Norrland is not observed, most probably because it has not been studied. Fig. 13 (next page). A rough chronological table showing the main trends in housepits and housepit sites in Fennoscandia. In Norrbotten column B, the stars mark the development in southern Norrland (S) and northern Norrland (N). * This is a rough estimation. A square with black outline in River Iijoki marks the appearance of the terrace house. The key to letters: A = increase in number of housepits/peak period B = beginning of maritime/aquatic orientation of housepit sites C = decrease in number of housepits D = beginning of rapid growth in housepit sizes E = largest housepits 56

4000 3000 2000 1000 cal BC Republic of Karelia, Russia A ( B) C D E Karelian Isthmus, Russia A B C D E Lake Saimaa area, Finland A B C D E River Iijoki, Finland A ( B) C D E River Kemijoki, Finland A ( B) C D E Norrbotten, Sweden A B C D E N Norway A B C D E?? none S N? * 4000 3000 2000 1000 cal BC The Beginning of the CORDED WARE/BATTLE AXE cultures ca. 2900/2800 cal BC 57

As with the other changes, the shift towards more aquatic site locations also took place in different environmental settings. On the Karelian Isthmus (Paper IV) and in the Lake Saimaa case study area (Paper I), the aquatically oriented and unsheltered housepit sites came predominantly into use around 3700 cal BC as a rough estimate. In southern Finland, the shift in housepit site location to islands did not take place only in larger inland lake areas. Smaller roundish housepits and larger, rectangular housepits surrounded by earth embankments are found also on islands in smaller lakes (Miettinen, T. 1998: 58; Kivimäki 2007; Lavento et al. 2007; Nordqvist 2007). These structures date to the Middle Neolithic, from Typical Comb Ware onwards. From coastal Ostrobothnia, there are only a few studies in which maritime site locations have been observed. For example, settlement in the Oulujoki River estuary moved to an island outside the river mouth ca. 2500 1800 cal BC (Ikäheimo 2002). The changes that befell housepits in the coastal areas as well as on large and small inland lakes provide insights into the nature of change, which obviously was not tied to specic environments. What is particularly interesting here is that the parallel developments in housepits extended over vast areas and into various cultures. The changes in the way pithouses were constructed and located were not limited to the sphere of a certain material culture complex. It seems quite justied to state that the development of societies as reected in the housepits was not brought on by environmental causes, as is proposed in the case of Northern Ostrobothnia (Núñez & Okkonen 1999; Vaneeckhout 2008a b; 2009 a c; Núñez 2009) and also in the Lake Saimaa area (Paper I). Environmental factors have certainly had an effect on the volume of the changes, but they were not the main reasons for the developments. My impression is that the changes in the pithouse building tradition were largely the result of applying new ideas of dwelling into practise, rather than merely societal responses to environmental causes (see also Paper III; Norberg 2008: 158 159, 179). In Paper III, I noted that the appearance of multi-room housepits was concurrent with the spread of the Corded Ware Culture to the north-eastern part of the Baltic Sea. I presumed that the multi-room housepits were dwelling structures that displayed the idea of a longhouse while the construction techniques followed the local pithouse building tradition. When writing that article, I did not question the traditional Finnish chronology, which dated the beginning of the Corded Ware Culture as early as 3200 cal BC. Now, when the actual beginning of the culture has been established at approximately 2900/2800 cal BC (see Chapter 2.2.2), the conclusion concerning the nature of multi-room housepits as dwellings displaying an idea similar to longhouses is possible, but the initial adoption of this idea from the Corded Ware Culture is improbable. In a number of cases introduced in Paper III, the dating of the sites is based on radiocarbon dates on wood charcoal. However, charred crust adhering to Pöljä Ware found in the terrace house at the Purkajansuo/Korvala site in Yli-Ii parish, northern Ostrobothnia, is dated to 3360 2930 cal BC (2 sigma, 4475±60 BP, Hela-136) (Schultz 2000). Likewise, the oldest radiocarbon dates of the Meskäärtty site date to the same period (Paper III). This gives a reason to believe that the rst appearance of multi-room houses actually pre-dates the earliest possible contact with the Corded Ware Culture. 58

Is it possible that the Neolithic ideas came rst, before the spread of the cultures that actually carried with them a Neolithic identity? In Norrland, Sweden, it has been proposed that Neolithic ideas were present around 2800 cal BC although the material culture exhibits only an ephemeral inuence of the Battle Axe Culture (Gustafsson 2007). In Norrbotten, the maritime orientation of the settlement, the increase in the number of housepits, the beginning of a rapid growth in housepit size, and the beginning of the Corded Ware/Battle Axe Cultures are temporally closer to each other than elsewhere in Fennoscandia. Christian Carpelan (2004) has demonstrated that the there were two peaks in the inuence of the Corded Ware/Battle Axe Cultures in northern areas. According to Carpelan, the rst peak at ca. 3000 2750 cal BC carried the material of the Corded Ware Culture as far north as the Arctic Circle in Finland, but not to modern Sweden. The rst peak cannot actually be quite so old, but it undoubtedly dates to the very beginning of the spread of the Corded Ware Culture to the North. During the second peak around 2500 2300 cal BC, the inuence of the Finnish Corded Ware Culture spread up to the Oulu area while regions further to the north received inuences from the sphere of the Swedish Battle Axe Culture (2900/2800 2400/2300 cal BC). Interestingly, northern Norway has also produced nds of imports from southern Scandinavian agricultural cultures (a battleaxe, arrows and int daggers) that are dating to the Gressbakken Phase (Olssen 1994: 91). Similarly, the inuence of the Corded Ware/Battle Axe Cultures and the largest housepits (and dwelling embankments and Giants Churches ) are in the main temporally and geographically concurrent phenomena around the Gulf of Bothnia. Now, after sifting through the main trends in the pithouse building traditions of Fennoscandia (Fig. 13), I believe that the periods of change were consequences of new cultural concepts that spread over northern Fennoscandia in two waves. The rst wave, the origin of which lies somewhere among the Comb Ware-related cultures ourishing south or south-east of Finland, disseminated a new idea of more sedentary settlement, probably supported by logistical mobility based on water transportation. This is reected in an increase in the number of housepits, often associated with village-like concentrations, and in a shift in the location of housepits in more aquatic environments than previously (Papers I and IV). This wave spread to northern Fennoscandia during the CW2 period, starting from the south-east and appearing a bit later in the north-west. It is also possible that the initial stage of cereal cultivation spread with this wave (Paper V). Because we do not know very much about the development of Stone Age water transportation, it is possible that no technical improvements occurred. In this case, the change in the rst wave was only about putting new ideas into practise. The second wave is associated with an increase in the size of pithouses and with more complex ground plans. The trend towards the largest pithouses all over Fennoscandia was a consequence of societal re-organization that followed the spread of Neolithic ideas, which had been introduced already some time before the northward spread of the Corded Ware/Battle Axe Cultures. This idea has been proposed be- 59

fore, as a cause for the cultural changes in Middle Norrland sites after 2800 cal BC (Gustafsson 2007: 321). The second wave spread from south to north. Although the beginning of this wave slightly preceded the appearance of the Corded Ware/Battle Axe Cultures on the eastern side of the northern Baltic Sea, it spread more full-blown Neolithic ideas than the rst wave, including new cultural concepts of what a dwelling should be. The second wave brought with it changes in society that favoured larger, more oblong or multi-room houses. At least in some cases, the locations of the pithouse changed to exhibit a less shorebound mode of settlement than before. These two waves met temporally in Norrland, Sweden. In Norrbotten, the contact with the Battle Axe Culture is presumed to have brought with it both agrarian elements and a re-organization of the settlement pattern. According to Norberg (2008: 202), this re-organization is observable as a decreasing number of housepits ca. 2300 cal BC, when other types of dwellings replaced pithouses. In Finland, as I have sought to illustrate here, the contacts between native cultures and Neolithic ideas took place already during an early phase, slightly before the Corded Ware/Battle Axe Cultures. After the arrival of the Corded Ware Culture, the inuence of the newcomer culture presumably coincided with the second wave. The contacts were followed by changes in the housepit building tradition. Erik Norberg (2008: 158 159, 178 179, 186)) presumes that the changes seen in Norrlandic and other Fennoscandian housepits during the 3 rd millennium cal BC are of social origin, and that one explanation for the parallel developments in vast areas might be found in cultural contacts maintained through contact networks. In Finland, the Neolithic cultural development by the Gulf of Bothnia has also been interpreted through the interaction of societies with different subsistence bases. Based on Ostrobothnian material, Jari Okkonen (2003; 2009) has interpreted the radical changes in building traditions ca. 3000 2500 cal BC ( Giant s Churches and the beginning of cairn building) to be the result of both an increase in social complexity resulting from economic growth and of new ideologies acquired through long-distance maritime connections with southern cultures that were already practicing agriculture. I agree with Norberg (2008: 178 179) that the reason for the parallel development of housepits in Fennoscandia is not explicable through the presence of certain resources. In addition, I have demonstrated that the parallel development of housepits took place earlier in the south-east and later in the north-west and north. Furthermore, I believe that the changes in housing were primarily caused by new ideas, new concepts of housing, the origin of which lies among the southern-southeastern Neolithic cultures. I believe that these new ideas were adopted into northern Fennoscandia beginning in the Typical Comb Ware period. As opposed to Okkonen s (2009) emphasis on direct contacts, I would submit that the trade networks were the instrument that transported the exotic materials as well as new ideas. Direct long-distance contacts were probably established later on, when we have more examples in the archaeological record of everyday objects travelling across the sea. I consider this to have taken place primarily during the 3 rd millennium cal BC. 60

5.5 Middle Neolithic sedensm An aquatic resource base provides a good foundation for a growing degree of sedentism, as noted in Chapter 4.1. In the Middle Neolithic, when an aquatic orientation is clearly detectable in the diet and in site location, there are also other traits that may point towards fully sedentary occupation at the housepit sites. According to Erica Engelstad (1990: 23), the growing signicance of a single dwelling site already indicates territoriality and a certain rate of sedentism. The concentration of the occupation to a few specic sites that were occupied either permanently or repeatedly on seasonal basis can be interpreted to mark the presence of a sedentary or semi-sedentary settlement pattern. In this respect, the question concerning the contemporaneity of housepits at one site, often raised in discussions of sedentism, is not the key issue. Sedentism can be manifested in long residential continuums. This is an interesting theme especially in Finland, where coastal areas are undergoing continual change due to the constant land uplift. In the Lake Saimaa case study area (Paper I), the growing signicance of particular places used for habitation began to be felt during the CW2 period. There, some of the village-like housepit sites that were occupied for the rst time during the CW2 period remained in use by later people with Kierikki and Pöljä asbestos wares (see also Karjalainen 1999; Halinen et al. 2002). In south-eastern Finland, the Meskäärtty site remained in use over a thousand years (Paper III). The results of the 2010 excavations showed that the occupation began earlier than had been supposed based on different ceramics types and radiocarbon dates (from 3300 to 2200 cal BC). The rst occupation phase at the site dates to the CW2 period (4000 3400 cal BC), when the site was located on an extremely small island. In the course of time, land uplift changed the site location to a spit by a small bay of the Baltic, and during the nal occupation phase, the site was located by a small lake. The change in attitudes towards regressive shorelines is seen also in Norrbotten, Sweden. After 2800 cal BC, a number of dwelling sites in Middle Norrland were in continuous use for longer periods. The occupation no longer followed the regressive shoreline as diligently as before (Gustafsson 2007; Runeson 2007). For example, at the Bjästamon housepit site in Middle Norrland, Sweden, the oldest occupation at the site, ca. 2800 cal BC, was located in the immediate vicinity of the beach, while during the youngest occupation phase ca. 2100 cal BC the distance between the shore and the settlement site was over one hundred metres (Spång 2007: 39). There, the change in the way the activities were carried out at the site was associated with large housepits and non-shorebound habitation. Similar observations have also been made at the Karlebotnbakken site in the Varangerfjord, Norway. There, a housepit dating to the Gressbakken phase (2200 1600 cal BC) is located far from the regressive shoreline, on a site where habitation commenced already ca. 3300 cal BC (Hood & Helama 2010). I cannot verify this statement properly, but I assume that the long continuum in habitation at certain dwelling sites has to do with a higher degree of sedentism and 61

the growing signicant of particular inhabited locations. I believe that the changes in settlement described above reect an altered attitude towards settlement practices. I will now return to the discussion on the maritime orientation of the resource base and the nature of habitation. In two papers, I have connected the changes in housepits and housepit sites to a growing degree of sedentism, logistical transportation based on watercraft, and a probable, albeit not yet veried, beginning of agricultural practices (Papers I, IV, and V). In this introduction paper, I have also illustrated how the aquatic orientation of settlement and/or an aquatic mode of subsistence occur together with the peak period in the number of housepits. It has been suggested that sedentary hunter-gatherer communities with effective water transportation could have integrated both domestic plants and animals into their economy without any radical dislocation of other aspects of their society (Bogucki 2004). This is an interesting suggestion, since the changes in housepit site locations to more unsheltered and aquatically oriented locations go together with the appearance of the rst signs of cultivation in coastal Central and Northern Sweden. Similar change occurred on the Karelian Isthmus (Paper IV) and in the Lake Saimaa area (Paper I) around the suggested, but not veried, beginning of cereal cultivation (Paper V). Unfortunately, there are no boat nds dating to the Stone Age in Finland. 11 The only direct references to boats are rock paintings in which the boat motifs are mostly found in the upper parts of the painting surfaces. In the Lake Saimaa area, the majority of boat motifs date to older than 3500 cal BC, i.e., to the CW2 period (Seitsonen 2005b). Although there is no concrete evidence of boats, they were undoubtedly present. I presume that the maritime diet and changes in site locations indicate an increasing importance of water transportation, and some technological innovations in watercraft might have been involved in the process. 5.6 At the northern limit of Corded Ware distribuon Because the interaction between the Corded Ware Culture and native cultural traditions are central in my interpretations, and because the contacts on the northern borderline of the Corded Ware culture have been touched in this study (Paper III, see Chapters 3.3.2 and 5.1), I will briey list a few observations concerning the interaction between the Corded Ware Culture and local cultural traditions. A certain degree of cultural hybridization and loans that most likely originate from interaction may be discerned around the northern limit of the Corded Ware Culture. In south-eastern Finland and on the Karelian Isthmus, Russia, the early beginning of contacts between the Corded Ware Culture and local cultural traditions apparently resulted in assimilation (Paper III, Carpelan et al. 2008). Similarly, in Southern Ostrobothnia, western Finland, the Corded Ware Culture and contemporaneous local pottery traditions are found in the same dwelling sites (Edgren 1997; Miettinen, M. 11 The only suggested candidate, the dugout found in Helsinki Aleksis Kiven katu, is probably of notably younger age (Mökkönen 2007; cf. Luho et al. 1956). 62

1998; Carpelan 2004) and some ceramics thought to represent hybrids of Corded Ware and Comb Ware traditions are present (Edgren 1997). The Corded Ware inuences on local pottery traditions, including proling and at bottoms, are present in Pöljä Ware (ca. 3500 2600/1900 cal BC) in the Finnish inland (O Caellacháin 2008) and in Palaiguba II pottery (ca. 2500 1500 cal BC) in the Republic of Karelia, Russia (Zhul nikov 1999). Likewise, the Middle Zone Pottery occurring in the Finnish inland close to the northern limit of the Corded Ware Culture exhibits elements from both the Corded Ware and the Comb Ware traditions (Carpelan 1979). In Eastern Sweden, the outcome of the hybridization of the Battle Axe and local Pitted Ware pottery traditions is called 3 rd group pottery and dated to the Middle Neolithic B (2800 2300 cal BC) (Larsson 2009: 356 ). Åsa Larsson (2009: 367) has suggested that this kind of pottery existed also in western and south-western Finland. The Bjästamon site (ca. 2800 2100 cal BC) with 3 rd group pottery in Middle Norrland is also of great interest (Lindholm et al. 2007; Larsson 2009: 365, Fig. 11.5). It is a housepit site with one two-room housepit that displays hybrid characteristics also in its construction (see Chapter 3.3.2, Fig. 12). One room with a horizontal timber framework suggests the northern pithouse building tradition, while the wall construction of the second room, which is made of upright poles, is known from the building tradition typical of Neolithic cultures further to the south. It seems that fairly analogous phenomena occur close to the northern limit of the distribution of the Corded Ware/Battle Axe Cultures all over Fennoscandia. It is interesting that contacts with the Corded Ware/Battle Axe Cultures evoke similar phenomena among different local cultural traditions. It seems that there is great variation in details in ceramics in each site, but the similar main trends are repeated from site to site with an emphasis on every site s particular characteristics. 5.7 Aer Typical Comb Ware: Populaon decline or merely cultural change? Several authors have proposed that the population of Finland declined during the Late Neolithic and the beginning of the Early Metal Period. This opinion is based on several observations: the decrease in the number of dwelling sites, low nd densities, thin cultural layers, a lack of substantial dwelling structures, a low number of radiocarbon dates, and a suggested population bottleneck in genetic studies ca. 1900 cal BC (Siiriäinen 1981; Sajantila et al. 1996; Okkonen 2003: 231; Tallavaara et al. 2010). It is also assumed that the Early Metal Period settlement was more mobile than the previous Neolithic settlement with housepits (Lavento 2001). The meagre available knowledge about the Late Neolithic material culture has even led some authors to suggest a discontinuity of settlement (Lavento 2001: 183). In this chapter, I propose an alternative hypothesis based on cultural changes, all of which had a negative impact on the visibility of archaeological materials. 63

It is clear that archaeological sites must be discovered before they can be studied, and that dwelling sites dating to certain periods are easier to nd than others. The archaeological visibility of the sites weakens radically after the decline in housepits. It is obvious that other type of dwellings replaced pithouses. The Rusavierto house gives a hint of this development. It is a dwelling with a timber frame, dating to 2300 1900 cal BC (Leskinen 2002), one end of which is slightly dug into a natural slope while other parts have been built above ground. I would suggest that the radical decline in the number of housepits is not the result of a decline in the number of dwellings per se, but a result of the shift from semi-subterranean dwelling structures to above-ground dwellings (see also Ikäheimo 2005), which obviously diminishes the archaeological visibility of dwelling structures detectable without excavations. Another change involves the prevailing settlement pattern. As noted above (Chapters 3.3 3.5), the increase in the size of housepits was already accompanied by a decrease in numbers, and during the period with the largest housepits, signs of a high degree of sedentism are already present. It is obvious that sedentary groups leave fewer archaeologically detectable and dateable sites than less sedentary groups do. A possible increase in indoor activities may also restrict the archaeological visibility outside the dwelling (see Chapters 3.4 3.5). There is a well-established opinion that Early Metal Period sites are more restricted in size than Stone Age dwelling sites (e.g. Lavento 1998). Janne Ikäheimo (2005) has noted that, due to this phenomenon, Bronze Age sites on the Oulujoki River have been discovered only in cases where the topsoil has been disturbed by modern land use activities. Further changes in material culture also took place in the course of the Neolithic. It seems to me that the CW2 period, with its great number of nd-rich sites, is used as a yardstick for evaluating the whole Neolithic Stone Age in Finland. If the material culture found in the excavations is poorer than the rich CW2 assemblages, the settlement at the site is thought to have been less pronounced. Furthermore, when we come to Early Metal Period sites, which are typically scarce in nds and have no easily detectable structures, the interpretations tend to diagnose a growing degree of residential mobility (Lavento 2001: 141 143). The decline of the Comb Ware tradition is followed by changes that indicate alterations in the functions of pottery. The size of vessels decrease, the residues adhering to shards of Kierikki Pöljä Wares point rather to cooking than to storage (Joensuu 2000), and storage pits appear (e.g. Vaneeckhout 2008a, Paper III). In this respect, there is also an interface between local cultural traditions and the Corded Ware tradition, where the pottery was not used for large-scale storage (e.g. Salo 1989: 11). The altered function of pottery and the reduced size of the pots results in fewer shards to be found. The suggested late Neolithic population decline has something to do with research history. I am of the opinion that the late Neolithic is so poorly studied as a period that the problem of the low number of sites may also be due to the inability to recognize late Neolithic material. Moreover, the Late Neolithic settlement was probably not as shorebound as it was during previous periods (Chapter 5.5), which makes them dif- cult to discover. 64

To sum up this short discussion, I would suggest that the following aspects collectively challenge the prevailing interpretation of a dramatic population decline: (1) a growing rate of sedentism (cf. Tallavaara et al. 2010), (2) a shift from semi-subterranean to above-ground dwellings, (3) a shift from outdoor to indoor activities, (4) changes in ceramics (smaller size, and altered function), (5) problems in identifying the material, and (6) the less-shorebound character of late Neolithic dwelling sites. These changes, which occurred mostly after the decline of the Comb Ware tradition, weaken the visibility of archaeological material. The latest paper dealing with the Neolithic population decline as seen through the probability distribution of radiocarbon dates (Tallavaara et al. 2010) admits to a lack of cultural processes in the analysis. Here, I have put forth my reasons for scepticism. My arguments are not bound to any environmental causes. I merely rely on general trends in cultural development and on the effect they have on archaeologists ability to identify late Neolithic material. The late Neolithic population decline holds true only if one believes that current archaeological knowledge on late Neolithic Stone Age and its material record is truly accurate. 65

66

6 CONCLUSIONS In this dissertation, I have tracked changes in the Neolithic pithouse building tradition in Fennoscandia. The chronological changes in the size, shape, environmental location, and clustering of housepits observed in the Lake Saimaa area, Finland, and on the Karelian Isthmus, Russia, are interpreted as marking a growing degree of sedentism (Papers I and IV). In Paper III, the multi-room pithouses that appeared during the last quarter of the 4 th millennium cal BC are interpreted as expressions of the idea of the long-house, borrowed from the contemporaneous Neolithic cultures further to the south. In this synthesis paper, I have explored the changes in the housepit building tradition, similar to those observed in the Papers, as they are to be seen in the Fennoscandian archaeological record on housepits. As a result, I have discovered that there are certain changes in size, shape, environmental location, and clustering of housepits that extended into various cultures and ecological zones in northern Fennoscandia. This dissertation argues for two mostly non-contemporaneous waves that spread certain changes in the pithouses themselves and in the practices of locating the dwellings in the environment/landscape. The beginning of the rst wave appears in the archaeological record during the early Typical Comb Ware period (4000 3400 cal BC, referred to as CW2). It is visible in the archaeological record as an increasing tendency to locate housepit sites in unsheltered and aquatically oriented locations. The environmental relocating of housepit sites was coincident with the appearance of village-like clusters of housepits, an increase in the number of housepits, a decrease in the number of other types of dwelling sites, the beginning of an increase in housepit size, and the increased signicance of dwelling sites as reected by longer continuums in occupation. I have interpreted these changes to reect a growing degree of sedentism. The changes connected with the rst wave appeared rst in the southeast and a bit later in the northwest. In the Lake Saimaa area, the changes took place in the middle of the CW2 period (Paper I), and on the western side of Gulf of Bothnia somewhat later. It is also possible that this rst wave brought with it the beginnings of cereal cultivation, although I do not consider even earlier experiments completely unthinkable (Paper V). The second wave of change introduced larger housepits with more oblong ground plans, multi-room houses, and in some cases less shore-bound occupation. On the eastern side of the Gulf of Bothnia, this wave probably precedes the beginning of the Corded Ware/Battle Axe Cultures (2900/2800 cal BC) by a few hundred years. On the western side, in Swedish Norrland, the changes of the rst and the second waves and the beginning of inuences from the Battle Axe Culture appeared simultaneously 67

around 2800 cal BC. Current data allows us to presume that ideas connected to truly Neolithic cultures might have spread to the North somewhat earlier than the material culture of the Corded Ware/Battle Axe Cultures. The waves I refer to here are not zones that spread everywhere. It is likely that the incorporation of new ideas might have been similar to the adoption of early agriculture in Central Europe. There, agriculture spread as scattered cells, and consequently there were groups living within a sphere of similar material culture but with different subsistence practices (see Paper V). I presume that something like this happened also in north northeastern Fennoscandia. It is likely that some groups were more receptive to new inuences than others were. Therefore, not all groups adopted the new ideas straight away. The evolution of housepits around the Bothnian Bay has previously been interpreted to have been caused by the adaptation of the local societies to prevailing environmental circumstances (Núñez & Okkonen 1999; 2005; Núñez 2004; 2009a; Okkonen 2003; Vaneeckhout 2008b; 2009a, c; 2010), and also by re-organization following contacts with the agrarian Corded Ware/Battle Axe Cultures spreading to North (Norberg 2008: 202). Erik Norberg (2008: 158 159, 178 179, 186) has proposed that on a Nordic scale, the developments in housepits were parallel and concurrent, and that the resource-based explanation was therefore very unlikely. He presumes that the changes were of social origin, and that the explanation for the development might be found in cultural contacts maintained through contact networks. It appears to me that the changes in the pithouse building tradition had more to do with applying new ideas of dwelling into practise than with societies merely responding to environmental stimuli (see also Norberg 2008: 158 159, 179; Paper III). Like Norberg (2008), I also propose that the ideas spread via the same trade networks that transported exotic material around Fennoscandia. In contrast to Norberg, I would submit that the spread of new ideas was not just a short concurrent episode but rather a long process with a varying speed of dispersion which began already during the CW2 period (4000 3400 cal BC). In Paper V and in the last section of this synthesis paper (Chapter 5), I have wished to take a closer look at a few cultural changes and contacts dating to the studied period. The results of Paper V showed that there was no unquestionable palynological or botanical evidence of cereal cultivation in Finland dating as early as the beginning of the changes in housepits during the CW2 phase (4000 3400 cal BC). What seems to be clear, however, is that the Corded Ware Culture was not the source of initial cereal cultivation. East of the Baltic Sea, cereal cultivation was practiced among cultures previously labeled Subneolithic or Forest Neolithic more than a thousand years before the emergence of the Corded Ware Culture. Therefore, it seems that the origin of the rst northern Fennoscandian cereal cultivation was to be found in eastern connections. This means that the contacts through which the knowledge of cultivation was adopted by the cultures inhabiting the north-eastern European forest zone took place in an area with active trade/social networks. In Chapter 5, I have briey discussed a few topics involved with cultural change, with emphasis on the 3 rd millennium cal BC. During this millennium, the number and 68

size of housepits rst peaked, and then during latter part, the number of housepits drastically decreased. Many of the discussed topics are still rather poorly explored. It appears, however, that the contacts between the Corded Ware/Battle Axe Cultures and local cultures commenced immediately, and the outcome of this interaction seems to have produced highly similar phenomena all around the northernmost limit of the distribution of the Corded Ware/Battle Axe cultures. Near that limit, there are signs of cultural loans and hybridization both in ceramics and in dwelling structures. Concurrently, there are some observations pointing to a less shore-bound mode of habitation and an intensication of indoor activities at the expense of outdoor activities. The changes in housing and in material culture that took place during the 3 rd millennium cal BC all had an effect on the archaeological visibility of dwelling sites. I am rather doubtful about the suggested Late Neolithic population decline in Finland. I would suggest that the small number of Late Neolithic sites and radiocarbon dates falling around the nal Neolithic Stone Age has more to do with cultural changes that weaken the archaeological visibility of the occupation, and with the (in)ability of archaeologists to identify sites of that period. From a Neolithic viewpoint, north-eastern Fennoscandia has always been a marginal area, at least when agriculture is considered a key factor. In this region, the Neolithic ideas were incorporated gradually. This process began during the Early Comb Ware phase, when the rst ceramics came into use around 5500 5100 cal BC. The efciency of the transportation routes in the north-eastern European forest zone is evident in the rapid tempo of the spread of early pottery technology (Carpelan & Parpola 2001: 63; Jordan 2010). During the Neolithic Stone Age, these networks spread artefacts and undoubtedly also ideas to vast areas, even to the most northern parts of Fennoscandia by the north Atlantic and the Arctic Ocean (see e.g. Torvinen 2000; Ramstadt et al. 2005; Damm 2008; Hood & Helama 2010). In this dissertation, I have argued for the presence of ideas and possibly technological innovations that spread over northern Fennoscandia and across different cultures and ecological zones. This study has revealed two waves of change in the housepit building tradition, both of which dispersed from the south south-east to the north north-west. Since the CW2 period, the ideas that brought with them the changes spread through the same networks that actively distributed commodities, exotic goods, and raw materials over vast areas between the southern Baltic Sea, the north-west Russian forest zone, and Fennoscandia. As I see it, the abandonment of pithouses during the Late Neolithic, which is most probably due to the replacement of pithouses by aboveground dwellings, was merely the nal phase of the developments in which the new concepts of housing were incorporated. 69

References Unpublished sources Excavaon and survey reports Enqvist, J. 2006. Hamina, osa I. Esihistoriallisten muinaisjäännösten inventointi 15.5. 9.6.2006. Survey report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Enqvist, J. 2007. Virolahti. Osa I: Esihistoriallisten muinaisjäännösten inventointi 3.9. 28.9. 2007. Survey report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Lavento, M., Nordqvist, K., Kivimäki, S., Lahelma, A., Miettinen, T., Mökkönen, T., Perttola, W., Pylkkö, I. & Seitsonen, O. 2007. Repoveden kansallispuiston ja Aarnikotkan metsän luonnunsuojelualueen inventointi 21. 25.8. 2006. Survey report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Lehtinen, L. 1987. Jäppilän muinaisjäännösten inventointi 1987. Survey report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Mökkönen, T. 1999. Kerimäen, Kesälahden, Kiteen ja Savonrannan osainventointi 1998. Saimaa-Laatokka projektin inventointi. Survey report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Schulz, H.-P. 2000. Yli-Ii Purkajansuo/Korvala. Excavation report. Excavation report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Sepänmaa, T. 1991. Kerimäen muinaisjäännösinventointi vuonna 1991. Survey report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Sepänmaa, T. 1995. Savonrannan inventointi 1995. Survey report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Vuoristo, K. 2009a. Hamina Virolahti 2009. Valtatien 7 tielinjan inventointi välillä Hamina Vaalimaa 15. 19.6.2009. Survey report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Vuoristo, K. 2009b. Virolahti Karpankangas. Kivikautisen asuinpaikan koekaivaus 7. 18.9. 2009. Excavation report. Dept. of Archaeology, Finnish National Board of Antiquities, Helsinki. Academic theses and term papers Carpelan, C. 1965. SÄR 2. Alustava katsaus erääseen rautakautiseen keramiikkaryhmään ja siihen liittyvään problematiikkaan. Licentiate thesis. Department of Archaeology, University of Helsinki. Joensuu, J. 2000. Kerimäen Raikuun Martinniemi 3:n keramiikka-aineisto. MA thesis. Department of Archaeology, University of Helsinki. Kotivuori, H. 2002. Alisen Kemijoen kivikautiset asumuspainanteet topograseen havainnointiin ja aineistovertailuun perustuva asutuskuva. Licentiate thesis. School of Cultural Research, Department of Archaeology, University of Turku. Kylli, J. 2000. Maatalouden alku Espoon ympäristössä: asutusmalli toimeentulostrategian 70

kuvastajana. MA thesis. Department of Archaeology, University of Helsinki. Mökkönen, T. 2000. Saimaan vesistöalueen pyyntikulttuurien toimeentulo ja asutusmallit: Kerimäen tapaustutkimus. MA thesis. Department of Archaeology, University of Helsinki. O Caellacháin, S. 2008. Pöljä keramiikkaryhmä vai reunamuoto?. Term paper, presented in seminar at the Department of Archaeology, University of Helsinki, 14.4. 2008. Pesonen, P. 1995. Varhainen asbestikeramiikka. MA thesis. Institute for Cultural Research, Archaeology, University of Helsinki. [the same manuscript was approved as a Licentiate thesis in 1996] Sartes, M. 1991. Kivikauden toimeentulostrategiat ja asutussysteemit. Turun alueen kivikautisten asuinpaikkojen topogra an ja keskinäisten suhteiden analyysi. MA thesis. School of Cultural Research, Department of Archaeology, University of Turku. Saukkonen, J. 1990. Kampakeraaminen asutus Närvijoen alueella. MA thesis. Department of Archaeology, University of Helsinki. Published Posters GTK 1996. Suursaimaa ja Muinaspäijänne laajimmillaan. Geological Survey of Finland, Espoo. [published poster] InterNet sources Furholt, M. von 2003. Absolutchronologie und die Entstehung der Scnurkeramik. 28 pages. <www.jungsteinsite.de> Hallgren, F. 2010. A short note about Stone Age farmers who did not adopt elk hunting, and elk hunters who did not adopt farming. In W. Østreng (ed.), Transference. Interdiciplinary Communications 2008/2009. Centre for Advanced Study, Oslo. InterNet publication. 12 pages. <http://www.cas.uio.no/publications_/transference. php> Jordan, P. 2010. Understanding the spread of innovations in prehistoric social networks: new insights into the origins and dispercial of early pottery in Northern Eurasia. In W. Østreng (ed.), Transference. Interdiciplinary Communications 2008/2009. CAS, Oslo. InterNet publication. 8 pages. <http://www.cas.uio.no/publications_/transference. php> Niemi, A. R. 2010. Life by the shore. Maritime dimensions of the Late Stoen Age, Arctic Norway. In W. Østreng (ed.), Transference. Interdiciplinary Communications 2008/2009. Centre for Advanced Study, Oslo. InterNet publication. 12 pages. <http:// www.cas.uio.no/publications_/transference.php> Pesonen, P. 1999b. Suomen esihistoriallinen keramiikka. InterNet pages for teaching Finnish prehistoric pottery. University of Helsinki, Helsinki. <http://www.helsinki./ hum/arla/keram/> (Read 28.12.2009) Ramstad, M., Hesjedal, A. & Niemi, A. R. 2005. The Melkøya project: maritime hunter- 71

sher island settlements and the use of space through 11 000 years on Melkøya, Arctic Norway. Antiquity Vol 79, No 304 (June 2005). < http://antiquity.ac.uk/projgall/ ramstad/index.html>, (Read 10.2.2010) Literature Ailio, J. 1909. Die Stenzeitlichen Wohnplatzfunde in Finland I-II. Helsingfors. Ailio, Julius 1915. Die geographische Entwicklung des Ladogasees in postglazialer Zeit und ihre Beziehung zur steinzeitlichen Besiedelung. (Bulletin de la Commission gé ologique de Finlande 45). Suomen geologinen komissioni, Helsingfors. Ailio, J. 1923. Ovatko Pohjanmaan jättiläislinnat muinaisjäännöksiä? Suomen Museo XXIX, 1 19. Ames, K. M. 2002. Going by boat. The foragercollector continuum at sea. In B. Fitzhugh & J. Habu (eds.), Beyond Foraging and Collecting. Evolutionary Change in Huntergatherer Settlement Systems. Kluwer/Plenum Press, New York, 19 52 Ames, K. M. 2003. The Northwest Coast. Evolutionary Anthropology 12, 19 33. Asplund, H. 1997. Kemiön suurpitäjän esihistoria. Kemiön suurpitäjän historia 1. Sagalundin museon kuntayhtymä, Kemiö, 213 82. Asplund, H. 2004. Problems of pre-roman Iron Age radiocarbon dating an example from SW Finland. In P. Uino (ed.), Fenno-Ugri et Slavi 2002. Dating and Chronology. (Museoviraston arkeologian osaston julkaisuja 10). Museovirasto, Helsinki, 9 14. Äyräpää, A. 1930. Die relative Chronologie der steinzeitlichen Keramik in Finnland I-II. Acta Archaeologica I. 165 190, 205 220. Binford, L. R. 1980. Willow smoke and dogs tails: hunter-gatherer settlement systems and archaeological site formation. American Antiquity 45(1), 4 20. Binford, L. R. 1990. Mobility, housing, and environment: a comparative study. Journal of Anthropological Research 46(2), 119 52. Bjerck, H. B. 1991. Boreal forangers at Vega Northern Norway. A study of site types and settlement patterns. In O. Grøn, E. Engelstad & I. Lindblom (eds.), Social space. Human spatial behaviour in dwellings and settlements. Odense University Press, Odense, 123 133. Björk, N, 2003. The Neolithic coastal settlements cosmology and ideology in a Baltic Sea perspective. In C. Samuelsson & N. Ytterberg (eds.), Uniting Sea. Stone Age Societies in the Baltic Sea region. (Occasional Papers in Archaeology 33). Uppsala universitet, Uppsala, 20 42. Bogucki, P. 2004. Observations from a distance on changing Stone Age societies in Central Sweden. In H. Knutsson (ed.), Coast to Coast: Arrival. Results and Re ections. Uppsala University, Uppsala, 369 376. Carpelan, C. 1979. Om asbestkeramikens historia I Fennoscandien. Finskt Museum 85, 5 25. Carpelan, C. 1999. Käännekohtia Suomen esihistoriassa aikavälillä 5100 1000 ekr. In Ed. P. Fogelberg. Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. (Bidrag till kännedom av Finlands natur och folk 153). Suomen Tiedeseura, Helsinki, 249 280. Carpelan, C. 2002. Esihistorian vuosiluvut, ajoitukset ja kronologia. In R. Grünthal (ed.), Ennen muinoin. (Tietolipas 180). Suomalaisen Kirjallisuuden Seura, Helsinki, 18 72

27. Carpelan, C. 2004. Corded Ware culture in Northern Finland. In M. Lavento (ed.), Early in the North, Volume 5. (Iskos 13). Archaeological Society of Finland & Finnish Antiquarian Society, Helsinki, 47 62. Carpelan, C. & Parpola, A. 2001. Emergence, contacts and dispersal of Proto-Indo- Europeans, Proto-Uralic and Proto-Aryan in archaeological perspective. In C. Carpelan, A. Parpola & P. Koskikallio (eds.), Early Contacts between Uralic and Indo-European: Linguistic and Archaeological Considerations. (Suomalais-ugrilaisen seuran toimituksia, 242.) Suomalais-ugrilainen seura, Helsinki, 55 150. Carpelan, C., Uino, P. & Gerasimov, D. V. 2008. Archaeology in the former municipality of Johannes. In K. Nordqvist (ed.), Karelian Isthmus: Stone Age Studies in 1998 2003. (Iskos, 16.) Finnish Antiquarian Society, Helsinki, 185 211. Costopoulos, A. 2005. Two multi-room Neolithic dwellings from Kierikki: Surprising likeness, thought provoking differences. Faravid 29, 5 17. Damas, D. (ed.) 1984. Handbook of North American Indians. Volume 5, Arctic. Smithsonian Instutution, Washington. Damm, C. 2008. Heterogene nettverk I det nordlige Fennoskandia. In P. Jensen, S.Sindbæk & H. Vandkilde (eds.), IX Nordic TAG. Globalization, identity, material culture and archaeology. Aarhus Univeristy, Aarhus. 12 16. Edgren, T. 1964. Jysmä i Idensalmi. En boplats med asbestkeramik och kamkeramik. Finskt Museum LXXI (1963), 13 37. Edgren, T. 1966. Jäkärlä-gruppen. En västnsk kulturgrupp under yngre stenålder. (Suomen muinaismuistoyhdistyksen aikakauskirja 64.). Suomen muinaismuistoyhdistys, Helsinki. Edgren, T. 1993. Den Förhistoriska tiden. In M. Norrback (red.), Finlands historia 1. Schildts Förlags, Esbo, 9 270. Edgren, T. 1997. Då snör- och kamkeramiker möttes i Österbotten. in K. Gulberg (red.), Arkeologi i Mittnorden. (Acta antiqua Ostrobotniensia, Studier i Österbottens förhistoria, 4.) Scriptum, Vasa, 151 174. Edgren, T. 1999. Käännekohtia Suomen kivikaudessa. In P. Fogelberg (ed.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. (Bidrag till kännedom av Finlands natur och folk 135). Finska vetenskaps-societeten, Helsinki, 281 93. Engelstad, E. 1984. Diversity in arctic maritime adaptations. An example from the Late Stone Age of Arctic Norway. Acta Borealia 2/1984, 3 24. Engelstad, E. 1988. Pit-houses in Arctic Norway. An investigation of their typology using multiple correspondence analysis. In T. Madsen (ed.), Multivariate archaeology: Numerical approaches to Scandinavian archaeology. (Jutland Archaeological Society Publications 84, 1988). Århus University press, Århus, 71 86. Engelstad, E. 1990. The meaning of sedentism and mobility in an archaeological and historical context. Acta Borealia 2/1990, 21 35. Engelstad, E. 1991. Book review. Prehistoric Hunter-Fishers of Varangerfjord, Northestern Norway: Reconstruction of settlement and subsistence during the Younger Stone Age. M. A. P. Renouf. BAR International Series 487. Arctic anthropology, Vol. 28, No. 2, 141 143. Fitszhugh, W. 1972. Environmental Archeology and Cultural Systems in Hamilton Inlet, Labrador: A Survey of the Central Labrador Coast from 3000 B.C. to the Present. 73

(Smithsonian Contributions to Anthropology, Number 16). Smithsonian Institution Press, Washington. Forss, A. 1981. Pattijoen Pikkuliekokankaan jätinkirkon kaivaukset kesällä 1979. Alustava raportti. Faravid 4. 77 86. Furholt, M. von 2008. Erscheinungen asynchroner kultureller Entwicklung am Übergang vom Spät- zum Endneolithikum in Mitteleuropa. Eine Untersuchung der Siedlungsfunde mit Schnurkeramik. In W. Dörer & J. Müller (eds.), Umwelt Wirtschaft Siedlungen im dritten vorchristlichen Jahrtausend Mitteleuropas und Südskandinaviens. (Offa Bücher 84). Untersuchungen aus dem Schleswig-Holsteinischen Landesmuseum fü r Vor- und Frü hgeschichte in Schleswig, dem Landesamt fü r Vor- und Frü hgeschichte von Schleswig-Holstein in Schleswig & dem Institut fü r Ur- und Frü hgeschichte an der Universität Kiel. Neumünster, 9 34. Gilman, P. 1987. Architecture as artefact: pit structures and pueblos in the American Southwest. American Antiquity 52(3), 538 64. Gjessing, G. 1941. Fangstfolk. Et streiftog gjennom nord-norsk førhistori. H. Aschenhoug & co, Oslo. Gjessing, G. 1942. Yngre steinalder i Nord-Norge. (Instituttet for sammenlignende kulturforskning. Serie B, Skrifter 39). Aschehoug, Oslo. Gjessing, G. 1945. Norges Steinalder. Norsk arkeologisk selskap, Oslo. Gjessing, G. 1975. Maritime adaptations in northern Norway s prehistory. In W. Fitzhugh (ed.), Prehistoric Maritime Adaptations of the Circumpolar Zone. Mouton, Hague, 87 100. Gustafsson, P. 2007. Ordning, oordning, nyordning. In P. Gustafsson & L.G. Spång (eds.), Stenålderns stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/Härnösand, 309 325. Halén, O. 1994. Sedentariness during the Stone Age of Northern Sweden in the light of the Alträsket site, c. 5000 B.C. and the Comb Ware site Lillberget, c. 3900 B.C. Source Critical Problems of Representativity in Archaeology. (Acta Archaeologica Lundensia, Series in 4:o, No. 20). Almqvist & Wiksell international, Stockholm. Halinen, P. 1997. Kaustisen Kankaan asuinpaikka ja punamultahaudat. Muinaistutkija 2/1997, 18 27. Halinen, P., Joensuu, J., Lavento, M. & Martio, L. 2002. House pit studies at Martinniemi in Kerimäki. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 201 210. Halinen, P., Joensuu, J., Lavento. M, Martio, L. & Mökkönen, T. 2003. Kerimäen Raikuun Martinniemen tutkimukset 1998 1999. Sihti 5, 9 26. Halinen, P., Katiskoski, K. & Sarkkinen, M. 1998. Yli-Iin Kuuselankankaan tutkimukset 1994 1996. In H. Ranta (ed.), Kentältä poimittua: kirjoitelmia arkeologian alalta 4. (Museoviraston arkeologian osaston julkaisuja 7) Museovirasto, Helsinki, 24 40. Halinen, P. & Mökkönen, T. 2009. Between Lake and Sea Stone Age Settlement by Ancient Lake Ladoga on the Karelian Isthmus. Fennoscandia archaeologica XXVI, 107 132. Halinen, P., Seitsonen, O., Seitsonen, S., & Nordqvist, K. 2008. Excavations at the Juoksemajärvi Westend Stone Age dwelling site in 2002. In K. Nordqvist (ed.), Karelian Isthmus Stone Age in 1998 2003.(Iskos 16). Finnish Anquarian Society, Helsinki, 233 265. 74

Hayden, B. 1997. The Pithouses of Keatley Creek: Complex Hunter-Gatherers of the Northwest Plateau. Harcourt Brace Collage Publishers, Fort Worth. Helm, J. 1981 (ed.). Handbook of North American Indians. Volume 6, Subarctic. Smithsonian Institution, Washington. Helskog, E. T. 1983. The Iversfjord locality: A study of behavioral patterning during the late Stone Age of Finnmark, North Norway. (Tromsø museums skrifter 19). Universitet I Tromsø, Tromsø. Helskog, K. 1984. The younger Stone Age settlements in Varanger, North Norway. Acta Borealia 1/1984, 39 70. Hertell, E. 2009. Ekologiat ja maatalouden alku Suomessa. In T. Mökkönen & S.-L. Seppälä (eds.), Arkeologipäivät 2008. Sosiaaliarkeologiaa yhteisöt arkeologisen aineiston taustalla & Tutkimushistorian painolasti. Suomen arkeologinen seura, Helsinki, 7 19. Hesjedal, A., Damm, C., Olsen, B. & Storli, I. 1996. Arkeologi på Slettnes. Dokumentasjon av 11.000 års bosetning. (Tromsø Museums skrifter XXVI). Tromsø Museum, Tromsø. Hiekkanen, M. 1984. Otliitel nye osobennosti postroek tipa Madeneva, otnosâihsâ k kamennomu veku. Madenevan tyypin kodanpohjat eräitä viitteitä kotakuntien perinne-eroista kivikaudella. In V. A. Rybakov (ed.), Novoe v arheologii SSSR i Finljândii. Nauka, Leningrad, 46 53. Hodgetts, L. 2010. Subsistence diversity in the Younger Stone Age landscape of Varangerfjord, northern Norway. Antiquity 84, 41 54. Hoffman, B. W. 1999. Agayadan Village: Household archaeology on Unimak Island, Alaska. Journal of Field Archaeology Vol. 26, No. 2, 147 161. Holback, T. J. 2007. En dare som bygge sitt thus på sand. In P. Gustafsson & L.G. Spång (eds.), Stenålderns stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/Härnösand, 161 185. Hood, B. C. 1995. Circumpolar comparison revisited: hunter-gatherer complexity in the North Norwegian Stone Age and the Labrador Maritime Archaic. Arctic Anthropology Vol. 32, No. 2, 75 105. Hood, B. C. & Helama, S. 2010. Karlebotnbakken reloaded; Shifting the chronological signicance of an iconic Late Stone Age site in Varangerfjord, North Norway. Fennoscandia archaeologica XXVII. 35 43. Huurre, M. 1984. Kainuu from the Stone Age to the Bronze Age. Finds and cultural connections. Fenno-Ugri et Slavi 1983. (Iskos 4.). The Finnish Antiquarian Society, Helsinki, 42 50. Huurre, M. 1986. Esihistoria. In M. Huurre & J. Keränen (eds.), Kainuun historia 1. Kainuun maakuntaliitto, Kajaani, 5 192. Huurre, M. 1990. 9000 vuotta Suomen esihistoriaa (4 th edition). Otava, Helsinki. Huurre, M. 2003. Viipurin läänin kivikausi. In M. Saarnisto (ed.), Viipurin läänin historia 1. Karjalan synty. Karjalan kirjapino, Lappeenranta, 151 244. Ikäheimo, J. 2002. Prehistoric dwelling remain in the lower Oulujoki River valley. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 129 136. Ikäheimo, J. 2005. Re-assessing the Bronze Age of coastal Northern Ostrobothnia the lower Oulunjioki river valley. In J. Goldhahn (red.), Mellan sten och järn. Dell 75

II.(Gotarc Serie C. Arkeologiska Skrifter No 59). Göteborgs universitet, Institutionen för arkeolog, Göteborg, 771 784. Ingold, T. 2000. The perception of the environment. Essays in livelihood, dwelling and skill. Routledge, London & New York. Itkonen, T. I. 1948. Suomen lappalaiset vuoteen 1945. Ensimmäinen osa. Werner Söderström osakeyhtiö, Porvoo. Jochelson, W. 1975 [1908]. The Koryak. (Memoirs of the American Museum of Natural History, vol. 10.) AMS Press, New York. Jones, B. 2002. Archaeological theory and scientic practice. Cambridge University Press, Cambridge. Jussila, P. 1992. Pieksamäen kasvijäännetutkimuksia. Sihti 2, 30 35. Jussila, P. 1994. Studies of plant remains at two Stone Age dwelling sites in Pieksämäki, Eastern Finland. In P. Purhonen (ed.), Fenno-Ugri et Slavi 1992. Prehistoric economy and means of livelihood. (Museoviraston arkeologian osaston julkaisuja 5). Finnish National Board of Antiquities, Helsinki, 16 22. Jussila, T. 1999. Saimaan kalliomaalausten ajoitus rannansiirtymiskronologian perusteella. In P. Kivikäs (ed.), Saimaan ja Päijänteen alueen kalliomaalausten sijainti ja syntyaika. (Kalliomaalausraportteja 1). Kopijyvä kustannus, Jyväskylä, 111 133. Jussila, T., Lehtinen, L., Nurminen, T., Raike, E. & Sepänmaa, T. 1992. Etelä-Savon muinaisjäännösinventointi 2000-projekti. Sihti 2, 20 29. Kankaanpää, J. 2002. The House Pits at Kauvonkangas, Tervola. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 65 77. Kankaanpää, J. 2003. Stone Age abandoment studies in Tervola, southern Lapland, Finland. In S. Samuelsson & N. Ytterberg (eds.), Uniting sea. Stone Age Societies in the Baltic Sea Region. (Occational Pappers in Archaeology 33). Uppsala univeristet, Uppsala. 103 116. Karjalainen, T. 1996a. Outokumpu Sätös ja Orov Navolok 16, talo 3. Muinaistutkija 1/1996, 13 18. Karjalainen, T. 1996b. Pithouse in Outokumpu Sätös excavated in 1992 1994. In T. Kirkinen (ed.), Pithouses and Potmakers in Eastern Finland. Reports of the Ancient Lake Saimaa Project. (Helsinki papers in archaeology No. 9). University of Helsinki, Finland, 71 88. Karjalainen, T. 1999. Sedentariness and dating Stone Age houses and sites. In M. Huurre (ed.), Dig It All. Papers dedicated to Ari Siiriäinen. The Finnish Antiquarian Society & The Archaeological Society of Finland, Helsinki, 185 200. Karjalainen, T. 2002. Comparisons between the artefact assemblages of six Neolithic houses. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 42 52. Katiskoski, K. 2002. The semisubterranean dwelling at the Kärmelahti site in Puumala, Savo province, Eastern Finland. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 171 200. Kelly, R. L. 1992. Mobility/sedentism: concepts, archaeological measures, and effects. Annual Review of Anthropology 21, 43 66. Kelly, R. L. 2007. The Foraging Spectrum: Diversity in Hunter-gatherer Lifeways. 76

Percheron Press, New York. Kent, S. 1990a. Activity areas and architecture: an interdiciplinary view of the relationship between use of space and domestic bult environments. In S. Kent (ed.), Domestic architecture and the use of space. An interdisciplinary cross-cultural study. Cambridge University Press, Cambridge, 1 8. Kent, S. 1990b. A Cross-cultural study of segmetation, architecture, and use of space. In S. Kent (ed.), Domestic architecture and the use of space. An interdisciplinary crosscultural study. Cambridge University Press, Cambridge, 127 151. Kivimäki, S. 2007. Esihistoriallisen ajan elinkeinot. In M. Lavento & A. Lahelma (eds.), Sama maisema, eri kulkijat Repoveden kansallispuisto kivikaudelta 1900-luvulle. (Metsähallituksen luonnonsuojelujulkaisuja. Sarja A 165: 52 53). Metsähallitus, Helsinki, 54 59. Koivunen, P. 1996. Kodanpohjia ja meripihkaa Kierikistä. Muinaistutkija 1/1996, 2 7. Koivunen, P. 1997. Teoria jätinkirkkojen käyttötarkoituksesta. Muinaistutkija 4/1997, 49 52. Kosmenko, M.G. 2004. The chronology of the Stone Iron Ages of the Karelian Republic. In P. Uino (ed.), Fenno-Ugri et Slavi 2002. Dating and Chronology. (Museoviraston arkeologian osaston julkaisuja 10). Museovirasto, Helsinki, 46 55. Kotivuori, H. 1993. Kivikauden asumuksia Peräpohjolassa. In R. Huopainen (ed.), Selviytyjät: näyttely pohjoisen ihmisen sitkeydestä. (Lapin maakuntamuseon julkaisuja 7). Lapin maakuntamuseo, Rovaniemi, 120 160. Kotivuori, H. 1996a. Pyytäjistä kaskenraivaajiksi. Rovaniemen asutus n. 6000 ekr. 1300 jkr. In M. Saarnisto et al. (ed.), Rovaniemen historia vuoteen 1721. Kotatulilta suvupirtin suojaan. Rovaniemi, Rovaniemi, 34 125. Kotivuori, H. 1996b. Mietteitä alisen Kemijoen kivikauden asuinmuistoista. Muinaistutkija 1/1996, 8 12. Kotivuori, H. 1998. Försänkta husgrunder från yngre stenåldern i nedre Kemiälven, norra Finland. In O. Kyhlberg (ed.), Hus och tomt i Norden under förhistorisk tid. (Bebyggelsehistorisk tidskrift Nr. 33). Stockholms universitet, Stockholm. 153 154. Kylli, J. 2001. Asutussysteemi ja toimeentulo muinaisessa Espoossa ja lähiympäristössä. Muinaistutkija 1/2001, 2 13. Larsson, Å. M. 2009. Breaking & Making Bodies and Pots. Material and ritual practicies in Sweden in the third millenium BC. (Aun 40). Uppsala Universitet, Uppsala. Lavento, M. 1998. Sisämaan vanhemman metallikauden väestö tutkimusongelmana. Muinaistutkija 4/1998, 46 55. Lavento, M. 2001. Textile ceramics in Finland and on the Karelian Isthmus. Nine variations and fugue on a theme of C. F. Meinander. (Suomen Muinaismuistoyhdistyksen Aikakauskirja 109). Suomen muinaismuistoyhdistys, Helsinki. Lavento, M. 2002. Milloin keramiikkatyyppi katoaa? ajoituksia ja ajoituksen losoaa. In P. Halinen, J. Kankaanpää & P. Pesonen (eds.), Arkeologipäivät 2001. Kronologia. Suomen arkeologinen seura, Helsinki, 39 45. Lavento, M. 2006. Kulttuuri ja kulttuurit suomalaisen arkeologian silmin katsaus 1950-luvulta 1990-luvun alkuun. In P. Pesonen & T. Mökkönen (eds.), Arkeologipäivät 2005. Arkeologia ja kulttuuri & Uutta kivikauden tutkimuksessa. The Archaeological society of Finland, Helsinki, 6 17. Lavento, M. 2008a. Archaeological research in the Saimaa district and in the Karelian 77

Isthmus in 1992-1999. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003. (Iskos 16). The Finnish Antiquarian Society, Helsinki, 26 45. Lavento, M. 2008b. The Saimaa Ladoga project: question framing and strategies. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003. (Iskos 16). The Finnish Antiquarian Society, Helsinki, 46 70. Lavento, M., Halinen, P. & Mökkönen, T. 2006. Subsistence strategies and changes of communities between 9000 1BC: an archaeological intensive-investigation in the western part of Lake Ladoga, on the Karelian Isthmus (Kaukola Räisälä project). In Herva, V.-P. (ed.), People, material culture and environment in the north. Proceedings of the 22nd Nordic Archaeological Conference, University of Oulu, 18-23 August 2004. (Studia Humaniora Ouluensia 1). University of Oulu, Oulu, 120 130. Lavento, M., Halinen, P., Timofeev, V., Gerasimov, D. & Saksa, A. 2001. An archaeological eld survey of Stone Age and Early Metal Period settlement at Kaukola (Sevastyanovo) and Räisälä (Melnikovo) on the Karelian Isthmus in 1999. Fennoscandia Archaeologica XVIII, 3 25. Lavento, M., Huttunen, R.-L. & Huttunen, A. 2004. Palynological investigations at the Ruhtinansalmi dwelling-site complex in Suomussalmi. Anthropogenic pollen evidence for a hunter-gatherer economy? In M. Lavento (ed.), Early in the North, Vol. 5. (Iskos 13). Suomen muinaismuistoyhdistys, Helsinki, 91 101. Lehtinen, L. & Sepänmaa, T. 1995. Surveys of archaeological sites in southern Savo. Fennoscandia archaeologica XII, 77 82. Leskinen, S. 2002. Late Neolithic house at Rusavierto. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 147 169. Leskinen, S. 2003. On the dating and function of the Comb Ceramics from Maarinkunnas. Finskt museum 102 (1995), 5 43. Liedgren, L. 1998. Förhistoriska byggnadskonstruktioner i Norrland. In O. Kyhlberg (red.), Hus och tomt i Norden under förhistorisk tid. (Bebyggelsehistorisk tidskrift Nr. 33). Stockholms universitet, Stockholm, 155 168. Lindqvist, A.-K. 2007. I vilken vid berget. In P. Gustafsson & L.G. Spång (eds.), Stenålderns stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/Härnösand, 116 145. Lindholm, P., Olsson, E. & Runeson, H. 2007. Grop, snörre och kam. In P. Gustafsson & L.G. Spång (eds.), Stenålderns stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/ Härnösand, 201 235. Lindholm, P. & Runeson, H. 2007. Lillmosjö en dagresa bort?. In P. Gustafsson & L.G. Spång (eds.), Stenålderns stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/ Härnösand, 147 159. Loefer, D. & Westfal, U. 1985. A well-preserved Stone Age dwelling site. Preliminary presentation of the investigations at Vuollerim, Lapland, Sweden. In M. Backe et al. (eds.), In honorem Evert Baudou (Archaeolgy and Environment 4). University of Umeå, Umeå, 425 434. Luho, V. 1948. Suomen kivikauden pääpiirteet. Otava, Helsinki. Luho, V., Hyyppä, E. & Gustafsson, Ch. 1956. En i Helsingfors funnen stenåldernskanot. 78

Finskt Museum LXIII (1956), 17 29. Lundberg, Å. 1985. Villages in the inland of Northern Sweden 5000 years ago. In M. Backe et al. (eds.), In honorem Evert Baudou (Archaeolgy and Environment 4). University of Umeå, Umeå, 293 301. Lundber, Å. 1997. Vinterbyar. Ett bandsamhällets territorier i Norrlands inland, 4500 3500 f. Kr. (Studia Archaeologica Universitatis Umensis 8). Umeå Universitet, Umeå. Lundberg, Å. 2007. Hus, hem och hamn. In P. Gustafsson & L.G. Spång (eds.), Stenålderns stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/Härnösand, 187 199. Mannermaa, K. 2008. The Archaeology of wings. Birds and people in the Baltic Sea region during the Stone Age. Kristiina Mannermaa, Helsinki. Matiskainen, H. 1989. Studies on the Cronology, Material Culture and Subsistence Economy of the Finnish Mesolithic, 10 000-6000 b.p.(iskos 8). Suomen muinaismuistoyhdistys, Helsinki. Matiskainen, H. & Jussila, T. 1984. Naarajärven kampakeraaminen asumus. Suomen Museo 1984. 17 52. Marshall, Y. 2006. Introduction: adopting a sedentary lifeway. In Y. Marshall (ed.), Sedentism in Nonagricultural Societies. (World Archaeology 38, Issue 2). Routledge, Oxon, 153 63 Meinander, C. F. 1940. Pyheensilta stenåldersboplats. Finskt Museum 1939, 28 43. Meinander, C. F. 1954a. Die Kiukaiskultur. (Suomen Muinaismuistoyhdistyksen Aikakauskirja 53). Suomen muinaismuistoyhdistys, Helsinki. Meinander, C. F. 1954b. Die Bronzezeit in Finnland. (Suomen Muinaismuistoyhdistyksen Aikakauskirja 54). Suomen muinaismuistoyhdistys, Helsinki. Meinander, C. F. 1976. Hyddbottnar av Madeneva typ. In T. Edgren (red.), Nordiska Arkeologimötet i Helsingfors 1967. En berättelse över mötet och dess förhandlingar sammanställd av Torsten Edgren. (Iskos 1). Suomen muinaismuistoyhdistys, Helsinki, 26 29. Miettinen, A. 2002. Relative Sea Level Changes in the Eastern Part of the Gulf of Finland During the Last 8000 Years. (Annales Academiae Scientiarum Fennicae, Geologica Geographica, 162.) Suomalainen tiedeakatemia, Helsinki. Miettinen, M. 1983. Stenåldersboplatsen Hundbacka I Pedersöre, Österbotten. Bottnisk kontakt 1. Föredrag vid maritimhistorisk konferens i Örnsköldsvik 12 14 februari 1982. (Skrifter från Örnsköldsviks museum nr 1), 14 21. Miettinen, M. 1983. Esihistoriallinen aika Etelä-Pohjanmaan Järviseudulla. In H. Junnila & H. Rantatupa (eds.), Järviseudun historia I. Esihistoriasta 1850-luvulle. Järviseutuseura, Evijärvi, 61 174. Miettinen, M. 1998. Asbestikeramiikkalöytöjä Etelä-Pohjanmaalta. In H. Ranta (ed.), Kentältä poimittua 4. Kirjoitelmia arkeologian alalta. (Museoviraston arkeologian osaston julkaisuja7). Finnish National Board of Antiquities, Helsinki, 60 71. Miettinen, T. 1998. Kymenlaakson esihistoriaa. (Kymenlaakson maakuntamuseon julkaisuja, 26.) Kymenlaakson maakuntamuseo, Kotka. Miettinen, T. 2004. Jaalan esihistoriaa. In E. Salminen, R. Hamari & T. Miettinen (eds.), Jaalan historia. Vesikansan vaiheet muinaisista ajoista 2000-luvulle. Jaalan kunta, Jaala, 89 136. 79

Moisanen, J. 1991. Tutkimuksia Kerimäen kivikautisilla asuinpaikoilla. Sihti 1, 25 32. Mökkönen, T. 2001. Hauki on kala: Saimaan vesistöalueen kivi- ja varhaismetallikauden osteologinen aineisto. Muinaistutkija 3/2001, 2 13. Mökkönen, T. 2007. Helsingin Aleksis Kiven kadun ruuhi. Muinaistutkija 4/2007, 33 37. Mökkönen, T. 2009. A Response to Samuel Vaneeckhout s social interpretation of the Kuuselankangas housepit site. Muinaistutkija 3/2009, 70 74. Mökkönen, T. 2010. A Response to Samuel Vaneeckhout s social interpretation of the Kuuselankangas housepit site: Part 2. Muinaistutkija 1/2010, 77 82. Mökkönen, T., Lavento, M. & Halinen, P. 2006. Neolithic Sites in Various Environments of the Karelian Isthmus. Kaukola Räisälä project s survey 2004. In Kirpichnikov, A. N., Nosov, E. N. & Saksa, A. I. (eds), Slavs, Finns and Ougriens. The Zones of Contacts and Cooperation. The papers of Russian-Finnish symposium on the Problems of Archaeology and History, Puškinskije Gory 2004, October, 7-10. Institute for the History of Material Culture Russian Academy of Sciences, Saint-Petersburg, 112 129. Mökkönen, T. & Nordqvist, K. 2006. Opettaja Sokolovin kivikauden kokoelma Karjalankannakselta. Muinaistutkija 2/2006, 7 18. Mökkönen, T. & Seitsonen, O. 2007. Virolahden pyyntikulttuurien inventointi vuonna 2005. Muinaistutkija 2/2007, 20 39. Murdock, G. P. 1967. Ethnographic atlas. University of Pittsburgh Press, Pittsburgh. Norberg, E. 2008. Boplatsvallen som bostad I Norrbottens kustland 5000 till 2000 före vår tideräckning. En studie av kontinuitet och förändring. (Studia Archaeologica Universitatis Umensis 23). Umeå Univeristet, Umeå. Nordqvist, K. 2007. Päättyvä kivikausi (2300 1800 eaa.). In M. Lavento & A. Lahelma (eds.), Sama maisema, eri kulkijat Repoveden kansallispuisto kivikaudelta 1900-luvulle. (Metsähallituksen luonnonsuojelujulkaisuja. Sarja A 165). Metsähallitus, Helsinki, 43 46. Nordqvist, K. & Lavento, M. 2008. Archaeological survey in Kaukola and Räisälä in 1999 and a study of environmental settings of the Stone Age dwelling sites in the area. In K. Nordqvist (ed.), Karelisn Isthmus Stone Age studies in 1998 2003. (Iskos 16). The Finnish Antiquarian Society, Helsinki, 140 163. Nordqvist, K. & Seitsonen, O. 2008. Archaeological research in the former municipalities of Koivisto and Kuolemajärvi, Karelian Isthmus in 2003: results and observations. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003. (Iskos 16). The Finnish Antiquarian Society, Helsinki, 215 234. Nordqvist, K., Seitsonen, O. & Lavento, M. 2009. Waterways and the Stone Age and Early Metal Period studies on the Karelian Isthmus The pre-world War II studies and research carried out by the University of Helsinki in 1998 2006. Quaternary International 203 (2009), 25 32. Nordqvist, K., Seitsonen, O. & Uino, P. 2008. Appendix 1. Stone Age and Early Metal Period sites in the studied municipalities. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003. (Iskos 16). The Finnish Antiquarian Society, Helsinki, 291 328. Núñez, M. 2004. All quiet on the eastern front? In H. Knutsson (ed.), Coast to Coast: Arrival. Results and Reections. Uppsala University, Uppsala, 345 367. 80

Núñez, M. 2009a. The Sea Giveth, The Sea Taketh: The Role of Marine Resources in Northern Ostrobotnia, Finland, 4000-2000 B.C. Arctic Anthropology, Vol. 46, Nos. 1-2, 167 175. Núñez, M. 2009b. What did jätinkirkko builders mean to see through their wall openings?. In J. Ikäheimo & S. Lipponen (eds.), Ei kiveäkään kääntämättä. Juhlakirja Pentti Koivuselle. Pentti Koivusen juhlatoimikunta, Oulu. 117 127. Núñez, M. & Okkonen, J. 1999. Environmental background for the rise and fall of villages and megastructures in north Ostrobothnia 4000 2000 cal BC. In M. Huurre (ed.), Dig It All. Papers Dedicated to Ari Siiriäinen. The Finnish Antiquarian Society & The Archaeological Society of Finland, Helsinki, 105 115. Núñez, M. & Okkonen, J. 2005. Humanizing of north Ostrobothnian landscape during the 4th and 3rd millennia BC. Journal of Nordic Archaeological Science, 15, 25 38. Nuñez, M. & Uino. P. 1997. Dwellings and related structures in prehistoric Finland. In O. Kyhlberg (red.), Hus och tomt i Norden under förhistorisk tid. (Bebyggelsehistorisk tidskrift Nr. 33). Stockholms universitet, Stockholm, 133 152. Nurminen, K. 2007. Kalanluulöytöjä Suomen neoliittisilta asuinpaikoilta. Muinaistutkija 1/2007. 2 17. Ojanlatva, E. & Alakärppä, J. 2002. Interpretation of the Peurasuo house pit in Oulu. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 109 122. Okkonen, J. 2001. Cairns and cultural landscape an attempt to dene Stone Age and Bronze Age land use and territoriality in Ostrobothnia, Finland. Faravid XXV, 23 35. Okkonen, J. 2003. Jättiläisen hautoja ja hirveitä kiviröykkiöit Pohjanmaan muinaisten kivirakennelmien arkeologiaa. (Acta Universitatis Ouluensis Series B Humaniora, 52). Oulun yliopisto, Oulu. Okkonen, J. 2009. Itämeri vuorovaikutusalueena kivikaudella. In K. Alenius, A. Honkala & S. Wunsch (eds.), Itämeren itälaidalla II, On the Eastern Edge of the Baltic Sea II. Rajat ja integraatio Fenno-Baltian historiassa.(studia Historica Septentrionalia 58). Pohjois-Suomen historiallinen yhdistys, Rovaniemi, 5 17. Okkonen, J. & Ikäheimo, J. 1993. Kannuksen Linnakankaan ja Himangan Hiidenlinnan muinaisjäännöksistä. Faravid 16, 7 26. Okkonen, J. & Ridderstad, M. 2009. Jätinkirkkojen aurinkosuuntauksia. In J. Ikäheimo & S. Lipponen (eds.), Ei kiveäkään kääntämättä. Juhlakirja Pentti Koivuselle. Pentti Koivusen juhlatoimikunta, Oulu. 129 135. Olsen, B. 1994. Bosetning og samfunn i Finnmarks forhistorie. Universitetsforlaget, Oslo. Olsen, B. 1998. Forhistoriske hus i Nord-Norge. In O. Kyhlberg (red.), Hus och tomt i Norden under förhistorisk tid. (Bebyggelsehistorisk tidskrift Nr. 33). Stockholms universitet, Stockholm, 185 194. Olson, G., Runeson, H., Sigvallius, B. & Storå, J. 2007. Människor och djur. In P. Gustafsson & L.G. Spång (eds.), Stenålderns stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/ Härnösand, 51 69. Pälsi, S. 1915. Riukjärven ja Piiskunsalmen kivikautiset asuinpaikat Kaukolassa. (Suomen Museo XXVIII), Suomen muinaismuistoyhdistys, Helsinki. 81

Pälsi, S. 1918. Kaivaus Pitkäjärven kivikautisella asuinpaikalla Räisälässä v. 1915. Suomen Museo XXV (1918), 25 34. Pesonen, P. 1996a. Early Asbestos Ware. In T. Kirkinen (ed.), Pithouses and Potmakers in Eastern Finland. Reports of the Ancient Lake Saimaa Project. (Helsinki Papers in Archaeology n:o 9). University of Helsinki, Helsinki, 9 40. Pesonen, P. 1996b. Posion Kuorikkikankaan asumus. Muinaistutkija 1/1996, 19 25. Pesonen, P. 1999a. Rekikylä kivikautinen kylä Ylikiimingissä. Muinaistutkija 1/1999, 2 15. Pesonen, P. 2001. Kiteen Sarvisuo - lisää varhaisesta asbestikeramiikasta. In H. Ranta (ed.), Kentältä poimittua 5, Kirjoitelmia arkeologian alalta. (Museoviraston arkeologian osaston julkaisuja 9). Museovirasto, Helsinki, 34 56. Pesonen, P. 2002. Semisubterranean houses in Finland a review. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 9 41. Pesonen, P. 2004. Neolithic pots and ceramics chronology AMS-datings of Middle and Late Neolithic ceramics in Finland. In P. Uino (ed.), Fenno-Ugri et Slavi 2002. Dating and Chronology. (Museoviraston arkeologian osaston julkaisuja10). Museovirasto, Helsinki, 87 97. Pesonen, P. 2006. One house two households? An investigation of a Late Subneolithic pithouse in Kuorikkikangas site, Posio, southern Lapland. In V.-P. Herva (ed.), People, Material Culture and Environment in the North. (Studia Humaniora Ouluensia 1.) University of Oulu, 198 213. Pesonen, P. & Leskinen, S. 2009. Pottery of The Stone Age Hunter-Gatherers in Finland. In P. Jordan & M. Zvelebil (eds.), Ceramics before farming. The dispercial of pottery among prehistoric Eurasian hunter-gatherers. Left Coast Press, Inc., Walnut Creek, 299 318. Rafferty, J.E. 1985. The archaeological record on sedentariness: recognition, development, and implications. In M.B. Schiffer (ed.), Advances in Archaeological Method and Theory 8. Academic Press, New York, 113 56. Rankama, T. 2002. Analyses of the quarz assemblages of houses 34 and 35 at Kauvonkangas in Tervola. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 79 108. Ranta, H. (ed.) 2002. Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki. Renouf, M. A. P. 1984. Northern coastal hunter-shers: an archaeological model. Coastal Archaeology. (World Archaeology Vol. 16, No. 1). Routledge, London & New York, 18 27. Renouf, M. A. P. 1986. Excavations at a Younger Stone Age settlement in Varangerfjord, Norway. Polar Record, Volume 23, Issue 144, 273 288. Renouf, M. A. P. 1988. Sedentary coastal hunter-sherers: an example from the Younger Stone Age of northern Norway. In G. Bailey & J. Parkington (eds.), The archaeology of prehistoric coastlines. (New directions in archaeology). Cambridge University Press, Cambridge. 102 115. Runeson, H. 2007. Den goda ordningen. In P. Gustafsson & L.G. Spång (eds.), Stenålderns stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/Härnösand, 73 115. 82

Saarnisto, M. 1970. The Late Weichselian and Flandrian History of the Saimaa Lake Complex. (Commentationes Physico-Mathematicae, Vol. 37). Societas Scientarium Fennica, Helsinki. Saarnisto, M. 2003. Karjalan arkeologisen aineiston radiohiiliajoituksia. In M. Saarnisto (ed.), Viipirin läänin historia 1. Karjalan synty. Karjalan kirjapaino, Lappeenranta, 512. Saarnisto, M. 2008. Emergence history of the Karelian Isthmus. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003. (Iskos 16). The Finnish Antiquarian Society, Helsinki, 128 139. Saarnisto, M. & Siiriäinen, A. 1970. Laatokan transgressioraja (Referat: Die Transgressionsgrenze des Ladogasees). Suomen Museo 77, 10 22. Sajantila, A., Salem, A.-H., Savolainen, P., Bauer, K., Gierig, C. & Pääbo, S. 1996. Paternal and maternal DNA lineages reveal a bottleneck in the founding of the Finnish population. Proceedings of the National Academy of Sciences of the United States of America 93, 12035 12039. Salo, U. 1989. Astian kulttuurihistoriasta Suomessa ja naapurialueilla. Suomen Museo 1989, 5 45. Sanders. D. 1990. Behavioral conventions and archaeology: Methods for the analysis of ancient architecture. In S. Kent (ed.), Domestic architecture and the use of space. An interdisciplinary cross-cultural study. Cambridge University Press, Cambridge, 43 70. Sartes, M. 1994. Subneolithic and Neolithic settlement systems in Southwest Finland. Discussion on resource areas. In P. Purhonen (ed.), Fenno-Ugri et Slavi 1992. Prehistoric economy and means of livelihood. (Museoviraston arkeologian osaston julkaisuja 5), 105 114. Saukkonen, J. 1994. Annual cycles, base camps and means of livelihood Stone-Age hunter-gatherer settlement in the Närvijoki river area of Southern Ostrobothnia. In P. Purhonen (ed.), Fenno-Ugri et Slavi 1992. Prehistoric economy and means of livelihood. (Museoviraston arkeologian osaston julkaisuja 5), 115 123. Schanche, K. 1995. Var det store landsbyboplasser i Finnmark i yngre steinalder? Debattinnlegg med utgangspunkt i bosetningsspor fra ca. 2000 f.kr. Arkeologiske skrifter 8. Arkeologisk Institutt & Bergen Museum, Bergen, 174 186. Schultz, H.-P. 2009. Honkobackharju ja Keski-Pohjanmaan jätinkirkot. In J. Ikäheimo & S. Lipponen (eds.), Ei kiveäkään kääntämättä. Juhlakirja Pentti Koivuselle. Pentti Koivusen juhlatoimikunta, Oulu. 137 149. Seitsonen, O. 2005a. Mesoliittinen ja neoliittinen kiviteknologia Räisälä Juoksemajärvi Westend asuinpaikalla. Muinaistutkija 4/2005, 15 31. Seitsonen, O. 2005b. Shoreline displacement chronology of rock paintings at Lake Saimaa, eastern Finland. Before farming 1/2005, article 4, 1 21. Seitsonen, O. 2006. Räisälä Pitkäjärvi revisited new interpretations of the dwelling remains. In P. Pesonen & T. Mökkönen (eds.), Arkeologipäivät 2005. Arkeologia ja kulttuuri & Uutta kivikauden tutkimuksessa. Suomen arkeologinen seura, Helsinki, 138 145. Seitsonen, O. & Gerasimov, D. V. 2008. Archaeological research in the Kurkijoki area in 2001 and 2003. a preliminary study of the Stone Age settlement in southern Ladoga Karelia. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003. 83

(Iskos 16). The Finnish Antiquarian Society, Helsinki, 164 184. Seitsonen, O., Nordqvist, K. & Gerasimov, D. V. 2009. Recent archaeological research in the Lake Pyhäjärvi microregion, Karelian Isthmus, Russia: The multi-period site of Pyhäjärvi Kunnianniemi and some Early Combed Ware period nds. Fennoscandia archaeologica XXVI, 163 171. Seitsonen, S. 2008. Osteological material from Stone Aeg and Early Metal Period sites in Karelian Isthmus and Ladoga Karelia. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003. (Iskos 16). The Finnish Antiquarian Society, Helsinki, 266 283. Siiriäinen, A. 1969. Über die Chronologie der steinzeitlichen Küstenwohnplätze Finnlands im Lichte der Uferverschiebung. Suomen Museo 76 (1969), 40 73. Siiriäinen, A. 1970. Archaeological Background of Ancient Lake Päijänne and geological Dating of the Meso-Neolithic Boundary in Finland. Bulletin of the Geological Society in Finland 42, 119 127. Siiriäinen, A. 1971. Shoreline Dating of the Särärisniemi 1 -Ceramics in Finland. Suomen Museo 78 (1971), 9 19. Siiriäinen. A. 1972. A Gradient/time curve for dating Stone Age shorelines in Finland. Suomen Museo 79 (1972), 5 18. Siiriäinen, A., 1974. Studies Relating to Shore Displacement and Stone Age Chronology in Finland. Finskt Museum 80, 5 22. Siiriäinen, A. 1978. Archaeological Shore Displacement Chronology in Northern Ostrobothnia, Finland. Iskos 2, 5 23. Siiriäinen, A. 1981. On the cultural ecology of Finnish Stoen Age. Suomen Museo 1980, 5 40. Siiriäinen, A. 1982. Recent Studies on the Stone Age Economy in Finland. Fennoscandia antiqua I [the rst number of current Fennoscandia Archaeologica], 17 26. Siiriäinen, A. 1987. On Archaeology and Land Uplift in Finland. In M. Perttunen (ed.), Fennoscandian land uplift: proceedings of a symposium at Tvärminne, April 10-11, 1986. (Geological Survey of Finland. Special Paper 2). Geological Survey of Finland, Espoo, 43 45. Siiriäinen, A. 1996. Foreword. In T.Kirkinen (ed.), Environmental Studies in Eastern Finland. Report of the Ancient Lake Saimaa Project. (Helsinki Papers in Archaeology, No. 8). University of Helsinki, Helsinki, 5 7. Sipilä, J. & Lahelma, A. 2006. War as a Paradigmatic Phenomenon: Endemic Violence and the Finnish Subneolithic. Journal of Conict Archaeology 2, 189-209 Simonsen, P. 1976. Steinalderens hustyper I Nord-Norge. In T. Edgren (red.), Nordiska Arkeologimötet i Helsingfors 1967. En berättelse över mötet och dess förhandlingar sammanställd av Torsten Edgren. (Iskos 1). Suomen muinaismuistoyhdistys, Helsinki, 23 25. Simonsen, P. 1979. Veidemenn på Nordkalotten. Hefte 3, Yngre steinalder og overgang til metall tid. (Stensilserie/Institutt for samfunnsvitenskap B, Historie 17). Universitet i Tromsø, Tromsø. Skandfer, M. 2009. Ethics in the Landscape: Prehistoric Archaeology and Local Sámi Knowledge in Interior Finnmark, Northern Norway. Arctic Anthropology 46 (1-2), 89 102. Spång, L. G. 2007. Säljägare torrläggs. In P. Gustafsson & L.G. Spång (eds.), Stenålderns 84

stationer. Arkeologi i Botniabanans spår. Riksantivarieämbetet & Murberget Länsmuseet Västernorrland. Stockholm/Härnösand, 33 49. Tallavaara, M., Pesonen, P. & Oinonen, M. 2010. Prehistoric population history in eastern Fennoscandia. Journal of Archaeological Science 37, 251 260. Thomas, J. 2001. Archaeologies of place and landscape. In I. Hodder (ed.), Archaeological Theory Today. Polity Press, Cambridge, 165 186. Tilley, C. 1994. A Phenomenology of Landscape. Places, paths and monuments. Berg Publishers, Oxford. Timofeev, V. I., Zaitseva, G. I., Lavento, M., Dolukhanov, P. & Halinen, P. 2004. he radiocarbon datings of the Stone Age Early Metal Period on the Karelian Isthmus. Geochronometria Vol. 23, 93 99. Torvinen, M. 2000. Säräisniemi 1 ware. Fennoscandia Archaeologica XVII, 3 35. Trigger, B. G. 1978. Handbook of North American Indians. Volume 15, Northeast. Smithsonian Institution, Washington. Ukkonen, P. 1993.The Post-glacial history of the Finnish mammalian fauna. Annales Zoologici Fennici 30, 249 264. Ukkonen, P. 1996. Osteological analysis of the refuse fauna in the Lake Saimaa area. In T. Kirkinen (ed.), Environmental studies in Eastern Finland. (Helsinki Papers in Archaeology No.8). University of Helsinki, Helsinki, 63 91. Uino, P. 1997. Ancient Karelia. Archaeological Studies Muinais-Karjala. Arkeologisia tutkimuksia. (Suomen Muinaismuistoyhdistyksen Aikakauskirja 104). Suomen muinaismuistoyhdistys, Helsinki. Uino, P. 2003. Karjalan arkeologiaa 150 vuotta. In M. Saarnisto (ed.), Viipurin läänin historia 1. Karjalan synty. Karjalan kirjapaino, Lappeenranta, 117 150. Vaara, R. 2000. Yli-Iin Kierikin kivikautinen kylä asumuskonstruktiot 1998 1999. Muinaistutkija 2/2000, 2 12 Vaneeckhout, S. 2008a. Dwelling depressions at Kierikki: results from GPS-survey. Faravid 32, 7 17. Vaneeckhout, S. 2008b. Sedentism on the Finnish northwest coast: shoreline reduction and reduced mobility. Fennoscandia Archaeologica XXV (2008), 61 72. Vaneeckhout, S. 2009a. 2500 years of social evolution in a northern Finnish Stone Age village. In T. Mökkönen & S.-L. Seppälä (eds.), Arkeologipäivät 2008. Sosiaaliarkeologiaa yhteisöt arkeologisen aineiston taustalla & Tutkimushistorian painolasti. Suomen arkeologinen seura, Helsinki, 63 71. Vaneeckhout, S. 2009b. Where is the social in social archaeology? Muinaistutkija 2/2009, 40 44. Vaneeckhout, S. 2009c. Aggregation and polarization in northwest coastal Finland. Socio-ecological evolution between 6500 and 4000 cal BP. Faculty of Humanities, Archaeology, University of Oulu, Oulu. Vaneeckhout, S. 2010. House Societies among Coastal Hunter-Gatherers: A Case Study of Stone Age Ostrobothnia, Finland. Norwegian Archaeological Review, Vol. 43, No. 1, 12 25. Vikkula, A. 1988. Östersjön som en förutsättning för mellan- och senneolitisk utveckling. Iskos 7, 59 66. Vikkula, A. 1994a. Ecological Approaches to the Stone Age of Ancient Lake Saimaa. In P. Purhonen (ed.), Fenno-Ugri et Slavi 1992. Prehistoric economy and the means of 85

livelihood. (Museovirasto, Arkeologian osaston julkaisuja 5). Finnish National Board of Antiquities, Helsinki, 167 179. Vikkula, A. 1994b. Stone Age environment and landscape changes on the eastern Finnish lake district. In I. Johnson (ed.), Methods in the Mountains, Proceedings of UISPP Commission IV Meeting, Mount Victoria, Australia, 9th 13th August 1993. (Sydney University Archaeological Methods Series # 2). University of Sidney, Sidney, 91 97. Vikkula, A. 1996. Foreword. In T. Kirkinen (ed.), Pithouses and Potmakers in Eastern Finland. Reports of the Ancient Lake Saimaa Project. (Helsinki Papers in Archaeology, No. 9). University of Helsinki, Helsinki, 5 7. Wodarczak, P. 2009. Radiocarbon and dendrogronological dates of the Corded Ware culture. Radiocarbon, Vol. 51, Nr 2, 737 749. Wilhelms, S. 1995. A natural-geographical study of prehistoric sites. Fennoscandia Archaeologica XII, 195 205. Zhul nikov, A. M. 1999. Eneolit Karelii (pamyatniki s poristoi i asbestovoi keramikoi) (with summary: Aeneolithic Age of Karelia [the sites with organic and asbestostempered ceramics]). Karelskii nauchnyi tsentr RAN, Petrozavodsk. Zhul nikov, A. M. 2003. Drevnie zhilishcha Karelii. Karel skii gosudarstvennyi kraevedcheskii muzei, Petrozavodsk. Zvelebil, M. 1981. From forager to farmer in the boreal zone. Reconstructing economic patterns through catchment analysis in prehistoric Finland. (BAR international series 115 (iii)). British Archaeological Reports, Oxford. 86

Paper I Mökkönen, T, 2002. Chronological variation in the locations of hunter-gatherer occupation sites vis-à -vis the environment. In H. Ranta (ed.). Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 53 64. [references p. 240 254]. Publisher: National Board of Antiquities, Helsinki. ISBN 951-616-079-4

Teemu Mökkönen CHRONOLOGICAL VARIATION IN THE LOCATIONS OF HUNTER-GATHERER OCCUPATION SITES VIS-À-VIS THE ENVIRONMENT 1. Introduction The location of occupation sites vis-à-vis the contemporaneous environment has been discussed in several papers dealing with settlement patterns, exploitation areas and subsistence. The location and the palaeoenvironment of occupation sites in Finland have often been investigated one site at a time only. More extensive surveys comprising several sites had not been conducted in Finland until the early 1980 s, when Zvelibil (1981) published his thesis on site catchment analysis of the Stone and Bronze Age settlement by the Kokemäenjoki River. At the same time, Siiriäinen published a paper in which he divides the coastal occupation sites into four ecological zones, reflecting the seasonal variation and the maximum possible exploitation of the environment in a small and restricted area (Siiriäinen 1981a; 1987). Broadbent (1979:192) and Welinder (1983:100) have suggested similar settlement patterns for neighbouring areas. After the early 1980s a number of studies of the settlement patterns and subsistence of hunter-gatherer cultures have been conducted in Finland, employing site catchment analysis (Saukkonen 1990 unpublished; Saukkonen 1994, Kylli 2000 unpublished; Kylli 2001; Mökkönen 2000 unpublished), nearest neighbour analysis (Sartes 1991 unpublished; Sartes 1994), or the arrangement of coastal settlements in the ecological zones proposed by Siiriäinen (Matiskainen 1989; Saukkonen 1990 unpublished). In Lapland the relations between settlement patterns, subsistence and a riverine environment have been investigated by Rankama (1996; 1997b). It was not until 1994 that computer resources were taken into use in environmental analyses, and the first archaeological studies using the GIS-method in Finland were conducted (Kirkinen 1994 unpublished; Kirkinen 1996; Vikkula 1994b). Previously, research on hunter-gatherer settlement patterns has focussed on coastal areas, where a number of strongly seasonal resources associated with the sea (e.g., salmon runs or seal hunting on the ice) have been central in creating the models. Often, a model of annual cycles incorporating travel between the coast and the inland regions, as dictated by the availability of natural resources, has been proposed. No previous research pertaining to exploitation areas or settlement patterns has been carried out in the inland lake areas, where the resources offered by the environment differ considerably from those of the coastal regions and the inland areas of Lapland. House pits have not been investigated before in this context, either. Studies analysing the locations of occupation sites have been rare in Finland. The only such analysis was undertaken in the area of the Lake Saimaa water system by the Ancient Lake Saimaa project of the University of Helsinki (Vikkula 1994a; Wilhelms 1995; Pesonen 1996 unpublished). The topographic 53

variables illustrating the Stone Age settlement of two research areas in the Lake Saimaa water system created by Vikkula (1994a) have been later employed by both Pesonen (1996 unpublished), Koivikko (1999 unpublished) and Mökkönen (2000 unpublished). The ritualistic landscape of hunter-gatherers has only been investigated in Finland in connection with rock paintings (Lahelma 2000 unpublished; Lahelma 2001). 2. The Research Methods The present research is based on four environmental reconstructions of different time horizons created with GIS, which were employed in a visual analysis of the distribution and topographic location of occupation sites. The aim of a visual analysis is to create a general view of the sites during the different time horizons. The resolution of this general view was enhanced by placing sites that differed from each other in their structures and their location in a distance diagram. In addition to the reconstructions of the areas of water and dry land in the different time horizons, the original investigation contained environmental reconstructions in which the vegetation and the water areas less than 3 metres deep were modelled. These environmental reconstructions were used in the exploitation area analysis (Site Catchment Analysis) of the sites, the results of which are not dealt with in this article (see Mökkönen 2000 unpublished). The aim was to use the visual settlement analysis based on the environmental reconstructions and the distance diagram to identify the changes in the settlement pattern that occurred in the region from the time of the Early Asbestos Ware to the Early Metal Age. 3. The Region and the Material The research area lies in eastern Finland, primarily in the municipalities of Kerimäki, Kesälahti and Savonranta, but including also parts of Punkaharju and Rääkkylä. The bulk of the area is in eastern Kerimäki. Waterways and topographic features were used to delineate a unified area some 30 x 40 km in extent. The backbone of the investigated region is an esker that runs through it towards the southeast. Areas of open water in the north and in the south clearly mark the borders. In the east an extensive area with flat contours borders the region, while in the west the landscape is heterogeneous and topographically varied, broken up by small lakes and marshes. Towards this direction, the border of the research area is not delineated with clear topographic features. The studied archaeological material includes the entire range of prehistoric remains classified as deriving from hunter-fisher-gatherer cultures discovered in the area, from the Early Asbestos Ware to the Early Metal Age. Agriculture was not unknown during the latter, but the most important means of subsistence were still hunting and fishing (Taavitsainen et al. 1998). Practically all of the archaeological material derives from the period after the Vuoksi River, the current outlet of the Lake Saimaa drainage, broke through the Salpausselkä end moraine due to transgression causing substantial lowering of the Lake Saimaa water level. The material, thus, dates from roughly between 4100 calbc and 300 AD. A basic survey of the prehistoric remains in the whole area has been done in the 1990 s, and in 1998 it was complemented by a one month survey concentrating on hunter-gatherer sites. The research material comprises a total of 105 sites related to hunting and fishing cultures, 80 of which are Stone Age or Early Metal Age occupation sites. Using the pottery and the elevation of the sites as reference points, the material was divided into four time horizons: 1) Early Asbestos Ware, 2) Typical Comb Ware, 3) Late Comb Ware and Kierikki/Pöljä Ware (the material does not contain pottery that could be defined as Jysmä Ware), 4) Early Metal Age. The analysis of the settlement pattern proceeds in chronological order. Many of the sites were occupied over more than one time horizon. In addition, the equivocal upper and lower limits of a number of sites that have not yielded pottery makes it possible to assign them to several time horizons. Two pre-existing sets of numerical environmental data were employed in creating the environmental reconstructions: the 25 m 2 grid elevation model and the land use and standing tree crop interpretation database. The shoreline inclination diagrams published by Jussila (1999) were used to tilt the elevation model to correspond to the situations during the different time horizons. This resulted in a reconstruction of the areas of water and dry land during the various time horizons. The mean values of the elevations at which the various pottery types are commonly found (according to Jussila 1999:120) were employed in the reconstruction. The land use and standing tree crop interpretation database was used to model the vegetation. 54

4. Changes in Settlement Pattern from Early Asbestos Ware to Early Metal Age Time Horizon 1: Early Asbestos Ware Early Asbestos Ware sites are spread relatively evenly and without clear clustering in the region. The occupation sites lie on narrow peninsulas, isthmuses, islands, and next to narrow waterway passages (Fig. 1). The distribution pattern agrees with the results of earlier surveys as regards to the degree of open water and aquatic conditions (see Pesonen 1996 unpublished). On the basis of their elevation and the presence of possible Early Asbestos Ware three house pits have been interpreted as part of the oldest time horizon, despite that fact that no extensively excavated hut depressions unequivocally associated with Early Asbestos Ware have been previously encountered (Pesonen 1996b:28). The house pits lie next to large bodies of open water, near the mouths of minor bays. Time Horizon 2: Typical Comb Ware Several changes occur in the settlement pattern during the Typical Comb Ware period. The number of house pits increases and their locations focus on two areas, both of which are at the crossroads of good routes of transportation (Fig. 2). House pits cluster around areas where eskers and narrow straits connecting two waterways meet. There is no clear common denominator in the locations of dwelling sites that do not include house pits. They have been found at the heads of peninsulas, on islands, and by straits the same way as during the Early Asbestos Ware period but also at the heads of bays. In general, the focus of settlement transferred during the Typical Comb Ware period to an environment with less open water. In order to gain a more detailed view of the development of the settlement pattern during the Typical Comb Ware period, the sites were placed in a distance diagram (Fig. 3). Judging from the diagram, it appears likely that the pattern of settlement changed during this period. The oldest hut depressions lay in the sheltered heads of southward facing bays, protected from the wind by an esker behind the site. In the latter part of the Typical Comb Ware period, the focus shifted to peninsulas and islands. At the same time, the number of house pits per site grew, and the sites were not always oriented southwards. In spite of the changes in the locations of the house pits, the same regions continued to be inhabited. The largest sites contain more than ten house pits, but isolated house pits have also been encountered. While we do not know how many of the pit houses were inhabited simultaneously at each site, the settlement pattern was probably village-like the same way as in the more extensive Ostrobothnian Stone Age communities (e.g., Nuñez & Okkonen 1999; Pesonen 1999b). In any case, the house pits on peninsulas and islands are, in their vulnerability to the elements, in an entirely different environment than before. During the Typical Comb Ware period the chronological relationship between the clusters of house pits and the sites without house pits cannot be unequivocally determined. Sites with and without house pits lie mostly at similar elevations. It is possible, however, that these sites are not contemporaneous. The seasonal variation of the water level in the Saimaa lake system probably differs today from what it was during the Typical Comb Ware period. The flooding during the fist half of the year was more intense than today, while during the late summer and early autumn the water level was considerably lower (see Mökkönen 2000). One may assume that the house pits were constructed beyond the reach of floodwaters, while sites that were in use for a short period only were probably located closer to the shore, and, in the autumn, at lower elevations. The number of sites without hut depressions probably begins to decrease already during the late Typical Comb Ware period, which might be considered an indication of increasing sedentariness (potentially denoting a change from seasonal sedentism to permanent sedentism?). Time Horizon 3: The Kierikki and Pöljä Ceramics and Late Comb Ware At the beginning of the Late Stone Age time horizon, house pits occur at the same sites as the village-type settlements characteristic for the late Typical Comb Ware period. After this period, house pits are found only individually on islands and peninsulas (Figs. 4 and 5). The period of Stone Age village settlement appears to have lasted only a short while in the region, from the late Typical Comb Ware period to the early phase of Time Horizon 3. Village-like settlement ends in the region after the time of the shore level dated at around 3250 BC (see Fig. 5) probably practically simultaneously with the introduction of Pöljä Ceramics (roughly 3100 calbc according to Carpelan 1999). During Time Horizon 3 the number of sites without house pits decreases considerably 55

56 Fig. 1. Occupation sites during Time Horizon 1 (Early Asbestos Ware). The water level corresponds to the upper level of the Ancient Lake Saimaa (c. 4100 calbc). (Based on Elevation Model 25 The National Land Survey of Finland, permission no. 205/MYY/02).

Fig. 2. Occupation sites during Time Horizon 2 (Typical Comb Ware). The water level has been lifted to correspond to the situation at c. 3250 calbc (Jussila s shoreline no. 2). (Based on Elevation Model 25 The National Land Survey of Finland, permission no. 205/MYY/02). 57

Fig. 3. Distance diagram of the occupation sites during Time Horizon 2 (Typical Comb Ware) and the ancient shorelines of Lake Saimaa. Shorelines according to Jussila (1999). For additional data on the sites, see Mökkönen 2000 unpublished: Appendix X:4. from what it was during the Typical Comb Ware period. The drop in the number of sites presumably began already during the late Typical Comb Ware period. Time Horizon 4: Early Metal Period There are no clearly discernible common features in the locations of sites during the Early Metal Age. Some of the sites lie on peninsulas, islands and by straits, i.e., in similar locations as Stone Age settlements (Fig. 6). The association of sites with the shoreline is no longer universal the locations of some of them differ considerably from the Stone Age pattern. Three house pits are interpreted as part of the Early Metal Age time horizon. One (Kesälahti 46 Kulhankangas B) lies clearly at a Stone Age elevation, but pottery finds from the site are probably Säräisniemi 2 Ware. Two other house pits lie on such low elevations that they most likely derive from the Early Metal Age (see Fig. 5). 5. Changes in the shape of the house pits Distinct differences can be observed in the present size and shape of the various house pits. Their shape in the research area will be examined on the basis of the observations presented in the research reports. Even though the current and the original shape of the house pit do not automatically correspond with each other, the pits described as rectangular are larger and deeper than the round/oval ones. The descriptions of house pits in research reports are very variable, and their designation as round, oval or rectangular is highly subjective. In this paper house pits whose shape has not been recorded, but whose size is expressed as a measurement of diameter, are interpreted as round. Similarly, depressions whose exact shape has not been described, but whose width and length differ, are interpreted as oval. Depressions that are clearly labelled as rectangular, and depressions that are surrounded by an earth 58

Fig. 4. Occupation sites during Time Horizon 3 (Kierikki/Pöljä Ware, Late Comb Ware). Some of the sites include Late Comb Ware, although it is not mentioned in the key. The water level has been lifted to correspond to the situation at c. 2750 calbc (Jussila s shoreline no. 4). (Based on Elevation Model 25 The National Land Survey of Finland, permission no. 205/MYY/02). 59

Fig. 5. Distance diagram of the house pits in the research area and the ancient shorelines of Lake Saimaa. Shorelines according to Jussila (1999). For additional data on the sites, see Mökkönen 2000 unpublished: Appendix X:2. wall, are included in the same category. The distance diagram (Fig. 7) based on the information about the depressions of different shapes shows that rectangular depressions and those surrounded by an embankment are a more recent phenomenon than round or oval depressions. The present-day shape of the of the house pits probably does not directly reflect the different building techniques used in constructing the walls and the roof (Vaara 2000:3), but is associated with the size of the pit. Rectangular house pits and those surrounded by an embankment are larger than round/oval depressions. The oldest rectangular house pits derive chronologically from the end of the Typical Comb Ware period and the beginning of the time horizon of the Kierikki/Pöljä Ware. Rectangular buildings were constructed throughout the Late Stone Age, but probably also during the Early Metal Age, as indicated by the rectangular house pits at Kulhangas B in Kesälahti (c. 13 x 17 m 2 ), which is surrounded by an embankment and dated, on the basis of the pottery finds, to that period. In addition to larger rectangular house pits, smaller depressions whose present day shape is circular or oval occur throughout the period. 6. Discussion The various sections of the research are accompanied by a number of uncertainties. One of the most important of these is the dating of the sites. Almost all of the material from the sites derives from field surveys, and the small amounts of material may have resulted in incorrect identifications of pottery types. Another problem is the dating of sites without pottery finds. If pottery or other datable finds are absent from a site, its elevation has been used in dating. The elevations of many sites are estimates made by archaeologists in the field on the basis of topographic maps, although 60

Fig. 6. Occupation sites during Time Horizon 4 (Early Metal Age). The water level has been lifted to correspond to the situation at c. 2100 calbc (Jussila s shoreline no. 5). (Based on Elevetion Model 25 The National Land Survey of Finland, permission no. 205/MYY/02). 61

Fig. 7. Distance diagram of the present-day shape of the house pits in the study area. Shorelines according to Jussila (1999). exact levelling of the shoreline has been carried out on a small number of sites. To further complicate the dating, land uplift has been so slow in the Lake Saimaa region that a site located at a particular elevation may be datable to more than one of the time horizons used here (for more information about shore displacement and the dating of pottery in the Lake Saimaa region, see Jussila 1999; Mökkönen 2000; Pesonen 1996b; 2001). Another problem is the fact that the time horizons used are defined on the basis of ceramic types, the dates of which partly overlap. In the Saimaa region, this is true of Early Asbestos Ware and Typical Comb Ware (Pesonen 1996b:28; 1997; 2001:48 cf. Carpelan 1999:273). Similarly, Typical Comb Ware, which defines Time Horizon 2, partially overlaps with the ceramics of Time Horizon 3, Kierikki and Pöljä Ware and Late Comb Ware (see Räihälä 1996:115 116; Pesonen 1997; Halinen et al. in press). Contradictory ideas about the origins of the Late Neolithic Pöljä- Jysmä Ware and the Early Bronze Age Textile Ceramics have been put forth in recent years. Carpelan (1999:270) suggests that late Neolithic Asbestos Wares was produced contemporaneously with Textile Ceramics at least until the period of the Säräisniemi 2 Ware. Lavento, however, argues that the contemporaneity clashes with the evidence provided by the material available today (Lavento 2001:183). The boundary between the Late Stone Age and the Early Metal Age may also be rather fluid. Given the small size, only 1200 km 2, of the investigated region, the results of the survey can not be generalised to cover the entire Lake Saimaa water system. A large number of hunter-gatherer sites are, however, known in the area. In addition, of the eleven sites with ten or more house pits in the Finnish lake district (Pesonen 1999b), two are within the present research area. The slow shore displacement of Lake Saimaa makes it difficult to judge from the available material whether or not the focus of settlement at some point outside our region. The material from our region does, however, match better the im- 62

Fig. 8. Settlement during the Typical Comb Ware period and the topography of the study area with the water level of c. 3250 calbc. The elevations have been indicated at 5 m intervals (white = 0 5 m above water, etc.). (Based on Elevetion Model 25 The National Land Survey of Finland, permission no. 205/MYY/02). 63

age of a central region than that of a periphery. In the context of this paper it is not possible to make a definite statement about the character of the village-like settlement that appears during the Typical Comb Ware period and disappears at the beginning of the period of the Kierikki/Pöljä Ware. In the research area there are, however, more previously unknown features associated with the village-like settlements than just the change in the location of the house pits from sheltered bays to the more exposed sites at the heads of peninsulas. Villages located on peninsulas are concentrated in specific areas, where eskers and straits connecting two water systems meet, and surrounded by extensive areas of low contours (Fig. 8). Next to the villages, areas of shallow water (less than 3 m in depth) are common (Mökkönen 2000 unpublished:138, 158). The largest villages in the studied region have been inhabited for long periods of time. The settlement has not been continuous, however, but interrupted by breaks, after which old house pits have been reused (Halinen et al. in press.) If the changes observed in the settlement pattern are real, and not merely a reflection of the focus of the settlement migrating from time to time outside the investigated region, questions arise about the possible chains of events behind the process. Changes in the settlement pattern probably cannot be explained through changes in hunting or fishing. The osteological record of the Saimaa lake district does not indicate great upheavals fishing was the basis of subsistence throughout the period. The gradual sift of the focus of hunting from aquatic to terrestrial mammals between the Early Asbestos Ware and the Early Metal Age is the only development of greater significance. This is indicated by the decrease of beaver and the increase of cervids in the osteological material (Mökkönen 2000 unpublished, Mökkönen 2001). To return finally to the village-like communities, the intensification of settlement requires an increase in the resource base or the more intensive use of already existing resources (Fletcher 1995:174 175). The Saimaa landscape, characterised by a rising shoreline during the transgression that prevailed before the Vuoksi River broke through the Salpausselkä end moraine, is likely to have been a relatively harsh environment when compared with the period after the event (c. 4100 4000 calbc), when increasing amounts of fine fertile soil were exposed by the lowering water level (Vikkula 1994b). It is likely that this exposure of fertile soil in the low-lying areas towards the end of the Climatic Optimum increased the resources offered by the environment and made it possible to maintain a higher population density than before. Similar environmental explanations have been proposed for the emergence and decline of the coastal Ostrobothnian Stone Age villages (Nun ~ ez and Okkonen 1999; Ylimaunu et al. 1999). The location of the villages in areas with gently sloping shores provided an environment rich in natural resources. The number of animal species in the osteological material from the Typical Comb Ware period is high compared to the material found in association with other pottery styles (Mökkönen 2001), which may be a reflection of the exploitation of the increased resources offered by the environment. 64

Unpublished Sources Alakärppä Jalo & Ojanlatva Eija 2000. Oulun Peurasuon myöhäiskivikautinen asuinpaikka. MA thesis, Department of Art Studies and Anthropology, University of Oulu. Europaeus, Aarne 1908. Kiinteitä muinaisjäännöksiä, Kivijärven ja Pihtiputaan pitäjät. Manuscript in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Foote, Don Charles 1965. Exploitation and Resource Utilization in Northwestern Arctic Alaska Before 1855. Ph. D. dissertation, McGill University. Montreal. Halinen, Petri 1995. Ounasjärven alueen esihistoriallisten peuranpyytäjäyhteisöjen asutusmallit. Lic. Phil. thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Halinen, Petri 2000. Kerimäki Raikuu Martinniemi. Kivikautisen asuinpaikan kaivaus 1998. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Halinen, Petri & Lavento, Mika 1997. Kesälahden inventointi 1997. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Havas, Honna 1999. Innland uten landegrenser. Bosetningsmodeller i det nordligste Finland og Norge i perioden 9000 6000 BP. MA thesis, University of Tromsø Department of Archaeology. Heinonen, Jari 2001. Saarijärvi, Rusavierto, alkuaineanalyysi (SEM/EDS). Report of trace element analysis made of the metal artefact (NM 32195:1) in Leskinen 2001: Appendix. Heinäaho, Heli 1997. Hylkeenpyynti ja traaninkäyttö Perämeren rannikolla myöhäisellä kivikaudella tarkasteltuna luuanalyysien sekä ekologisen ja etnologisen aineiston valossa. Seminar paper, Department of Art Studies and Anthropology, University of Oulu. Helskog, Ericka T. 1980. The Iversfjord Locality: A Study of Behavioral Patterning during the Late Stone Age of Finnmark, North Norway. Ph. D. dissertation. University of Wisconsin, Madison. 240 Joensuu, Jutta 2000. Kerimäen Raikuun Martinniemi 3:n keramiikka-aineisto. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Jussila, Pirjo 1992. Kasvijäännetutkimus kahdella kivikautisella asuinpaikalla. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Jussila, Timo 1991. Puumala. Muinaisjäännösten täydennysinventointi 1991. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Jussila, Timo 1997. Pihtiputaan muinaisjäännösten inventointi 1996. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Kankaanpää, Jarmo 1999. Tervola 94a Kauvonkangas. Kivikautisten asuinpainanteiden kaivaus 3.7. 31.8.1998. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Katiskoski, Kaarlo 1999. Puumala Kärmelahti. Kivikautisen asumuspainanteen ja asuinpaikan kaivaus 1998. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Katiskoski, Kaarlo 2000. Puumala Kärmelahti. Kivikautisen asumuspainanteen ja asuinpaikan kaivaus 1999. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Katiskoski, Kaarlo & Mökkönen Teemu 2000. Puumala Paimenhuhta. Kivikautisen asuinpaikan tarkastus 1999. Inspection report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Kirkinen, Tuija 1994. Etelä-Savon ja Karjalan nuoremman rautakauden asutuksen suhde luonnonympäristöön ja taloudelliseen ympäristöön. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Koivikko, Minna 1999. Etelä-Saimaan vedenalaiset asuinpaikat kivikautisten kohteiden inventointi ja topografisen mallin rakennus. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki.

Kotivuori, Hannu 1987. Tervola 94a Kauvonkangas, Siirtola. Map on file in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Kotivuori, Hannu, Heikkinen, Markku & Pärssinen, Timo 1985. Tervola 94 Kauvonkangas, Siirtola. Map on file in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Kuokkanen, Timo 2001. Nakkilan Rieskaronmäen pronssikautinen talo. Paper presented at the seminar Tila mielentila in Hämeenlinna, 28.4.2001. Kylli, Jukka 2000. Maatalouden alku Espoon ympäristössä; asutusmalli toimeentulostrategian kuvastajana. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Lahelma, Antti 2000. Landscapes of the Mind. A Contextual Approach to Finnish Rock-art. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Lavento, Mika 2001. Kerimäki Raikuu Martinniemi. Kaivaus 1999. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Lehtinen, Leena & Sepänmaa, Timo 1990. Puumalan muinaisjäännösinventointi 1989. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Leskinen, Sirpa 2000. Saarijärvi Rusavierto. Kivikautisen asuinpaikan kaivaus 1999. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Leskinen, Sirpa 2001. Saarijärvi Rusavierto. Kivikautisen asuinpaikan kaivaus 2000. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Mannermaa, Kristiina 1999a. Puumala Kärmelahti. Osteological analysis report in Katiskoski 1999: Appendix. Mannermaa, Kristiina 1999b. Puumala Kärmelahti. Osteological analysis report in Katiskoski 2000: Appendix. Mannermaa, Kristiina 1999c. Puumala Paimenhuhta. Osteological analysis report in Katiskoski & Mökkönen 2000: Appendix. Mannermaa, Kristiina 2000. Saarijärvi Rusavierto. Osteological analysis report in Leskinen 2000: Appendix. Martio, Lotta. 2000. Kerimäki Raikuu Martinniemi 3. Kivikautisen asuinpainanteen kohteensisäinen analyysi. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Mökkönen, Teemu 1999. Saimaa Laatokka-projektin inventointi. Kerimäen, Kesälahden, Kiteen ja Savonrannan osainventointi 1998. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Mökkönen, Teemu 2000. Saimaan vesistöalueen pyyntikulttuurien toimeentulo ja asutusmallit Kerimäen tapaustutkimus. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Niskanen, Markku 1999. Peurasuon analysoitu luumateriaali vuodelta 1999. Osteological analysis report, Laboratory of Archaeology, University of Oulu. Norberg, Erik 1997. Saivaara II. En studie av kvartsmaterialet från tre anläggningar och tankar kring detta. MA thesis in archaeology, University of Umeå. Okkonen, Jari 1994. Ii Olhava Ritakangas S. Röykkiö- ja asuinpaikka-alueen tarkastus 1994. Inspection report, Laboratory of Archaeology, University of Oulu. Okkonen, Jari 1997. Haukipudas Navettakangas SE. Asumuspainanteen ja kuopan tarkastus 1997. Inspection report, Laboratory of Archaeology, University of Oulu. Okkonen, Jari 1998. Muinaiset kivirakennelmat Keskija Pohjois-Pohjanmaalla. Lic. Phil. thesis, Department of Art Studies and Anthropology, University of Oulu. Peltonen, Nina 2001. Saarijärvi, Rusavierto. Osteological analysis report in Leskinen 2000: Appendix. Pesonen, Petro 1996. Varhainen asbestikeramiikka. Lic. Phil. thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Rautiainen, Pirjo 1995. Kuivaniemen Veskankankaan myöhäismesoliittinen asuinpaikka, luonnonresurssit ja rakennelmat. MA thesis, Institute for Cultural Studies, Department of Comparative and Finnish Archaeology, University of Turku. Riekki, Jari 1999. Villi Pohjola. Todisteet aggressiosta Kierikin nuolenkärjissä. Seminar paper, Department of Art Studies and Anthropology, University of Oulu. Roine, Leena 1964. Kertomus Pihtiputaan muinaisjäännösten inventoinnista kesällä 1964. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Ruonavaara, Leena 1988. Porvoon Böle varhaiskampakeraamisena aikana. Kivikautisen asuinpaikan löytö- 241

aineiston tilastollinen käsittely. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Sarkkinen, Mika 1998a. Pattijoen inventointi 1997. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Sarkkinen, Mika 1998b. Ruukin inventointi 1997. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Sarkkinen, Mika 1998c. Siikajoen inventointi 1997. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Sartes, Minna 1991. Kivikauden toimeentulostrategiat ja asutussysteemit. Turun alueen kivikautisten asuinpaikkojen topografian ja keskinäisten suhteiden analyysi. MA thesis, Institute for Cultural Studies, Department of Comparative and Finnish Archaeology, University of Turku. Saukkonen, Jyri 1990. Kampakeraaminen asutus Närvijoen alueella. MA thesis, Institute for Cultural Studies, Department of Archaeology, University of Helsinki. Schanche, Kjersti 1988. Mortensnes En studie av samfunn og materiell kultur gjennom 10.000 år. MA thesis in archaeology, University of Tromsø, Department of Archaeology. Schulz, Eeva-Liisa 1986. Tervola 30 Törmävaara. Kertomus vuoden 1985 kaivauksesta. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Schulz, Hans-Peter 1996a. Saarijärven Summassaaren ja Mäkelänlammen kaivaukset 30.8. 30.9.1995. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Schulz, Hans-Peter 1996b. Saarijärvi Summassaari ja Pyhäjärvi. Kivikautisten asuinpaikkojen koekaivaukset ja pyyntikuoppajärjestelmän kartoitus 30.8. 30.9.1994. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Sepänmaa, Timo 1992. Kerimäen muinaisjäännösinventointi vuonna 1991. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Sepänmaa, Timo 1995. Savonrannan inventointi 1995. Survey report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Setälä, Jukka 1988. Tervola 94a Kauvonkangas, Siirtola. Map on file in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. Tenhunen, Tanja 1999. Makrofossiilitutkimus. Puumala Kärmelahti. Macrofossil analysis report in Katiskoski 1999: Appendix. Tenhunen, Tanja 2001. Makrofossiilitutkimus. Saarijärvi Rusavierto. Macrofossil analysis report in Leskinen 2001: Appendix. Ukkonen, Pirkko 1995. Saarijärvi Rusavierto. Osteological analysis report in Schulz 1996a: Appendix. Vanhatalo, Simo 2000. Saarijärvi 55 Haikanniemi N. Kivikautisen asuinpaikan ja painanteen koekaivaus 1998. Excavation report in the archive of the National Board of Antiquities, Department of Archaeology, Helsinki. 242

Literature Ailio, Julius 1909a. Die Steinzeitlichen Wohnplatzfunde in Finland. I. Teil. Übersicht der Funde. Helsingfors. Ailio, Julius 1909b. Die Steinzeitlichen Wohnplatzfunde in Finland. II. Teil. Beschreibung der Funde. Helsingfors. Ailio, Julius 1923. Ovatko Pohjanmaan jättiläislinnat muinaisjäännöksiä? Suomen Museo 1922, 1 19. Alakärppä, Jalo 1999. Kuoppa-asumusten lämmityskokeilu Kierikissä. Muinaistutkija 4/1999, 50 57. Alakärppä, J. & Ikäheimo, J. & Ojanlatva, E. 1998. Oulun Peurasuo 1997. Myöhäiskivikautisen asuinpainanteen kaivaustutkimus. Meteli. Oulun yliopiston arkeologian tutkimusraportti 15. Almgren, Oscar 1912. En uppländsk bronsåldershydda. Fornvännen 1912, 132 151. Andersen, Niels H. 1975. Sarup. Et befaestet neolitisk anlaeg på Sydvestfyn. KUML 1973 74, 109 120. Andrefsky, William Jr. 1998. Lithics. Macroscopic approaches to analysis. Cambridge. Appelgren, Hjalmar 1891. Suomen muinaislinnat. Suomen Muinaismuistoyhdistyksen Aikakauskirja XII: I LXXV, 1 220. Asplund, Henrik 1997. Kemiön suurpitäjän esihistoria. Kemiön suurpitäjän historia I, 213 282. Baudou, Evert 1995. Norrlands forntid ett historiskt perspektiv. Kungl. Skytteanska samfundets handlingar No 45 1995. Bengtsson, Lisbet 1998. Dilemmas in Interpreting a Mesolithic Site. Proceedings from the Third Flint Alternatives Conference at Uppsala, Sweden, October 18 20, 1996. Occasional Papers in Archaeology 16. Uppsala, 55 60. Bertvall, Cecilia 1991. Fattenborg: Enorma hyddbottnar från bronsåldern. Populär arkeologi. Årg 9 Nr 2 1991, 17 19. Bilund, Antti 1993. Enonkosken muinaisjäännösinventointi 1992. Sihti 3, Savonlinnan maakuntamuseon julkaisuja, 9 12. Broadbent, Noel 1979. Coastal Resources and Settlement Stability. A Critical Study of a Mesolithic Site Complex in Northern Sweden. Aun 3. Broadbent, Noel D. & Knutsson, Kjel 1975. An Experimental Analysis of Quartz Scrapers. Results and Applications. Fornvännen 70, 113 128. Callahan, Errett 1987. An Evaluation of the Lithic Technology in Middle Sweden during the Mesolithic and Neolithic. Aun 8. Callahan, Errett, Forsberg, Lars, Knutsson, Kjel & Lindgren, Christina 1992. Frakturbilder. Kulturhistoriska kommentarer till det säregna sönderfallet vid bearbetning av kvarts. Tor 24, 27 63. Carpelan, Christian 1976. Nordfinska kåtabottnar från förhistorisk och historisk tid. Iskos 1, 30 34. Carpelan, Christian 1992. Juikenttä näkökulma saamelaiseen yhteiskuntaan. Suomen varhaishistoria. Studia historica septentrionalia 21, 34 44. Carpelan, Christian 1999. Käännekohtia Suomen esihistoriassa aikavälillä 5100 1000 ekr. Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. Bidrag till kännedom av Finlands natur och folk 153, 249 280. Carpelan, Christian 2001. Zur zeitgenössischen archäologischen Forschung in Finnland. Finnisch- Ugrischen Forschungen 56, 254 354. Carpelan, Christian & Lavento, Mika 1996. Soil phosphorus survey at subrecent Saami winter village sites near Inari, Finnish Lapland A preliminary report. Proceedings from the 6th Nordic Conference of Scientific Methods in Archaeology, Esbjerg 1993. Arkæologiske Rapporter fra Esbjerg Museum nr. 1, 1996, 97 107. Casselberry, Samuel E. 1974. Further refinement of formulae for determining population from floor area. World Archaeology 6 (1), 117 122. Casteel, Richard W. 1972. Two static maximum population-density models for hunter-gatherers: a first approximation. World Archaeology 4 (1), 19 40. Clark, Anette McFadyen & Clark, Donald W. 1974. Koyukon Athapaskan houses as seen through oral tra- 243

dition and through archaeology. Arctic Anthropology XI Suppl., 29 38. Clark, Donald W. 1981 Prehistory of the Western Subarctic. Handbook of North American Indians Vol. 6, Subarctic, 107 129. Cook, Sherburne F. 1972. Prehistoric Demography. Current Topics in Anthropology: theory, methods and content. Addison-Wesley Module in Anthropology No 16. Crowell, Aron 1988. Dwellings, Settlements, and Domestic Life. In Fitzhugh, W. W. & Crowell, A. (eds.), Crossroads of Continents; Cultures of Siberia and Alaska. Washington D.C., 194 208. Donner, Kai 1979. Siperian samojedien keskuudessa vuosina 1911 1913 ja 1914. Keuruu. Earle, Timothy, Beck, Jens-Henrik, Kristiansen, Kristian, Aperlo, Peter, Kelertas, Kristina & Steinberg, John 1998. The Political Economy of Late Neolithic and Early Bronze Age Society: The Thy Archaeological Project. Norwegian Archaeological Review. Vol. 31. No. 1. 1998, 1 28. Edgren, Torsten 1966. Jäkärlä-gruppen. En västfinsk kulturgrupp under yngre stenålder. Suomen Muinaismuistoyhdistyksen Aikakauskirja 64. Edgren, Torsten 1984. Kivikausi. Suomen historia I. Espoo, 8 97. Edgren, Torsten 1999. Alkavan rautakauden kulttuurikuva Länsi-Suomessa. Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. Bidrag till kännedom av Finlands natur och folk 153, 311 333. Elgstrøm, Ossian 1922. Karesuandolapparna. Stockholm. Engelmark, Roger 1985. Carbonized seeds in postholes a reflection of human activity. Iskos 5, 205 209. Engelstad, Ericka 1988. Pit-houses in Arctic Norway. An investigation of their typology using multiple correspondence analysis. Multivariate Archaeology. Numerical Approaches in Scandinavian Archaeology. Jutland Archaeological Society Publications XXI, 1988, 71 84. Erä-Esko, Aarni 1974. Kemijoen alueen arkeologisten tutkimusten historiaa. Kemijoki 8000. Kemijoen alueen muinaisuutta 8000 vuoden ajalta arkeologisten tutkimusten valossa, 5 14. Europaeus, Aarne 1913. Paavolan pitäjän jättiläiskirkot. Suomen Museo 1913, 80 91. Europaeus, Aarne 1916. Ojalankankaan kivikauden löytöpaikka Alavuudella. Suomen Museo 1916, 73 82. Europaeus, Aarne 1922. Fornfynd från Kyrkslätt och Esbo socknar. Suomen Muinaismuistoyhdistyksen Aikakauskirja XXXII, 1 208. Fletcher, Roland 1995. The limits of settlement growth. A theoretical outline. Cambridge. Florek, Waclaw, Koutaniemi, Kielo & Koutaniemi, Leo 1987. Past and present of the Oulujoki river and adjoining areas. Nordia 21/2, 119 132. Forsberg, Lars 1985. Site Variabilty and Settelement Patterns. Archaeology and environment 5. Forss, Aulis 1993. Jätinkirkkojen lähettyvillä olevat epämääräiset röykkiöt. Lapinraunioita ja hiidenkiukaita. Julkaisu/Museovirasto, arkeologian osasto, N:o 3, 10 16. Forss, Aulis 1996. Jätinkirkot Pohjanlahden pohjoisen rannikkoalueen arvoituksellinen muinaisjäännösryhmä. Muinaistutkija 1/1996, 26 35. Fortelius, Mikael 1981. Johdatus arkeologiseen luuanalyysiin. Museovirasto. Esihistorian toimiston julkaisuja n:o 1. Gestrin, Tryggve 1994. Gropar i klapper. Kentältä poimittua 2. Kirjoitelmia arkeologian alalta. Julkaisu/ Museovirasto, arkeologian osasto n:o 4, 17 21. Giddings, James L. 1964. The Archeology of Cape Denbigh. Brown University Press. Providence, RI. Gjessing, Gutorm 1942. Bøl ljegammen. En hypotese om opphavet. Norsk Geografisk Tidskrift, IX (2), 41 88. Gould, Richard A. 1966. Some Stone Artifacts of the Wonkongurun of South Australia. American Museum Novitates 2249. Gould, Richard A., Koster, Dorothy A. & Sontz, Ann H. L. 1971. The Lithic Assemblage of the Western Desert Aborigines of Australia. American Antiquity 36(2), 149 169. Grigalavičciene Elena 1995. Zalvario ir ankstyvasis geležies amžius Lietuvoje. Vilnius. Gurina, N. N. 1951 = Gurina, N. N. 1951. Posledeniö äpoxi neolita i rannego metalla na severnom pobereq<e Oneqskogo ozera. Materialy i issledovaniö po arxeologii SSSR, 20, 77 142. Gurina N. N. 1967 = Gurina, N. N. 1967. Iz istorii drevnyx plemen zapadnyx oblastej SSSR (po materialam narvskoj äkspedicii). Materialy i issledovaniö po arxeologii SSSR, 144. Leningrad. 244

Gurina N. N. 1989 = Gurina, N. N. 1989. Mezolit Karelii. Arxeologiö SSSR. Moskva, 27 31. Gustavsson, Kenneth 1997a. Multidisciplinary analyses on Bronze Age pottery from Otterböte, Åland. Iskos 11, 193 202. Gustavsson, Kenneth 1997b. Otterböte. New Light on a Bronze Age Site in the Baltic. Theses and Papers in Archaeology B:4. Stockholm University. Göthberg, Hans 1995. Huskronologi i Mälarområdet, på Gotland och Öland. Hus & gård. Artikeldel. Hus & gård i det förurbana samhället - Rapport från ett sektorforskningsprojekt vid Riksantikvarieämbetet. Riksantikvarieämbetet. Arkeologiska undersökningar. Skrifter nr 14, 65 109. Halén, Ove 1994. Sedentariness during the Stone Age of Northern Sweden in the light of the Alträsket Site, c. 5000 B.C. and the Comb Ware Site Lillberget, c. 3900 B.C. Source Critical Problems of Representativity in Archaeology. Acta Archaeologica Lundensia. Series in 4:o. No. 20. Halinen, Petri 1992. Enontekiön Markkina kolmen kodan kaivaukset vuonna 1990. Kentältä poimittua. Kirjoitelmia arkeologian alalta. Museoviraston esihistorian toimiston julkaisuja N:o 2, 39 44. Halinen, Petri 1997a. Ajoitustuloksia. Muinaistutkija 3/ 1997, 53. Halinen, Petri 1997b. Kaustisen Kankaan asuinpaikka ja punamultahaudat. Muinaistutkija 2/1997, 18 27. Halinen, Petri 1999a. Burial Practices and the Structure of Societies During the Stone Age in Finland. In Huurre, Matti. (ed.): Dig it all. Papers dedicated to Ari Siiriäinen. Helsinki, 173 180. Halinen, Petri 1999b. Esihistorialliset saviastiat ovat olleet pyöreitä. Muinaistutkija 2/1999, 12 15. Halinen, Petri 1999c. Suomen liikenteen esihistoria: kivi- ja pronssikausi. In Mauranen, Tapani (ed.): Maata, jäätä, kulkijoita. Tiet, liikenne ja yhteiskunta ennen vuotta 1860. Tuhat vuotta teitä, kaksisataa vuotta tielaitosta. Helsinki, 36 51. Halinen, Petri, Joensuu, Jutta, Lavento, Mika, Martio, Lotta & Mökkönen, Teemu Kerimäen Raikuun Martinniemen tutkimukset 1998 1999. Sihti 5, Savonlinnan maakuntamuseon julkaisuja. In press. Halinen, Petri, Katiskoski, Kaarlo & Sarkkinen, Mika 1998. Yli-Iin Kuuselankankaan asuinpaikan tutkimukset 1994 1996. Kentältä poimittua 4. Kirjoitelmia arkeologian alalta. Museoviraston arkeologian osaston julkaisuja N:o 7, 24 40. Halkka, Antti 1995. Jään ja hylkeiden meri. In Lokki, Juhani & Miettinen, Kaarina (eds.): Kaunis kotimaa, osa 3. Helsinki, 30 35. Hall, Edwin Jr 1976. A Preliminary Analysis of House Types at Tutko Lake, Northern Alaska. Archaeological Survey of Canada, no. 49, 1 101. Hamari, Pirjo 1996. Suorakaiteen muotoiset kivilatomukset Pohjois-Suomessa. Kentältä poimittua 3. Kirjoitelmia arkeologian alalta. Museoviraston arkeologian osaston julkaisuja N:o 6, 46 58. Hamari, Pirjo & Halinen, Petri 2000. Saamelaisperäisten muinaisjäännösten inventointi. In Kirkinen, Tuija & Maaranen, Päivi (eds.): Arkeologinen inventointi. Opas inventoinnin suunnitteluun ja toteuttamiseen. Helsinki, 153 171. Hayden, Brian, Franco, Nora & Spafford, Jim 1996. Evaluating Lithic Strategies and Design Criteria. In Odell, George H. (ed.): Stone Tools: Theoretical Insights into Human Prehistory. New York, 9 49. Helliwell, Christine 1996. Space and sociality in a Dayak Longhouse. In Jackson, Michael (ed.): Things as They Are: new directions in phenomenological anthropology. Indiana University Press, 128 148. Helskog, Ericka Thrash 1983. The Iversfjord Locality. A Study of Behavioral Patterning during Late Stone Age of Finnmark, North Norway. Tromsø Museums Skrifter Vol. XIX. Helskog, Knut 1984. The younger Stone Age Settlements in Varanger, North Norway. Acta Borealia 1/1984, 39 70. Hertell, Esa & Manninen, Mikael A. Rävåsenin kvartsiaineisto. Finskt Museum 1996. In press. Hesjedal, Anders, Damm, Charlotte, Olsen, Bjørnar & Storli, Inger 1996. Arkeologi på Slettnes. Dokumentasjon av 11.000 års bosetning. Tromsø Museums Skrifter XXVI. Hiekkanen, M. 1984 = Xiekkanen, M. 1984. Otlihitel<nye osobennosti postroek tipa madeneva, otnosöщixsö k kamennomu veku. Madenevan tyypin kodanpohjat eräitä viitteitä kotakuntien perinne-eroista kivikaudella. In Rybakov, V. A. (ed.): Novoe v arxeologii SSSR i Finlöndii. Leningrad, 46 53. Hiekkanen, Markus & Seger, Tapio 1988. Beyond Postholes: An investigation of Pre-Roman house remains at Mickels in Espoo, S. Finland. Fennoscandia Archaeologica V (1988), 21 34. Hirviluoto, Anna-Liisa 1985. Rautakausi. Kivikaudesta keskiaikaan. Porvoon seudun esihistoriaa. Porvoon museoyhdistyksen julkaisuja nro 1, 58 73. 245

Hirviluoto, Anna-Liisa 1991. Salon esihistoria. Jyväskylä. Honkanen, Pekka 1981. Uudenmaan kansainvaellus- ja merovingiaika. Helsingin yliopiston arkeologian laitos. Moniste n:o 26. Hunter-Anderson, R. L. 1977. A theoretical approach to the study of house form. In Binford, L. R. (ed.): For Theory Building in Archaeology. Essays on Faunal Remains, Aquatic Resources, Spatial Analysis, and Systemic Modeling. Studies in Archaeology. New York, 287 315. Huurre, Matti 1983. Pohjois-Pohjanmaan ja Lapin esihistoria. Pohjois-Pohjanmaan ja Lapin historia 1. Kuusamo. Huurre, Matti 1986. Ihmisenkuvia. Uusia ja vanhoja löytöjä Suomesta. Iskos 6, 46 51. Huurre, Matti 1991. Oulujokilaakson esihistoria. In Huurre, Matti & Vahtola, Jouko (eds.): Oulujokilaakson historia kivikaudelta vuoteen 1865. Oulu, 11 70. Huurre, Matti 1998a. Arkeologinen tutkimus Kainuussa. Varhain Pohjoisessa. Johdanto. Varhain Pohjoisessa -hankkeen artikkeleita. Helsinki Papers in Archaeology no. 11, 20 24. Huurre, Matti 1998b. Kivikauden Suomi. Helsinki. Indreko, R. 1932. Der Siedlungsfund von Moksi, Gemeinde Võisiku. Sitzungsberichte der Gelehrten Estnischen Gesellschaft 1930. Tartu, 197 218. Itkonen, Ilmari 1913. Tietoja Inarin kirkonkylän seudun muinaisuudesta. Suomen Museo 1913, 2 9. Itkonen, T. 1918. Eräs kaivaus Kuolanniemellä. Suomen Museo 1918, 35 38. Itkonen, T.I. 1948. Suomen lappalaiset vuoteen 1945 I. Porvoo. Itkonen, T.I. 1984. Suomen lappalaiset vuoteen 1945. Ensimmäinen osa. Porvoo. Jaanits, L. Ju. 1954 = Önits, L. Å.1954. Novye dannye po neolitu Pribaltiki. Sovetskaö arxeologiö, XIX, 159 204. Jaanits, L. Ju. 1959a = Önits, L. Å.1959a. Neolotiheskoe poselenie Valma. Trudy Pribaltijskoj ob edinennoj äkspedicii, I. Moskva, 114 123. Jaanits, L. Ju. 1959b = Önits, L. Å.1959b. Poseleniö äpoxi neolita i rannego metalla v prisust<e r. Ämajygi (Ästonskaö SSR). Tallin. 246 Jaanits, L. Ju. 1973 = Önits, L. Å. 1973. Neolitiheskie pamötniki Ästonii i ix xronologiö. Ätnokul<turnye obщnosti lesnoj i lesostepnoj zony evropejskoj hasti SSSR v äpoxu neolita. Materialy i issledovaniö po arxeologii SSSR, 172. Leningrad, 202 209. Jaanits, L. 1988. Eesti sooarheoloogiast. Eesti sood. Tallinn, 217 221. Jaanits, L. & Jaanits, K. 1975. Frühmesolithische Siedlung in Pulli. Eesti NSV Teaduste Akadeemia Toimetised. Ühiskonnateadused, 1, 64 70. Jaanits, L. & Jaanits, K. 1978. Ausgrabungen der frühmesolithischen Siedlung von Pulli. Eesti NSV Teaduste Akadeemia Toimetised. Ühiskonnateadused, 1, 56 63. Jaanits, L., Laul, S., Lõugas, V. & Tõnisson, E. 1982. Eesti esiajalugu. Tallinn. Janes, Robert R. 1983. Archaeological ethnography among Mackenzie basin Dene, Canada. The Arctic Institute of North America. Technical Paper No. 28. Jarva, Eero & Okkonen, Jari 1993. Röykkiöiden inventointikokemuksia Pohjois-Suomesta. Lapinraunioita ja hiidenkiukaita. Julkaisu/Museovirasto, arkeologian osasto n:o 3, 93 108. Jussila, Pirjo 1992. Pieksämäen kasvijäännetutkimuksia. Sihti 2, Savonlinnan maakuntamuseon julkaisuja, 30 35. Jussila, Pirjo 1994. Studies of Plant Remains at two Stone Age Dwelling Sites in Pieksämäki, Eastern Finland. Fenno-Ugri et Slavi 1992. Prehistoric economy and means of livelihood. Julkaisu/Museovirasto, arkeologian osasto n:o 5, 16 22. Jussila, Pirjo 1996. Site exploitation through macrofossil analysis at different settlement stages of the Pörrinmökki site in Rääkkylä. Environmental Studies in Eastern Finland. Reports of the Ancient Lake Saimaa Project. Helsinki Papers in Archaeology no. 8, 119 128. Jussila, Timo 1993. Kerimäki Raikuu Martinniemi Kivi-, pronssi- ja rautakautisen pyyntiasuinpaikkaalueen fosfori- ja yleiskartoitus. Sihti 3, Savonlinnan maakuntamuseon julkaisuja, 29 38. Jussila, Timo 1994. Kaivausteknisiä kokeiluja. Kentältä poimittua 2. Kirjoitelmia arkeologian alalta. Julkaisu/ Museovirasto, arkeologian osasto, n:o 4, 130 140. Jussila, Timo 1999. Saimaan kalliomaalausten ajoitus rannansiirtymiskronologian perusteella. Saimaan ja Päijänteen kalliomaalausten sijainti ja syntyaika. Kalliomaalausraportteja 1/1999, 113 133.

Jussila, Timo 2000. Pioneerit Keski-Suomessa ja Savossa. Rannansiirtymisajoitusmenetelmien perusteita ja vertailua. Muinaistutkija 2/2000, 13 28. Jussila, Timo, Lehtinen, Leena, Nurminen, Teija, Raike, Eeva & Sepänmaa Timo 1992. Etelä-Savon muinaisjäännösinventointi 2000 -projekti. Sihti 2, Savonlinnan maakuntamuseon julkaisuja, 20 29. Jutikkala, Eino 1987. Kuolemalla on aina syynsä: maailman väestöhistorian ääriviivoja. Porvoo. Karjalainen, Taisto 1996a. Kuhmo 14 Vasikkaniemi SW asuinpaikkavyöhykkeen tutkimus- ja tulkintaongelmia. Kentältä poimittua 3. Museoviraston arkeologian osaston julkaisuja N:o 6, 15 22. Karjalainen, Taisto 1996b. Outokumpu Sätös ja Orov Navolok 16, talo 3. Muinaistutkija 1/1996, 13 18. Karjalainen, Taisto 1996c. Pithouse in Outokumpu Sätös excavated in 1992 1994. Pithouses and Potmakers in Eastern Finland. Reports of the Ancient Lake Saimaa Project. Helsinki Papers in Archaeology no. 9, 71 88. Karjalainen, Taisto 1998. Lintutorni ja Ritokangas painanneasuinpaikkojen tutkimusmateriaalien eroja. Kentältä poimittua 4. Museoviraston arkeologian osaston julkaisuja N:o 7, 88 94. Karjalainen, Taisto 1999. Sedentariness and Dating Stone Age Houses and Sites. In Huurre, Matti (ed.): Dig it all. Papers dedicated to Ari Siiriäinen. Helsinki, 185 190. Katiskoski, Kaarlo 1997. Uusimmat tutkimukset Vantaan Stenkullan kivikautisella asuinpaikalla. Helsingin pitäjä 1997, 6 9. Katiskoski, Kaarlo. Asumusjäänteet Vantaan Stenkullan kivikautisella asuinpaikalla. Finskt Museum 1996. In press. Kirkinen, Tuija 1996. Use of a Geographical Information System (GIS) in Modelling the later Iron Age settlement in eastern Finland. Environmental Studies in Eastern Finland. Reports of the Ancient Lake Saimaa Project. Helsinki Papers in Archaeology no. 8, 19 61. Kivikoski, Ella 1946. Husgrunderna i Storhagen, Kulla, Finström. Suomen Muinaismuistoyhdistyksen Aikakauskirja XLVII:3. Knutsson, Kjel 1988a. Making and Using Stone Tools. The analysis of the lithic assemblages from Middle Neolithic sites with flint in Västerbotten, northern Sweden. Aun 11. Knutsson, Kjel 1988b. Patterns of tool use. Scanning electron microscopy of experimental quartz tools. Aun 10. Knutsson, Kjel 1998. Convention and lithic analysis. Proceedings from the Third Flint Alternatives Conference at Uppsala, Sweden, October 18 20, 1996. Occasional Papers in Archaeology 16, 71 93. Koivunen, Pentti 1996. Kodanpohjia ja meripihkaa Kierikistä. Muinaistutkija 1/1996, 2 7. Koivunen, Pentti 1997. Teoria jätinkirkkojen käyttötarkoituksesta. Muinaistutkija 4/1997, 49 52. Koivunen, P. & Makkonen, M. 1998. Yli-Iin Kuuselankankaan kaivaukset 1993 1994. Meteli. Oulun yliopiston arkeologian laboratorion tutkimusraportti 16. Kokkonen, Jyri 1978. Kymin Niskasuon keramiikkalöydöt. Helsingin yliopiston arkeologian laitos. Moniste n:o 17. Kol cov, L. V. 1989 = Kol<cov L. V. 1989. Mezolit Volgo-Okskogo meqdureh<ö. Arxeologiö SSSR. Moskva, 68 86. Koponen, Onni E. 1983. Pielisen museon kertomaa III. Pielisen Karjalan talonpoikaisrakennuksista. Pielisen museon julkaisuja numero 3. Korteniemi, Markku 1998. Jätinkirkkojen funktiosta. Muinaistutkija 2/1998, 45 49. Kotivuori, Hannu 1988. Varhaismetallikautisen asuinpaikan koekaivaus Vöyrissä. Faravid 11, 1987, 35 52. Kotivuori, Hannu 1993a. Kivikauden asumuksia Peräpohjolassa vertailua ja rakenteellisia tulkintoja. Selviytyjät. Näyttely pohjoisen ihmisen sitkeydestä. Lapin maakuntamuseon julkaisuja 7, 120 160. Kotivuori, Hannu 1993b. Pohjanlahden kiveliöt muinaisen toiminnan tyyssijoina. Lapinraunioita ja hiidenkiukaita. Julkaisu/Museovirasto, arkeologian osasto, N:o 3, 17 30. Kotivuori, Hannu 1994. Sierijärven tupasija ja muita lappalaismuistoja Rovaniemellä. Kentältä poimittua 2. Julkaisu/Museovirasto, arkeologian osasto, N:o 4, 41 50. Kotivuori, Hannu 1996a. Mietteitä alisen Kemijoen kivikauden asuinmuistoista. Muinaistutkija 1/1996, 8 12. Kotivuori, Hannu 1996b. Pyytäjistä kaskenraivaajiksi. Kotatulilta savupirtin suojaan. Rovaniemen historia vuoteen 1721. Jyväskylä, 36 125. Kotivuori, Hannu 1998. Försänkta husgrunder från yngre stenålder i nedre Kemiälvdalen, norra Finland. Hus 247

och tomt i Norden under förhistorisk tid. Bebyggelsehistorisk tidskrift Nr 33 1997, 153 154. Kriiska, Aivar 1996a. Archaeological excavations on the Neolithic site of Riigiküla IV. Eesti Teaduste Akadeemia Toimetised. Humanitaar- ja sotsiaalteadused, 45:4, 1996, 410 419. Kriiska, Aivar 1996b. Archaeological studies on the Kõpu Peninsula. Eesti Teaduste Akadeemia Toimetised. Humanitaar- ja sotsiaalteadused, 45:4, 398 409. Kriiska, Aivar 1998. Mesoliitilised asustusjäljed Loode- Saaremaal. Ajalooline Ajakiri, 1 (100), 13 22. Kriiska, Aivar 2000. Corded Ware Culture Sites in North-Eastern Estonia. De temporibus antiquissimis ad honorem Lembit Jaanits. Muinasaja teadus, 8. Tallinn, 59 79. Kriiska, Aivar & Lõugas, Lembi 1999. Late Mesolithic and Early Neolithic Seasonal Settlement at Kõpu, Hiiumaa Island, Estonia. Environmental and Cultural Hiatory of the Eastern Baltic Region. PACT 57, 157 172. Kriiska, Aivar & Saluäär, Ulla 2000. Archaeological fieldwork on the island of Ruhnu. Arheoloogilised välitööd Eestis 1999, 18 28. Kylli, Jukka 2001. Asutussysteemi ja toimeentulo muinaisessa Espoossa ja lähiympäristössä. Muinaistutkija 1/2001, 2 13. Lahelma, Antti 2001. Kalliomaalaukset ja shamanismi. Tulkintaa neuropsykologian ja maalausten sijainnin valossa. Muinaistutkija 2/2001, 2 21. Lang, Valter 1995. Varane maaviljelus ja maaviljelusühiskond Eestis: ääremärkusi mõningate arengutendentside kohta. Eesti arheoloogia historiograafilisi, teoreetilisi ja kultuuriajaloolisi aspekte. Muinasaja teadus, 3, 116 181. Lang, Valter 1996. Muistne Rävala. Muistised, kronoloogia ja maaviljelusliku asutuse kujunemine Loode-Eestis, eriti Pirita jõe alamjooksu piirkonnas. Muinasaja teadus 4. Lang, Valter 1999. Kenneth Gustavsson, Otterböte. New Light on a Bronze Age Site in the Baltic. Theses and Papers in Archaeology B:4. Stockholm University 1997, 141 p. + 8 appendices. Fennoscandia archaeologica XVI, 81 84. Lanne, Erkki & Pernu, Teuvo 1974. Muhoksen muodostuman seismisesti määritetyt rajat. Department of Geophysics, University of Oulu, Contribution No. 46. Larsson, Mats 1998. Under tak. Från jägarens hydda till vikingens långhus i Sydsverige. Hus och tomt i Norden under förhistorisk tid. Bebyggelsehistorisk tidskrift Nr 33 1997, 53 62. Lavento, Mika 2001. Textile Ceramics in Finland and on the Karelian Isthmus. Nine Variations and Fugue on a Theme of C.F. Meinander. Suomen Muinaismuistoyhdistyksen Aikakauskirja 109. Lee, Richard B. & DeVore, Irven 1984. Problems in the Study of Hunters and Gatherers. In Lee, Richard B. & DeVore, Irven (eds.): Man the Hunter. Chicago, 3 12. Leem, Knud 1767. Beskrivelse over Finmarkens Lapper. København. Lehtinen, Leena & Sepänmaa, Timo 1995. Surveys of Archaeological Sites in Southern Savo. Fennoscandia archaeologica XII, 77 82. Liedgren, Lars 1989a. Bebyggelseutveckling i Forsa, Hälsingland, under den äldre järnåldern. Arkeologi i norr 2, 45 81. Liedgren, Lars 1989b. Boplatsundersökningen på Kalaschabrännan: Preliminära aspekter på bebyggelsen. Studia Archaeologica Ostrobotniesia 1988, 31 56. Lindgren, Christina 1994. Ett bipolärt problem om kvartsteknologi under mesolitikum. Aktuell Arkeologi IV. Stockholm Archaeological Reports 29, 77 86. Lindgren, Christina 1996. Kvarts som källmaterial exempel från den mesolitiska boplatsen Hagtorp. Tor 28, 29 52. Lindgren, Christina 1998. Shapes of quartz and shapes of minds. Proceedings from the Third Flint Alternatives Conference at Uppsala, Sweden, October 18 20, 1996. Occasional Papers in Archaeology 16, 95 103. Lindman, Gundela 1985. Förhistoriska aggressionsstrukturer i det västsvenska landskapet. Gotarc. Series B. Gothenbur Archaeological Theses No 2. Loeffler, David & Westfal, Ulf 1985. A Well-preserved Stone Age Dwelling Site. Preliminary Presentation of the Investigations at Vuollerim, Lapland, Sweden. In honorem Evert Baudou. Archaeology and Environment 4, 425 434. Luho, Ville 1948. Suomen kivikauden pääpiirteet. Helsinki. Luho, Ville 1956. Die Askolakultur. Suomen Muinaismuistoyhdistyksen aikakauskirja 57. Luho, Ville 1967. Die Suomusjärvi-Kultur. Die Mittelund Spätmesolithische Zeit in Finnland. Suomen Muinaismuistoyhdistyksen Aikakauskirja 66. 248

Lundberg, Åsa 1997. Vinterbyar ett bandsamhälles territorier i Norrlands inland, 4500 2500 f.kr. Studia Archaeologica Universitatis Umensis 8. Løken, Trond 1998. Det forhistoriske huset i Rogaland belyst ved flateavdekkende utgravninger. Hus och tomt i Norden under förhistorisk tid. Bebyggelsehistorisk tidskrift Nr 33 1997, 169 184. Malmer, Mats 1988. Alvastra pålbyggnad. In Burenhult, Göran (ed.): Länkar till vår forntid. Stockholm. Matiskainen, Heikki 1989. The Paleoenvironment of Askola, Southern Finland. Mesolithic settlement and subsistence 10 000 6000 b.p. Studies on the Chronology, material Culture and Subsistence Economy of the Finnish Mesolithic, 10 000 6000 b.p.. Iskos 8, 1 97. Matiskainen, Heikki & Jussila, Timo 1984. Naarajärven kampakeraaminen asumus. Suomen Museo 1984, 17 52. Meinander, C. F. 1940. Pyheensilta stenåldersboplats. Finskt Museum 1939, 28 43. Meinander, C. F. 1950. Esihistoria. Etelä-Pohjanmaan historia I. Helsinki. Meinander, C. F. 1954a. Die Bronzezeit in Finnland. Suomen Muinaismuistoyhdistyksen Aikakauskirja 54. Meinander, C. F. 1954b. Die Kiukaiskultur. Suomen Muinaismuistoyhdistyksen Aikakauskirja 53. Meinander, C. F. 1964. Pihtiputaan kivikautta. Pihtiputaan kirja. Jyväskylän yliopiston kotiseutukirja 2. Pieksämäki, 29 40. Meinander, C. F. 1971. Radiokarbondateringar till Finlands stenålder. Societas Scientiarum Fennica. Årsbok vuosikirja XLVIII, 1969 1970, 3 14. Meinander, C. F. 1976. Hyddbottnar av Madeneva-typ. Iskos 1, 26 29. Melander, Jan 1989. Analys av lerklining från ett järnåldershus i Hälsingland. Arkeologi i norr 2, 83 94. Miettinen, Mirja 1982a. Den förhistoriska tiden i Petalax. Petalax historia I, 7 78. Miettinen, Mirja 1982b. Stenåldersboplatsen Hundbacka i Pedersöre, Österbotten. Bottnisk kontakt I. Föredrag vid maritimhistorisk konferens i Örnsköldsvik 12 14 februari 1982. Skrifter från Örnsköldviks museum nr 1, 14 21. Miettinen, Mirja 1983. Esihistoriallinen aika Etelä-Pohjanmaan Järviseudulla. Järviseudun historia I, 61 174. Miettinen, Mirja 1989. Den österbottniska kustbosättningen under tidig metallålder ca 1000 BC 200 AD. Sammanfattning av undersökningar på 1970- och 1980-talen. Bottnisk kontakt IV, 99 107. Miettinen, Mirja 1992. Laukaan Hartikan kivikautinen kalmisto. Keski-Suomi 18, 8 23. Miettinen, Mirja 1994. Tidig metallålder i Österbottens kustland: Nya arkeologiska forskningsresultat. Järnåldern i Mittnorden. Studier i Österbottens förhistoria nr 3. Acta antiqua Osterbotniensia, 155 172. Miettinen, Mirja 1999a. Finds of East Swedish Pitted Ware from the Rävåsen site in Kristiinankaupunki, Ostrobothnia. Fennoscandia archaeologica XVI, 71 74. Miettinen, Mirja 1999b. Muinaistutkimusta 119 vuotta Pihtiputaalla. Pihtiputaan Joulu 1999, 27 29. Miettinen, Mirja 2000. Pihtiputaan esihistoriaa. Pihtiputaan kirja II. Jyväskylän yliopiston ylioppilaskunnan julkaisusarja 37, 29 47. Miettinen, Timo 1998. Kymenlaakson esihistoriaa. Kymenlaakson maakuntamuseon julkaisuja no 26. Miettinen, Timo 1999. Mitä Susikopinharjulla tapahtui? Muinaistutkija 2/1999, 35 39. Moisanen, Jukka 1991. Tutkimuksia Kerimäen kivikautisilla asuinpaikoilla. Sihti 1, Savonlinnan maakuntamuseon julkaisuja, 25 32. Muurimäki, Eero 1995. Saarijärven museon kivikauden kylän rakennusennallistukset teoreettista taustaa. Muinaistutkija 2/1995, 3 11. Mäkivuoti, Markku 1995. Haukiputaan esihistoriaa muinaisjäännösten inventoinnin ja Hiidenkankaan tutkimusten valossa. Esitelmiä Haukiputaasta. Kellon Haukiputaan Kotiseutujulkaisu IV, 6 27. Mökkönen, Teemu 2000. Saimaan vedenkorkeuden vuodenaikaisvaihtelut nykyistä lämpimämmässä ilmastossa. Muinaistutkija 2/2000, 29 38. Mökkönen, Teemu 2001. Hauki on kala Saimaan vesistöalueen kivi- ja varhaismetallikauden osteologinen aineisto. Muinaistutkija 3/2001, 2 13. Nabakov, Peter & Easton, Robert 1989. Native American Architecture. New York. Naroll, Raoul 1962. Floor area & settlement population. American Antiquity 36, 587 589. Nelson, Edward William 1983 (1899). The Eskimo About Bering Strait. Bureau of Ethnology. Smithsonian Institution. Washington. 249

Nielsen, Poul Otto 1998. De ældste langhuse. Fra toskibede til treskibede huse i Norden. Hus och tomt i Norden under förhistorisk tid. Bebyggelsehistorisk tidskrift Nr 3 1997, 9 30. Niskanen, Markku 1998. Animal bones. Oulun Peurasuo 1997 myöhäiskivikautisen asuinpainanteen kaivaustutkimus. Meteli. Oulun yliopiston arkeologian laboratorion tutkimusraportti 15, 28 30. Nomina Rerum Mediaevalium 1998. Kuvitettu nimikkeistö esinekuvauksia varten. Göteborg. Nuñez, Milton 1990a. On Subneolithic pottery and its adoption in Late Mesolithic Finland. Fennoscandia archaeologica VII, 27 52. Nuñez, Milton 1990b. On the Food Resources Available to Man in Stone Age Finland. Finskt Museum 1990, 24 54. Nuñez, Milton & Okkonen, Jari 1999. Environmental Background for the Rise and Fall of Villages and Megastructures in North Ostrobothnia 4000 2000 cal BC. In Huurre, Matti (ed.): Dig it all. Papers dedicated to Ari Siiriäinen. Helsinki, 105 115. Nuñez, Milton & Uino, Pirjo 1998. Dwellings and related structures in prehistoric mainland Finland. Hus och tomt i Norden under förhistorisk tid. Bebyggelsehistorisk tidskrift Nr 33 1997, 133 152. Nyman, Adolf Leonard 1882. Muinaismuistoja Laukaan kihlakunnasta. Suomen Muinaismuistoyhdistyksen aikakauskirja V, 3 129. Okkonen, J. & Petäjä-Ronkainen, A. 1996. Geokemiallisia havaintoja Kemijokivarren kivikautisista asumuspainanteista. Meteli. Oulun yliopiston arkeologian laboratorion tutkimusraportti 10. Olsen, Bjørnar 1994. Bosetning og samfunn i Finnmarks forhistorie. Oslo. Olsen, Bjørnar 1998. Forhistoriske hus i Nord-Norge. Hus och tomt i Norden under förhistorisk tid. Bebyggelsehistorisk tidskrift Nr 33 1997, 185 194. Oswalt, Wendell H. 1967. Alaskan Eskimos. Chandler Publications in Anthropology and Society. San Francisco, 92 112. Oulujoki 1959. Oulujoki source of power. The power plants of Oulujoki power company, Helsinki. Pankruev, G. A. 1994. Karjalan mesoliittinen ja neoliittinen kausi. 1. Mesoliittinen kausi. Helsinki. Pankrušev, G. A. & Zuravlev, A. P. 1966.= Pankruwev, G. A. & Quravlev. A. P. 1966. Stoöanka Vigajnavolok I. In Pankruwev, G. A. (ed.): Novye pamötniki istorii drevnej Karelii. Moskva, Leningrad, 152 177. Pesonen, Petro 1994. Löytöjen dokumentointi, fosforianalyysi ja kodanpohjat Rääkkylän Pörrinmökin asuinpaikalla. Kentältä poimittua 2. Kirjoitelmia arkeologian alalta. Julkaisu/Museovirasto, arkeologian osasto, N:o 4, 56 67. Pesonen, Petro 1995. Hut Floor Areas and Ceramics Analysis of the Excavation Area in the Rääkkylä Pörrinmökki Settlement Site, Eastern Finland. Fennoscandia archaeologica XII, 139 150. Pesonen, Petro 1996a. Archaeology of the Jaamankangas area with special reference to the Rääkkylä Pörrinmökki Stone Age settlement site. Environmental Studies in Eastern Finland. Reports of the Ancient Lake Saimaa Project. Helsinki Papers in Archaeology no. 8, 93 117. Pesonen, Petro 1996b. Early Asbestos Ware. Pithouses and Potmakers in Eastern Finland. Reports of the Ancient Lake Saimaa Project. Helsinki Papers in Archaeology no. 9, 9 40. Pesonen, Petro 1996c. Posion Kuorikkikankaan asumus. Muinaistutkija 1/1996, 19 25. Pesonen, Petro 1996d. Takymetri arkeologisella kaivauksella. Muinaistutkija 2/1996, 31 33. Pesonen, Petro 1997. Suomen esihistoriallinen keramiikka. http://www.helsinki.fi/hum/arla/keram. Updated 27.4.1997. Pesonen, Petro 1998. Vihi Kampakeraaminen asuinpaikka Rääkkylässä. Muinaistutkija 1/1998, 23 30. Pesonen, Petro 1999a. Radiocarbon Dating of Birch Bark Pitches in Typical Comb Ware in Finland. In Huurre, Matti (ed.): Dig it all. Papers dedicated to Ari Siiriäinen. Helsinki, 191 200. Pesonen, Petro 1999b. Rekikylä kivikautinen kylä Ylikiimingissä. Muinaistutkija 1/1999, 2 15. Pesonen, Petro 1999c. Suomen esihistoriallinen keramiikka. http://www.helsinki.fi/hum/arla/keram. Updated 10.3.1999. Pesonen, Petro 1999d. Takymetri löytödokumentoinnissa ja löytöjen levinnän kuvaaminen. Historiallisen ajan arkeologian menetelmät. Seminaari 1998. Museoviraston rakennushistorian osaston julkaisuja 20, 39 42. Pesonen, Petro 2001 Kiteen Sarvisuo lisää varhaisesta asbestikeramiikasta. Kentältä poimittua 5. Kirjoitelmia arkeologian alalta. Museoviraston arkeologian osaston julkaisuja N:o 9, 34 56. 250

Purhonen, Paula & Ranta, Helena (eds.) 1991. Arkeologia Suomessa Arkeologi i Finland 1986 1987. Purhonen, Paula & Ranta, Helena (eds.) 1994. Arkeologia Suomessa Arkeologi i Finland 1988 1989. Pälsi, Sakari 1913. Palomäen kivikautinen asuinpaikka Urjalassa. Suomen Museo 1913, 63 79. Pälsi, Sakari 1918. Kaivaus Pitkäjärven kivikautisella asuinpaikalla Räisälässä v. 1915. Suomen Museo 1918, 25 34. Pälsi, Sakari 1920. Riukjärven ja Piiskunsalmen kivikautiset asuinpaikat Kaukolassa. Suomen Muinaismuistoyhdistyksen Aikakauskirja XXVIII, 1 192. Pälsi, Sakari 1939. Esihistorian tutkimuskentiltä. Jyväskylä. Ramqvist, Per H. 1985. Gene. On the origin, function and development of sedentary Iron Age settlement in Northern Sweden. Archaeology and environment 1. Ramqvist, Per H. 1997. Inte bara väggar. Analys av bränd lera från järnåldern. Umeå. Rankama, Tuija 1994. Muutamia ajatuksia kiviteknologiaan liittyvästä terminologiasta. Muinaistutkija 4/ 1994, 7 13. Rankama, Tuija 1996. Prehistoric riverine adaptationsin subarctic Finnish Lapland: the Teno river drainage. Dissertation Submitted in Partial Fulfilment of the Requirements for the Degree of Doctor of Philosophy in the Department of Anthropology at Brown University. Rankama, Tuija 1997a. Ala-Jalve. Spatial, technological, and behavioral analyses of the lithic assemblage from a Stone Age Early Metal Age site in Utsjoki, Finnish Lapland. BAR International Series 681. Rankama, Tuija 1997b. Tenojoen vesistöalueen ympäristön ja asutuksen vaiheet lyhennelmä väitöskirjasta. Muinaistutkija 1/1997, 1 13. Rauhala, Pirjo 1977. Liljendalin Kvarnbacken. Helsingin yliopiston arkeologian laitos. Moniste n:o 13. Reisborg, Synnöve 1994. En husgrund från Nibble, Ösmo socken, Södermanland. Tor 26, 57 66. Rikkinen, Kalevi 1994. Suomen aluemaantiede. 2. uudistettu painos. Helsingin yliopisto. Lahden tutkimusja koulutuskeskus. Risla, Pentti 2001. Jättekyrkor i Österbotten. Bottnisk Kontakt X, 163 169. Rothschild, Nan A., Mills, Barbara J., Ferguson, T.J. & Dublin, Susan 1993. Abandonment at Zuni farming villages. In Cameron, Catherine M. & Tomka, Steven A. (eds.): Abandonment of settlements and regions. Ethnoarchaeological and archaeological approaches. Cambridge University Press, Cambridge, 123 137. Räihälä, Oili 1996. A Comb Ware house in Outokumpu Sätös Some remarks on the application of ceramic typologies. Pithouses and Potmakers in Eastern Finland. Reports of the Ancient Lake Saimaa Project. Helsinki Papers in Archaeology no. 9, 89 118. Räihälä, Oili 1997. Kuoppatalon merkitys, Muinaistutkija 4/1997, 37 44. Räihälä, Oili 1998. Suomussalmen Salonsaari kivikautinen leiripaikka Kiantajärven rannalla. Kentältä poimittua 4. Kirjoitelmia arkeologian alalta. Museoviraston arkeologian osaston julkaisuja N:o 7, 5 23. Räihälä, Oili 1999. Tutkimuksia Suomussalmen kivikautisesta asutuksesta kvartsien fraktuurianalyysin avulla. Studia Historica Septentrionalia 35. Rajamailla V, 117 136. Salo, Unto 1962. Pronssikauden tutkimuksia Nakkilassa. Suomen Museo 1962, 27 73. Salo, Unto 1970. Metallikautinen asutus Kokemäenjoen suussa I. Muinaisjäännökset ja muinaislöydöt. Pori. Salo, Unto 1976. Bronsåldershus i Satakunta. Iskos 1, 51 54. Salo, Unto 1981. Satakunnan pronssikausi. Satakunnan historia 1,2. Satakunnan maakuntaliitto ry. Rauma. Salo, Unto 1984. Pronssikausi ja rautakauden alku. Suomen historia 1. Espoo, 98 249. Salo, Unto 1997. Ihmisen jäljet Satakunnan maisemassa. Kulttuurimaiseman vuosituhannet. Suomalaisen Kirjallisuuden Seuran Toimituksia 677. Rauma. Salo, Unto 1999. Kotimaakuntamme Satakunta. Katsaus Satakunnan asuttamisen, organisoitumisen, talouden ja kulttuurin vaiheisiin. Pori. Sarkkinen, Mika & Mäkivuoti, Markku 2000. Inventointia Pohjois-Pohjanmaalla. In Kirkinen, Tuija & Maaranen, Päivi (eds.): Arkeologinen inventointi. Opas inventoinnin suunnitteluun ja toteuttamiseen. Helsinki, 140 152. Sartes, Minna 1994. Subneolithic and Neolithic settlement systems in South-West Finland. Discussion on resource areas. Fenno-Ugri et Slavi 1992. Prehistoric economy and means of livelihood. Julkaisu/Museovirasto, arkeologian osasto, N:o 5, 105 114. 251

Sarvas, Anja 1973. Tervolan Törmävaara kaivaustuloksia vuodelta 1972. Totto XI, 14 18. Saukkonen, Jyri 1994. Annual cycles, base camps and means of livelihood Stone-Age hunter-gatherer settlement in the Närvijoki river area of southern Ostrobothnia. Fenno-Ugri et Slavi 1992. Prehistoric economy and means of livelihood. Julkaisu/Museovirasto, arkeologian osasto, N:o 5, 115 123. Schanche, Kjersti 1989. Nye funn fra yngre steinalder i Varanger. Viking 1989, 53 71. Schanche, Kjersti 1995. Var det store landsbyboplasser i Finnmark i yngre steinalder? Debattinnlegg med utgangspunkt i bosetningsspor fra ca. 2000 f.kr. Arkeologiske Skrifter fra Arkeologisk institutt i Bergen Museum. Universitetet i Bergen No. 8, 174 186. Schiffer, Michael B. 1976 Behavioral Archaeology. Academic Press, New York. Schiffer, Michael B. 1996. Formation Processes of the Archaeological Record. University of Utah Press. Salt Lake City. Schulz, Hans-Peter 1996. Pioneerit pohjoisessa. Suomen varhaismesoliittinen asutus arkeologisen aineiston valossa. Suomen Museo 1996, 103, 5 45. Schulz, Hans-Peter 1990. On the Mesolithic Quartz Industry in Finland. Iskos 9, 7 23. Schönbäck, Bengt 1952. Bronsåldershus i Uppland. Tor 1949 51, 23 45. Seger, Tapio 1986a. Orrmoan: A Pre-Roman Dwelling site in Korsnäs, S. Ostrobothnia, Finland. Finskt Museum 1986, 22 32. Seger, Tapio 1986b. Trofastbacken: Excavation of a Pre- Roman House in Korsnäs, S. Ostrobothnia, Finland. Iskos 6, 175 184. Seger, Tapio 1987. Refuse faunas from the Pre-Roman sites of Trofastbacken and Orrmoan in Korsnäs, S. Ostrobothnia, Finland. Suomen Museo 1987, 40 44. Seger, Tapio & Nieminen, Eeva-Liisa 1983. Kaivausdokumentoinnista ja löytöjen levinnän graafisista esityksistä. Karhunhammas 7, 142 155. Seppälä, Sirkka-Liisa & Haimila, Miikka 1998. Tavara tarkasti talteen löytöjen numeerisesta paikannuksesta ja käsittelystä esimerkkinä polttokenttäkalmisto. Kentältä poimittua 2. Kirjoitelmia arkeologian alalta. Museoviraston arkeologian osaston julkaisuja N:o 7, 104 113. Sepänmaa, Timo & Bilund, Antti 1993. Rantasalmen muinaisjäännösinventointi vuonna 1993 yhteenvetoa 252 tuloksista. Sihti 3, Savonlinnan maakuntamuseon julkaisuja, 22 28. Siiriäinen, Ari 1964. Kemijärven Jatulinsaari. Suomen Museo 1964, 26 39. Siiriäinen, Ari 1967. Yli-Iin Kierikki. Asbestikeraaminen asuinpaikka Pohjois-Pohjanmaalla. Suomen Museo 1967, 5 37. Siiriäinen, Ari 1974. Tervolan Törmävaara. Kemijoki 8000 näyttelyopas. Kemijoen alueen muinaisuutta 8000 vuoden ajalta arkeologisten tutkimusten valossa. Helsinki, 42 43. Siiriäinen, Ari 1981a. On the Cultural Ecology of the Finnish Stone Age. Suomen Museo 1980, 5 40. Siiriäinen, Ari 1981b. Problems of the East Fennoscandian Mesolithic. Finskt Museum 1977, 5 31. Siiriäinen, Ari 1983. Humppilan Järvensuon kivikautinen löytöpaikka. Karhunhammas 7, 79 87. Siiriäinen, Ari 1986. Kärräniemi in Rovaniemi. A middle subneolithic site with a palisade in northern Finland. Iskos 6, 185 198. Siiriäinen, Ari 1987. On archaeology and land uplift in Finland. Geological Survey of Finland. Special Paper 2, 43 45. Simonsen, Povl 1976. Stenalderens hustyper i Nord- Norge. Iskos 1, 23 25. Simonsen, Povl 1979. Veidemenn på Nordkalotten, hefte 3. Stencilseries B. ISV. University of Tromsø. Simonsen, Povl 1986. Fortsatte undersøkelser ved Virdnejavri, Kautokeino k., Finnmark. Arkeologisk feltarbeid i Nord-Norge 1985. Tromura, kulturhistorie 6, 1 11. Simonsen, Povl 1996. Steinaldersbosetningen i Sandbukt på Sørøya, Vest-Finnmark. Tromsø Museums Skrifter XXVII. Sirelius, U.T. 1921. Suomen kansanomaista kulttuuria II. Helsinki. Sohlström, Beatrice 1992. En stenåldershydda en bosättningsanalys. Kentältä poimittua. Kirjoitelmia arkeologian alalta. Museoviraston esihistorian toimiston julkaisuja N:o 2, 27 38. Soininen, Tuija-Liisa 2000. Arkeologisen toiminnan kehitys Pirkanmaalla 1991 1999. Arkeologia Suomessa 1997 1998, 22 40. Spång, Lars Göran 1986. Stenåldersbostaden. Studier i Norrländsk forntid II. Acta Bothniensia Occidentalis. Umeå.

Strandberg, Nina 1996. Kaarina Hulkkio varhaismetallikauden ja vanhemman rautakauden kohtaamispaikka. Kentältä poimittua 3. Kirjoitelmia arkeologian alalta. Museoviraston arkeologian osaston julkaisuja N:o 6, 37 45. Strandberg, Nina 1998. Rakennukset Kaarinan Hulkkion varhaismetallikautisella ja rautakautisella asuinpaikalla. Muinaistutkija 1/1998, 2 12. Stuiver, Minze & Kra, Renee S. (eds.) 1986. Radiocarbon 28 (2B), 805 1030. Stuiver, Minze & Reimer, Paula J. 1993. Extended 14 C Data Base and Revised CALIB 3.0 14 C Age Calibration Program. Radiocarbon 35 (1), 215 230. Suominen, Esa 1998. Kainuun museon arkeologinen toiminta (1979 1996). Varhain Pohjoisessa. Johdanto. Varhain Pohjoisessa -hankkeen artikkeleita. Helsinki Papers in Archaeology no. 11, 25 26. Sussman, Carole 1988. A Microscopic Analysis of Use- Wear and Polish Formation on Experimental Quartz Tools. BAR International Series 395. Oxford. Taavitsainen, J.-P. 1982. Tikku-ukko Rääkkylästä. Fennoscandia Antiqua I. 13 16. Taavitsainen, J.-P., Simola, Heikki & Grönlund, Elisabeth 1998. Cultivation History Beyond the Periphery: Early Agriculture in the North European Boreal Forest. Journal of World Prehistory 12:2, 199 253. Talve, Ilmar 1990. Suomen kansankulttuuri. Suomalaisen Kirjallisuuden Seuran Toimituksia 514. Tanner, Väinö 1929. Om Petsamo-kustlapparnas sägner om forntida underjordiska boningar, s.k. jennam vuölas kuatt. Finskt Museum 1928, 1 24. Taskinen, Juha 1995. Laatokan poikia viiksekkäitä. Helsingin Sanomat. Kuukausiliite n:o 3. 4. helmikuuta 1995, 30 35. Tesch, Sten 1993. Houses, Farmsteads, and Long-term Change. A Regional Study of Prehistoric Settlements in the Köpinge Area, in Scania, Southern Sweden. Lund. Thalbitzer, William 1914. The Anmassalik Eskimo: contributions to the ethnology of the East Greenland natives. Meddelser om Grønland 39 40. Vol. 39. Torvinen, Markku 2000a. Kokemuksia kivi- ja varhaismetallikautisten kohteiden inventoinnista. In Kirkinen, Tuija & Maaranen, Päivi (eds.): Arkeologinen inventointi. Opas inventoinnin suunnitteluun ja toteuttamiseen. Helsinki, 133 139. Torvinen, Markku 2000b. Säräisniemi 1 Ware. Fennoscandia archaeologica XVII, 3 35. Tuovinen, Tapani 1990. Fornfynden och deras uttolkning. In Zilliacus, Kurt (ed.): Finska skären. Studier i åboländsk kulturhistoria utgivna av Konstsamfundet till dess 50-årsjubileum 1990. Helsingfors, 15 113. Uino, Pirjo 1986. An Iron Age Community at Ketohaka in Salo and Other Remains of Metal Period Buildings in Finland. Iron Age Studies in Salo I II. Suomen Muinaismuistoyhdistyksen Aikakauskirja 89:1, 25 201. Ukkonen, Pirkko 1996. Osteological analysis of the refuse fauna in the Lake Saimaa area. Environmental Studies in Eastern Finland. Reports of the Ancient Lake Saimaa Project. Helsinki Papers in Archaeology no. 8, 63 91. Vaara, Rauno 2000. Yli-Iin Kierikin kivikautinen kylä asumuskonstruktiot 1998 1999. Muinaistutkija 2/2000, 2 12. Vankina, L. V. 1970. Vankina, L. V. 1970. Torfönikovaö stoönka Sarnate. Muzej istorii latvijskoj SSR. Riga, izdatel<stvo Zinatne. Sārnates purva apmetne. Latvijas PSR vēstures Muzejes. Rigā, izdavnieciba Zinātne Vasks, A. 1994. Agrā dzelzs laikmeta apmetne Kerkūzos. Arheoločija un etnogrāfija XVI, 46 64. Vasks, Andrejs 1995. New data on Early Iron Age settlement in south-eastern Latvia. In Kazakevi ius, Vytautas & Sidrys, Raymond (eds.): Archaeologia Baltica. Vilnius, 57 80. Viitanen, Eeva-Maria 1996. Palanutta savea seinistä, katosta ja vähän muualtakin Hämeenlinnan Varikonniemen asuinpaikan palaneen saven analyysi. Kentältä poimittua 3. Kirjoitelmia arkeologian alalta. Museoviraston arkeologian osaston julkaisuja N:o 6, 86 98. Vikkula, Anne 1987. The Stone Age graves of the Nästinristi site in Laitila, SW Finland. Suomen Museo 1986, 5 18. Vikkula, Anne 1994a. Ecological approaches to the Stone Age of Ancient Lake Saimaa. Fenno-Ugri et Slavi 1992. Prehistoric economy and means of livelihood. Julkaisu/Museovirasto, arkeologian osasto, N:o 5, 167 179. Vikkula, Anne 1994b. Stone Age environment and landscape changes on the eastern Finnish lake district. Methods in the Mountains. Sidney University Archaeological Methods Series # 2, 91 97. Vilkuna, Janne 1996. Keski-Suomen muinaismuistojen inventointiprojekti. Arkeologia Suomessa 1993 1994, 33 38. 253

Vilkuna, Kustaa 1933. Kodasta pirttiin. Suomen kulttuurihistoria I, 158 169. Vuorinen, Juha-Matti 1997. Rakentamisesta Raision Mullissa rautakauden lopulla ja varhaiskeskiajalla. Muinaistutkija 4/1997, 45 48. Welinder, Stig 1983. Ecosystems change at the Neolithic transition. Norwegian Archaeological Review vol. 16 N:o 2, 99 105. Whittle, Alasdair 1996. Europe in the Neolithic. The creation of new worlds. Cambridge World Archaeology. Cambridge University Press. Wilhelms, Saara 1995. A natural-geographical study of prehistoric sites. Fennoscandia archaeologica XII, 195 205. Wyszomirski, Bozena 1975. Människofigurer på den skånska gropkeramiken. Fornvännen 1975/3 4, 129 137. Ylimaunu, Juha, Ylimaunu, Timo & Okkonen, Jari 1999. Hylkeenpyynnin kehityksestä ja merkityksestä Itämerellä esihistoriallisella ajalla. Faravid 22 23, 131 158. Ylimaunu, T. 1998. Kvartsiesineet ja -iskokset. Meteli. Oulun yliopiston arkeologian tutkimusraportti 15, 23 28. Zvelebil, Marek 1981. From Forager to Farmer in the Boreal Zone. Reconstructing economic patterns through catchment analysis in prehistoric Finland. BAR International Series 115. 254

Paper II Mökkönen, T., Nordqvist, K. & Belskij, S. 2007. The Rupunkangas 1A site in the archipelago of ancient Lake Ladoga: a housepit with several rebuilding phases. Fennoscandia archaeologica XXIV, 3 28. Publisher: The Archaeological Society of Finland, Helsinki ISSN 0781-7126

Fennoscandia archaeologica XXIV (2007) Teemu Mökkönen, Kerkko Nordqvist & Stanislav Bel skij THE RUPUNKANGAS 1A SITE IN THE ARCHIPELAGO OF ANCIENT LAKE LADOGA: A HOUSEPIT WITH SEVERAL REBUILDING PHASES Abstract A small-scale excavation at the Rupunkangas 1A site in 2005 yielded artefacts dating from the Mesolithic Stone Age to the Early Metal Period. In this article, all sites in the Rupunkangas area are discussed briefly. However, the main focus is on the interpretation of the Rupunkangas 1A site a housepit with thick blackish cultural layers. Even though the stratigraphy did not yield any clear surfaces that could be interpreted as separate floor layers, demolition layers, or layers with collapsed earth walls or roof, the stratigraphy and finds together with the radiocarbon dates indicate several rebuilding phases of pithouses in the excavated housepit. The maximum thickness of the cultural layers within the partly excavated housepit was ca. 1.5 meters. The oldest radiocarbon date obtained from the site was 8770 ± 85 BP, rendering the oldest occupation at the site approximately synchronous with the isolation of Lake Ladoga from the Baltic Sea Basin. This is also the oldest radiocarbon date obtained from any housepit in the region. The youngest finds date most probably to the early Early Metal Period, prior to the formation of the River Neva ca. 1300 cal. BC. In this article, it is argued that the Rupunkangas 1A site should be interpreted as a seasonally occupied hunting camp as opposed to the typical interpretation of sites with housepits as permanent residential camps in a (semi-)sedentary settlement system. Keywords: Housepit, Stone Age, Mesolithic, Neolithic, Early Metal Period, Karelian Isthmus, Lake Ladoga, sedentariness Teemu Mökkönen & Kerkko Nordqvist, Institute for Cultural Research, Archaeology, P.O.Box 59, FIN- 00014 University of Helsinki, Finland, teemu.mokkonen@helsinki.fi, kerkko.nordqvist@helsinki.fi; Stanislav Bel'skij, Peter the Great Museum of Anthropology and Ethnography (Kunstkamera), Universitetskaja nab. 3, RU-199034, St. Petersburg, Russia, stanislav.belskiy@kunstkamera.ru INTRODUCTION The former Finnish parish of Kaukola (now Russian Sevast janovo) is located in the valley of the Vuoksi River on the north-western shore of Lake Ladoga, on the Karelian Isthmus, Russia (Fig. 1). About 5 km north-east of the municipal centre of Kaukola and ca. 12 km north-west of the city of Käkisalmi (Ru. Priozersk) lies a pine barren area called Rupunkangas, which used to be a large island in the archipelago of Ancient Lake Ladoga. The large dwelling site areas around Piiskunsalmi and Riukjärvi, most of which were excavated in the beginning of the 20 th century (see Ailio 1909; Pälsi 1915; Uino 1997: 32 33; 2003; Huurre 2003: 202 203), are located some ten kilometres north-west of the Rupunkangas area. The nature of the sites on the former island of Rupunkangas is diverse. There are pit structures of various shapes and sizes. Many sites consist of several find areas, some of which are located at different elevations and on separate shore terraces. However, the most interesting results derived from the Rupunkangas 1A site (known in Russian as Protochnoe IV), where a housepit 1 that was only 3

Fig. 1. The hunter-gatherer sites in the Vuoksi River Valley and the location of the Rupunkangas area. In the larger map the former Rupunkangas island is marked with a square. The extent of Ancient Lake Ladoga in the Vuoksi River Valley is illustrated in the larger map with the help of the 20 m a.s.l. contour line. The sites and areas mentioned in the text: 1) sites around the Piiskunsalmi area (Kaukola), 2) sites around the Riukjärvi area (Kaukola), and 3) the Pitkäjärvi site (Räisälä). Maps: T. Mökkönen & K. Nordqvist. 4

partly excavated proved to have been reoccupied several times during the Early Mesolithic to the Early Metal Period. A housepit with several rebuilding phases is a rarity in this region. However, the most remarkable characteristic of the housepit at the Rupunkangas 1A site is the total duration of the utilization period. The rebuilding phases of individual housepits usually fall within a few hundred years (e.g., Renouf & Murray 1999). In North America there are records of housepits with several occupation phases, each lasting some hundreds of years. In these cases, the intervening unoccupied periods separating the building phases have lasted longer than the occupation phases, and the total utilization period of the housepit may have lasted up to some 3000 years (see Hayden 2000b; Johnson & Wilmerding 2001: 141; Prentiss et al. 2003). At Rupunkangas 1A the span between the oldest and the youngest rebuilding phases, as indicated by radiocarbon dates and ceramics found at the site, is strikingly long, ca. 6500 calendar years. A short article dealing with this site will be published in Russian (Mökkönen et al., in press). The interpretation of the site presented here is not fully congruent with the previously written article. The sites in the Rupunkangas area are also briefly discussed in another article presenting the results of the 2004 survey (Mökkönen et al. 2007). THE RUPUNKANGAS AREA AN ISLAND IN ANCIENT LAKE LADOGA Rupunkangas is a moraine esker that formed a large island before the rapid regression of the water level after the formation of the River Neva ca. 1300 cal. BC 2 (Saarnisto & Grönlund 1996). Prior to the formation of the Neva, the water level of Ancient Lake Ladoga was at a higher elevation than today. During the Ancient Lake Ladoga Phase covering the period from the isolation of Lake Ladoga from the Baltic Sea Basin some 9800 9700 years ago (ca. 7800 7700 cal. BC) (Saarnisto 2003: 64) to the formation of the Neva (ca. 1300 cal. BC), the water level in the Rupunkangas area varied between ca. 20.5 and 21.7 m a.s.l. (Saarnisto & Siiriäinen 1970). The present water level of the lakes near the Rupunkangas area varies between ca. 7.5 and 8 meters above sea level. Fig. 2. Dwelling sites and changes in the water level in the Rupunkangas area. The shore levels are based on Saarnisto and Siiriäinen (1970). Contours on the ancient Rupunkangas island are at 2.5 meter intervals. The numbered sites on the map: 1) Pontuksenhauta 1, 2) Pontuksenhauta 2, 3) Pontuksenhauta 3, 4) Rupunkangas 4, 5) Rupunkangas 2, 6) Rupunkangas 1, and 7) Rupunkangas 3. Map: T. Mökkönen. Since the Rupunkangas area is located approximately on the same land uplift isobase as the outlet of Lake Ladoga before the birth of the River Neva, it is presumed that the water level in the area remained roughly constant during the whole period from the isolation to the formation of the Neva outlet (Fig. 2). Some changes in the water level have undoubtedly occurred, but these changes have been minor, only slightly more than 1 meter at the maximum. Following the isolation of Lake Ladoga, the water level can be tracked through archaeological sites. In the Rupunkangas area the Rupunkangas 3 site is very suitable for this purpose. It 5

Fig. 3. General map of the Rupunkangas 1 site. The excavated part of the find area Rupunkangas 1A is located on a strip of land between two sand pits. Map: T. Mökkönen. is located on sandy soil on top of a roughly 50 cm high shore terrace. The terrain at the site and behind it is totally flat. A radiocarbon date on burnt bone dates the site to 7880 7610 cal. BC (59.7% probability, Hela-1165, 8740 ± 80 BP) which is approximately synchronous with the isolation of Lake Ladoga. The Rupunkangas 3 site lies at an elevation of 22.0 to 23.0 m a.s.l. The base of the terrace at the site lies approximately at an elevation of 21.5 m a.s.l., which corresponds to the 6

water level at the time the site was settled or to the later modification of the terrace during the maximum transgression. According to observations on the soil profile of a small sand pit located on top of the terrace, no traces of transgression layers covering the cultural layer are present. Later, after the formation of the River Vuoksi (ca. 4000 cal. BC), which runs from Lake Saimaa in Finland to Lake Ladoga in Russia, the transgression accelerated also in the Rupunkangas area. As reconstructed, the water level preceding the formation of the River Vuoksi and the accelerated transgression lay at an elevation of 20.8 to 20.4 m a.s.l. in the Rupunkangas area 3 (according to Saarnisto & Siiriäinen 1970). Following the transgression, the maximum water level, that is, ca. 21.7 m a.s.l. in the Rupunkangas area, was reached just before the formation of the River Neva (Saarnisto & Siiriäinen 1970). After the formation of the Neva outlet the water level of Lake Ladoga dropped rapidly by some ten meters (Saarnisto & Grönlund 1996: 207), and the Rupunkangas area ceased to be an island (see Fig. 2). ARCHAEOLOGICAL SITES IN THE RUPUNKANGAS AREA All the sites in the Rupunkangas area have been discovered in surveys carried out by the University of Helsinki Department of Archaeology in cooperation with Russian archaeologists from the Institute for the History of Material Culture, Russian Academy of Sciences (St. Petersburg) and the Museum of Anthropology and Ethnography named after Peter the Great, Kunstkamera (St. Petersburg). The first sites in the Rupunkangas area were discovered in 1999 (Halinen et al. 1999; Lavento et al. 2001). The latest surveys and excavations have been carried out during the Kaukola-Räisälä Project (see Lavento et al. 2006; Mökkönen et al. 2007) organized and largely financed by the University of Helsinki in 2004 (Halinen & Mökkönen 2004) and 2005 (Mökkönen 2005a; 2005b; Mökkönen & Nordqvist 2005; Mökkönen et al. 2005). The University of Tartu in Estonia has also taken part in the project s fieldwork. In all, seven dwelling sites were found in the surveys in an area of some 400 x 800 meters (Fig. 2). The sites are concentrated in the northern part of the former island, where the soil is sandy till. All of the sites have been located right next to the water on the top of terraces or slightly further back. Some of the sites consist of several find areas. At two of the sites archaeological material is found on successive terraces at different elevations. However, the narrow terraces are located on a sheer slope, and therefore the difference in the elevation does not automatically mean that the materials on the separate terraces are of different ages. During 2005, small-scale excavations were carried out at two of the sites, namely Rupunkangas 1A and 4. The dating of the sites is based on radiocarbon and pottery typology. Actually, only two of the sites have been dated by radiocarbon. In addition to the Rupunkangas 1A site, the main subject of this article, there is also another Early Mesolithic radiocarbon date from the Rupunkangas 3 site. The dating of the other sites rests on the dating of the ceramics found at the sites in the surveys (Table 3 at the end of the paper). The dating of the sites as a whole covers approximately the span from ca. 8000 cal. BC to ca. 1300 cal. BC. According to the present archaeological data gathered from the Rupunkangas area, the only temporal gap in the material is in the Early Neolithic Stone Age, which is not represented in the radiocarbon dates or in the pottery. The excavated site Rupunkangas 1A will now be discussed in detail. The other sites with housepits in the Rupunkangas area are discussed at the end of this article. RUPUNKANGAS 1A A HOUSEPIT WITH SEVERAL REBUILDING PHASES Before the formation of the River Neva, the large island formed by the Rupunkangas area was located in the outer zone of the inner archipelago of Ancient Lake Ladoga. The Rupunkangas 1A site, located at the eastern extremity of the ancient island, lay on an exposed shore (see Figs. 1 2). The 2.5 to 4 kilometre-wide stretch of open water opposite the site was separated from the larger open water areas of Ancient Lake Ladoga by a group of rocky islands. The site was located on top of a sheer shore terrace facing the open water and devoid of a sheltered background. The excavated find area Rupunkangas 1A is situated on top of a high bank at an elevation of ca. 22.0 to 23.5 m a.s.l. The base of the bank lies at ca. 20 m a.s.l. 7

Fig. 4. Kerkko Nordqvist standing in the housepit at the Rupunkangas 1A site in May 2005. The overgrown early 20 th century sand pit is on the left. The excavation area was located next to the sand pit to the right of Mr. Nordqvist. Photo: T. Mökkönen. Fig. 5. Southern profile section of the excavation area at the Rupunkangas 1A site. The radiocarbon dates obtained from samples 1 to 4 are shown in the Fig. 7. Radiocarbon sample 5 (Hela-1183, 205 ± 40 BP) dates the traditional tar-burning pit. The Late Comb Ware sherds were found in layer No. 7. Map: T. Mökkönen & K. Nordqvist. 8

During the surveys (2004 2005) finds pointing to both the Mesolithic Stone Age and the Early Metal Period were found in the find area referred to as Rupunkangas 1A (Halinen & Mökkönen 2004; Mökkönen et al. 2005; see also Mökkönen et al. 2007). There are altogether five find areas around the sand pits, and most of the site was presumably destroyed by sand extraction. The excavation area in find area 1A was situated on a narrow strip of land between the two sand pits (Figs. 3 4). The overgrown sand pit on the southeastern side of the excavation area is of Finnish origin and is marked on a Finnish topographic map from 1939. The other sand pit on the north north-western side, opened up by the Russians, is still in more or less active use. The excavation area was modest. It originally measured 1 x 3 meters, but during the excavation the area was expanded on the lower elevations towards the sand pit and finally covered 2 x 3 meters. During the surveys, a depression located on the narrow spit between the sand pits was perceived to resemble a housepit. Due to its location between two sand pits and the coarse till exposed on the surface in some areas inside the pit, however, it was not originally interpreted as a housepit but as a by-product of sand extraction activities. Hence, the true nature of the depression as a housepit became evident only in the course of the excavation. Stratigraphy and finds A two-week long small-scale excavation revealed multiple cultural layers with a total depth of 1.5 meters (Fig. 5). The dating of the layers, as presented here, is based on stratigraphic observations, ceramic typology, and radiocarbon dates. In the beginning of the excavation, the till area exposed on the surface at the bottom of the larger pit indeed turned out to be a by-product of the sand extraction activities. However, it was not formed as was first thought during sand quarrying activities carried out on the brink of the sand pits. On the contrary, the till proved to be fill in an older pit and was most probably deposited in connection with sand extraction. This till-filled pit structure was a simple tar-burning pit, dated by radiocarbon most probably to the late 17 th or 18 th century cal. AD (Hela-1183, 205 ± 40 BP). The tar-burning pit had gouged out parts of the older cultural layers. The finds from the stratigraphically uppermost yellowish brown sandy cultural layer (ca. 22.4 22.7 m a.s.l.) included some quartz artefacts (flakes and flake fragments), a large number of small, burnt fish bone fragments, and a small amount of pottery dating most probably to the early part of the Early Metal Period. Some of the pottery sherds were decorated with two parallel rows of small roundish pits (Fig. 6). One sherd shows that the vessel has been profiled and the rows of pits were placed approximately at a 45 degree angle in relation to the turn in the vessel s neck. In the 2005 survey one body sherd of pottery with very coarse net or textile impression was found (Fig. 6, artefact 3). This is the only textile impressed piece from the site. A number of other sherds have a striated surface. All of the pieces contain sand and organic temper. In the striated pieces the amount of hair temper is notable. The thickness of the sherds varies between 9.5 and 13 mm. All of the pottery sherds from the uppermost layers probably derive from a single vessel. The pottery could best be classified as Textile Ceramics. There are also some porous hair-tempered undecorated pieces that could be categorized on the basis of the fabric as Late Neolithic or Early Metal Period ceramics. Considering the environment of the site, this occupation phase dates most likely to the period preceding the formation of the River Neva (ca. 1300 cal. BC). The appearance of Textile Ceramics in archaeological material in Karelia around 1900/1800 cal. BC (Carpelan 1999: 268; Carpelan & Parpola 2001: 87; Lavento 2001: 106; Zhul nikov 2005: 114) sets the oldest time limit for the Early Metal Period settlement phase. Under the uppermost layers containing Textile Ceramics there was a weakly coloured cultural layer with only a few finds, namely quartz flakes, flake fragments, and a few large sherds of a single Late Comb Ware vessel (ca. 22.15 22.50 m a.s.l.). Below this, the soil changed into coarser sandy till that contained more charcoal than the upper layers. The lowest layers from ca. 21.3 to 22.9 m a.s.l. were Mesolithic. The upper parts of these layers were partly mixed with younger Neolithic layers, but most of the lowest layers (thickness ca. 50 70 cm) were purely Mesolithic. The Mesolithic layers were very coarse sandy till heavily mixed with ash and charcoal and containing lots of small stones. Due to the large amount of charcoal, the colour of these cultural 9

Fig. 6. Finds from the Rupunkangas 1A site. Ceramics 1-4: 1-2) Hair tempered Early Metal Period ceramics, 3) Textile Ceramics, and 4) Late Comb Ware. Lithic artefacts made of quartz from Mesolithic layers: 5-6) and 8-11) blades and blade fragments, 7) multi-platform core, 12) scraper, 13-14) platform-on-anvil cores. All lithics were found during the 2005 excavation. The length of the scale bars is 3 cm. Photos: T. Mökkönen, Drawings: O. Seitsonen, Layout: T. Mökkönen & K. Nordqvist. 10

layers varied from dark brown to black. These layers did not contain any precisely datable artefacts the Mesolithic date is based on radiocarbon dates. The lithic material analysed by Oula Seitsonen is mainly composed of quartz, but some lithic artefacts made of quartzite and rock crystal are also included (see Mökkönen 2005a). The largest fraction of the material consists of flakes and flake fragments (Table 1). The material included some blades (according to form and dimensions), cores (bipolar, hammer-on-anvil, and platform cores) and one scraper. Clearly identifiable changes in the find profile within the Mesolithic layers could not be discerned. In total, the number of finds is low. Osteological Material Table 1. Lithic artefacts found during the excavation of the Rupunkangas 1A site. All identified bones in the osteological material recovered in the excavation were fish (Mökkönen 2005a). The excavation yielded only 380 bone fragments of which 115 were identifiable (Table 2). The osteological analysis was carried out by Sanna Seitsonen. Most of the bones identified as to species, genus or family were found in the uppermost layers together with the Textile Ceramics and Late Comb Ware. The poor preservation of burnt bone in coarse sandy till is demonstrated by the fact that only nine bone fragments were found in the Mesolithic layers. In addition to bone fragments found in the excavation, a fragment of elk/reindeer bone and some fish bones were found in the surveys (Puttonen 2004; 2005). The excavated osteological material from the site consists exclusively of fish. Most of the fish species in the assemblage, such as perch, are relatively easy to catch year round, even though their spawning season is in the spring and summer. However, salmonid fishes, represented in the assemblage by whitefish, are easiest to catch during their spawning season in the autumn and early winter. The osteological material gathered from the nearby Mesolithic site Rupunkangas 3 includes, in addition to salmonid fishes, also three seal bone fragments (see the analysis by Sanna Seitsonen in Mökkönen & Nordqvist 2005), which suggests that the site was inhabited during the cold season. The best seal hunting season on Lake Ladoga is in the late winter when the seal pups are easiest to catch on the ice at the margin of the open water. However, seal hunting during the open water season is also known from ethnological sources (Lehtonen 1974; see also Ylimaunu 2000: 169 83). The osteological materials from the sites Rupunkangas 1A and Rupunkangas 3 are in harmony with the environmental location of the sites in the archipelago. The only non-aquatic species in the material from these sites are dog and elk/ reindeer, both represented in the material by one bone fragment. In the osteological material there are implications pointing to both cold and warmer seasons. However, this material is too small to be used as Table 2. Bone fragments identified as to species, genus, or family from the Rupunkangas 1A site. 11

Fig. 7. Radiocarbon dates from the Mesolithic layers. The stratigraphic location of the samples is shown in Fig. 5. a basis for defining the period of occupation of the site. In addition, the occupation period could have varied at different times. Mesolithic dates In all, dates were obtained from five radiocarbon samples from different stratigraphic units at Rupunkangas 1A (Figs. 5 & 7). There was not enough datable material in the uppermost layers, the tar-burning pit being an exception. All dated samples were charcoal. Four of the samples were dated to the Mesolithic Stone Age. The partly mixed nature of the cultural layers is evident in the case of sample 3 (Hela-1196, 7550 ± 75 BP) taken from the bottom of a layer in which the Late Comb Ware sherds were found. The youngest Mesolithic date was produced by sample 4, taken from a stone-lined pit structure that was dug into older layers. The most surprising feature of the dating results is the temporal distribution of the radiocarbon dates. These cover nearly the whole Mesolithic period. On the basis of the radiocarbon dates, the relative dating of the ceramics, and the dating of the formation of the River Neva, the dwelling site has been in episodic use during a time span extending from ca. 8000 to ca. 1300 cal. BC and covering the whole period of the Ancient Lake Ladoga Phase. For a long time, the record for the oldest radiocarbon dates from the Karelian Isthmus was held by a fishing net found at the Korpilahti site in the former parish of Antrea (Vuoksenranta) (Pälsi 1920). Two floats made of pine bark have been dated to the late Pre-Boreal period (Hel-269, 9230 ± 210 BP and Hel-1303, 9310 ± 140 BP, e.g., Siiriäinen 1974: 11; Matiskainen 1989: 71) and an AMS analysis of the net cord has yielded an only slightly younger date (Hela-404, 9140 ± 135 BP, Carpelan, in press; Miettinen et al., in press; Takala 2004a: 151 with references). Lately, the possibility of dating burnt bone material has increased the number of old sites dated by the radiocarbon method. At the moment, the oldest radiocarbon-dated dwelling site on the Karelian Isthmus is the Suuri Kelpojärvi site in the former parish of Antrea (Hela-931, 9275 ± 120 BP, Takala 2004a: 152), which is approximately synchronous with the Antrea net. The oldest date from Rupunkangas 1 falls in the same period as the next-oldest sites from the Karelian Isthmus. Two sites in the former parish of Heinijoki near the Vetokallio outlet of Ancient Lake Ladoga, namely Valkialampi 1 (Hela-743, 8765 ± 65 BP, burnt bone) and Valkialampi 2 (Hela-744, 8720 ± 70 BP, burnt bone) are also dated to the very beginning of the Boreal period (Takala 2004a: 154, 161, Fig. 159). The oldest date from the Rupunkangas 1 site (Hela-1182, 8770 ± 85, charcoal) and the date from Rupunkangas 3 (Hela-1165, 8740 ± 75 BP, burnt bone) are approximately synchronous with the Valkialampi sites. In addition, there is one unpublished new dating from the Juhola 2 site in the former parish of Kirvu, which is about 200 radio- 12

Fig. 8. Cut 2, made in the profile of the sandpit next to the excavation area. The cultural layers have clearly formed in a man-made depression that cuts into the natural layers. The location of Cut 2 is illustrated in Figs. 3 and 9. Map: T. Mökkönen & K. Nordqvist. carbon years older than the oldest sites in Rupunkangas and Valkilampi areas (Petri Halinen, pers. comm.). Even though the oldest dates from the Rupunkangas area are among the six oldest radiocarbon dates obtained from dwelling sites on the Karelian Isthmus, these dates are still young when comparing to the oldest dates from north Karelia, south-eastern Finland, and southern Finland. There, the oldest known dwelling sites are dated older than 9000 BP, that is ca. 8250 cal. BC. Compilations of the oldest dates from Finland and the Karelian Isthmus can be found in Takala (2004a: 161, Fig. 159) and Pesonen (2005; see also Matiskainen 1996; Schulz 1996; Forsberg 2006: 13). PREHISTORIC STRUCTURES AT THE RUPUNKANGAS 1A SITE Due to the small size of the excavation area, only a few structures were discovered. As mentioned above, the uppermost feature of the excavation area was a simple tar-burning pit filled with till, dating to the late 17 th or 18 th century AD. The following discussion will concentrate on the rebuilding phases of the housepit and the structures inside the pit. The housepit An interesting structure and a key for understanding the stratigraphy of the site was revealed when the edge of the sand pit next to the excavation area was sectioned for the purpose of recording the layers (hence referred to as Cut 2; see Figs. 3, 8 & 9). Due to the very real threat of a landslide, the section was not cut until the recording of the sections in the excavation area was finished. Cut 2 shows that the Mesolithic layers clearly intrude into the natural layers (Fig. 8). This means that the Mesolithic cultural layers were formed in a manmade depression. The depression, visible on the 13

Fig. 9. Surface elevation model of the Rupunkangas 1A site. Contours are given at 10 cm intervals (m a.s.l.). The shaded area shows the shape of the last housepit on the site. Map: T. Mökkönen & K. Nordqvist. surface of a narrow strip of land between the sand pits, was recognized already in the surveys, but at that time it was not considered to be a housepit for several reasons (see above). After the excavation and exposing of Cut 2, it was obvious that the excavated layers had actually formed inside a man-made depression interpreted as a housepit. Without Cut 2, this interpretation could have been possible, but it would have been much more poorly arguable than now. The information gained from Cut 2 gave a context for the layers found in the excavation area. At the bottom, the Mesolithic layers were clearly cutting into the natural layers, the Neolithic layers above were partly cutting into the Mesolithic layers below, and the top layers containing Textile Ceramics were partly cutting into both the Mesolithic and the Neolithic layers. Therefore, the dwelling depression was first dug during the early Mesolithic Stone Age and later reused several times in the course of the Mesolithic. The interpretation of the weak cultural layers associated with Late Comb Ware is problematic. The data gathered during the excavation do not allow us to determine whether these layers bear evidence of a Late Neolithic rebuilding phase or of some other Late Neolithic activity, such as trash dumping. Judging by the stratigraphy of the site, the latest rebuilding phase took place during the Early Metal Period. The layer containing Textile Ceramics in the excavation area continued into Cut 2, where it intersects the dark and coarse Mesolithic layers (Fig. 8). It is clear that the layer with Textile Ceramics was formed in a pit dug into the Mesolithic and Neolithic layers a pit that can be interpreted as a housepit. The shape and size of the final rebuilding phase of the pithouse at the site are definable. Based on the surface elevation model, there are two entrances or antechambers, one at each end of the depression (Fig. 9). The entrances are visible as smaller depressions (ca. 2 x 2 meters in size) connected to the larger depression in the middle (ca. 5 x 6 meters in size). The depth of the depression as measured from the surrounding terrain is ca. 35 cm. The surface elevation model illustrates the last 14

building of a pithouse in the housepit, which probably took place during the Early Metal Period. According to the elevation model, the excavation area was located by the entrance and near the corner between the entrance and the main room. The relationship between the excavation area and the older building phases of the pithouses is not known, but the Mesolithic structure may have been larger than the one visible in the surface elevation model (concerning Mesolithic housepits in the Karelian Republic, see Zhul nikov 2003). In North America, housepits with several rebuilding phases have been studied in several locations where organic materials have been preserved so well that it has been possible to distinguish separate floor levels and layers with collapsed roofing material (e.g., Hayden 1997: 36 43; Hayden 2000a; Johnson & Wilmerding 2001; Prentiss et al. 2003). This was not the case at Rupunkangas. The deepest thin and totally black charcoal layer was the only clearly definable interface, although it did not form a completely continuous layer covering the whole excavation area. The concentrations of firecracked and unburnt stones did not form any identifiable surfaces to be interpreted as separate floor layers, demolition layers, or layers with collapsed earth wall or roof material formed after the abandonment of the structure (see below). Stone structures inside the housepit Five identified stone structures could be discerned in the prehistoric layers. In the lowest part of the Neolithic layers a concentration of fire-cracked stones was interpreted at first as a fireplace (see Mökkönen et al., in press). On top of the Mesolithic layers there were two other obscure pile-like concentrations of fire-cracked rocks that could not reasonably be interpreted as fireplaces because no burnt sand was associated with the piles. The Mesolithic layers actually contained a large number of stones, both fire-cracked and unburnt, in more or less blurry concentrations. Nevertheless, it was not possible to distinguish any clear interfaces within the Mesolithic layers even with the help of the stones. Most probably these stone concentrations have not originally lain on the floor of the pithouse but have presumably collapsed into the pit from the bank or from the lowest part of the roof when the structures have deteriorated or caved in (see e.g., Hayden 1997: 37 8, 40 2, 63 4; Kankaanpää 2002: 71; Ojanlatva & Alakärppä 2002: 117 18). In addition to these more or less obscure stone structures there were two small pit features. One of these was a pit ca. 20 cm in diameter surrounded by small stones and filled with bright red burnt sand. This stone-lined feature was dug at an angle into the older Mesolithic layers. It can be interpreted as a post hole or as a pit used for keeping a fire. This feature is dated to the Late Mesolithic (see Figs. 5 & 7, the radiocarbon sample 4). Another pit feature was found in the lowest Mesolithic layer, where a pit dug ca. 20 cm deeper than the other Mesolithic layers did not contain any finds. DISCUSSION ON THE RUPUNKANGAS 1A SITE The former Rupunkangas island is located in an aquatic milieu. Looking at the overall distribution of Stone Age and the Early Metal Period dwelling sites in the valley of the Vuoksi River (Fig. 1; see also Lavento et al. 2001: Appendix 1; Lavento et al. 2006: Fig. 1; Mökkönen et al. 2007), it is clear that the Rupunkangas area, located in the archipelago of the Ancient Lake Ladoga, is not in the heartland of the hunter-gatherer settlement of the area. In the following, issues such as site function and the formation process of thick multi-period cultural layers will be discussed in the light of the Rupunkangas 1A site. Thick and heavily coloured cultural layers Thick cultural layers such as those of the Rupunkangas 1A site are not regularly found in the area. However, there are also a few other sites with abnormally thick cultural layers on the Karelian Isthmus. At the Telkkälä site in former Muolaa Parish, the overall thickness of the superimposed layers as measured from the top to the bottom of the lowest cultural layer was 2.5 meters (Takala & Sirviö 2003; Takala 2004b). In Telkkälä, the upper and the lower cultural layers were separated by 20 90 cm thick layers deposited by the transgression of Ancient Lake Ladoga. Another location where the layers reached a combined thickness of ca. 2.5 to nearly 3 meters was at the 15

Stone Age dwelling site Kunnianniemi, on the shore of Lake Kiimajärvi in the former parish of Pyhäjärvi (Seitsonen & Nordqvist 2006; Gerasimov et al. 2007). There, four cultural layers were separated by transgression layers. At both sites the thick cultural layers were observed in small-scale excavations, and therefore it has not been possible to connect the layers to any definable structures. The lowest, Mesolithic layers at the Rupunkangas 1A site were heavily coloured by charcoal and ash. A similar phenomenon has also been observed at the Telkkälä site, where the lowest layers of black-coloured sand dating to the Early Neolithic and the Late Mesolithic Stone Age contained an abundant amount of charcoal as compared to the stratigraphically superior, lightly coloured Neolithic layers (Takala & Sirviö 2003: 62, 68, Fig. 8). Similar observations have also been made at the Kunnianniemi site and at the Juoksemajärvi site in Räisälä Parish (Halinen et al., in press). In this light, there may be a tendency of the Mesolithic layers to be coloured blackish by charcoal and soot to a greater degree than the Neolithic layers, which indicates if some underlying geological process is not responsible changes in the activities carried out at the dwelling sites. The layers that have built up inside the housepit discovered at the Rupunkangas 1A site are unusual as regards the thickness of the layers and the amount of charcoal they contain. The radiocarbon dates indicate several occupation or rebuilding phases at the site, which naturally partly explains the thickness of the layers. A housepit with several rebuilding phases is a rarity in the general vicinity on the Karelian Isthmus, in Russian Karelia, and in Finland. On a global scale, there are a considerable number of known examples of housepits with several rebuilding phases (e.g., Schlanger 1988; Hayden 1997: 247; 2000a: 313; Renouf & Murray 1999; Johnson & Wilmerding 2001; Prentiss et. al. 2003; for ethnological references see LaMotta & Schiffer 1999: 23). The black charcoal layers in the housepits excavated at the Keatley Creek site on the Canadian Northwest Plateau have been thought to result either from the mixing of charcoal ground underfoot with decomposed organic waste in the floor sediments, or from the collapse of burnt roof structures into the housepit (Hayden 1997: 37 8, 40 2). Some of the material on the wall embankments or on the roof, for example, fire- cracked stones, may also have fallen into the housepit following the collapse of the wooden supporting structures (Hayden 1997: 63 4; Johnson & Wilmerding 2001; Kankaanpää 2002; Ojanlatva & Alakärppä 2002). In the areas with better preservation of organic material, distinct layers containing floor assemblages and remains of collapsed roofs can be identified (e.g., Schlanger & Wilshusen 1993; Hayden 1997; 2000a). In some cases charred wooden structures have preserved also in Finland (Karjalainen 1996; Katiskoski 2002; Leskinen 2002) and in Russian Karelia (Zhul nikov 2003: 180 3), but multiple building phases have not been observed at these sites. Since separate floor layers or layers formed by collapsed structures were not identified at the Rupunkangas 1A site, partly because of the poor preservation of organic material and partly because of the somewhat mixed nature of the layers, it is not possible to determine the formation process of the blackish layers with certainty. Nonetheless, it is clear that the use of fire is somehow involved. Experimental burnings of pithouse replicas, based on the Anasazi pithouses found in south-western Colorado, have shown that the accidental burning of such a pithouse is very unlikely (Wilshusen 1986: 247; Schiffer 1987: 92 with references). It should be remembered that in Anasazi pithouses the roof, based on vertical posts, was covered with earth in contrast to Neolithic pithouses found in Fennoscandia, where birch bark covered with sod has been considered the most typical roofing material (Sirelius 1921: 198 9; Kotivuori 1993: 143 4; 2002: 83; Vaara 2000: 6 7; Muurimäki 2007: 102 103). In such a pithouse only the lowest part of the roof might have been covered with actual soil. Still, since it is not known how easily sod-covered pithouses catch fire, it cannot be said how likely it is for a pithouse like this to burn down accidentally. Moreover, our understanding of Mesolithic housepits and their constructions is still in its infancy the Mesolithic housepit at the Rupunkangas 1A site may not necessarily have been constructed like the Neolithic examples. The Mesolithic housepits at the Rupunkangas 1A site have been dug so deep that the pithouse could 16

easily have been completely under ground, like a dugout. Even though there is not much that we can say for sure about Mesolithic pithouses, it is possible that the Mesolithic occupation phases of the pithouses at Rupunkangas 1A ended with the burning of the old structures. The function of the site The presence of housepits is usually connected to a fairly sedentary way of life. According to ethnographic sources, pithouses are used in nontropical climates and inhabited during the wintertime by populations with at least a bi-seasonal settlement system (Gilman 1987, see also Binford 1990). However, in areas with a relatively cool climate year-round, such as coastal Siberia and Newfoundland, pithouses could be used also as summer houses (Gilman 1987: 542 3; Renouf & Murray 1999: 121). In the archaeological literature dealing with the northernmost areas of Europe, housepits are generally connected to (semi-sedentary) winter habitation (e.g., Kotivuori 1993: 142, 145 9; Lundberg 1997: 125 6, 136; Katiskoski 2002; Ojanlatva & Alakärppä 2002: 121; Pesonen 2006) or year-round permanent settlement (e.g., Engelstad 1984: 19 20; Helskog 1984: 65; Schanche 1993; Halen 1994: 177; Karjalainen 1999; Leskinen 2002: 168; Núñez & Okkonen 1999). Despite the general tendency to see a relatively strong connection between pithouses and at least a semi-sedentary settlement pattern, pithouses located in environmentally and ecologically marginal 4 areas are not necessarily occupied for long periods. In areas where predictable resources are available during some time of the year, pithouses could be repeatedly used for short-term occupation (Renouf & Murray 1999; Smith 2003: 163, 165, 180 2; see also Fletcher 1995: 168). In the case of Rupunkangas 1A, the low density of finds (as compared to find material found in the core area of the Stone Age settlement) with respect to the thickness of the layers can be seen to indicate short-term occupation of the site (see Kent 1992: 641 3; Smith 2003: 173 4). In the case of Rupunkangas 1A, it is assumed that the actual duration of the occupation of the site is reflected more reliably by the number of finds than by the thick cultural layers, the formation of which is presumably connected more closely with abandonment and post-abandonment processes than with the actual occupation phase of the site. The building of facilities such as pithouses more costly than ephemeral structures such as simple huts or wind shelters at a site located in a marginal area indicates a multi-year planning range in the exploitation of predictable resources (Smith & McNees 1999: 118; Smith 2003: 182 3). The existence of permanent hunting or fishing camps with pithouses is not strictly connected to any particular type of residential mobility. Permanently used sites with periodical seasonal occupation indicate the existence of a logistically organized system that can occur alongside different types of residential mobility from largely sedentary to wholly mobile (Diehl 1997: 181 2, see also Binford 1980: 12 13, 19; Kent 1992: 637; Hood 1995: 99). Following the arguments of Wandsnider (1992) it can be supposed that permanently attractive locations with facilities such as pithouses would be reused for long periods. Once the pithouse or houses were no longer inhabitable but the location was still attractive, a new pithouse would be built at a new site within the same location. In the Rupunkangas area the seasonal settlement could have persisted on the same ancient island from year to year with a slight shift in location every now and then (see also Kotivuori 1993: 143 5; Kankaanpää 2002: 73 4). The location of the Rupunkangas 1A site, as well as of the other sites in Rupunkangas area, provides some hints as to the function of the site. Other sites in the Vuoksi River Valley with long periods of occupation are located on the shores of narrow fjord-like bays near the mainland, where the heartland of the hunter-gatherer settlement in the area has clearly been. The location of the ancient Rupunkangas island is very similar to the locations of ethnologically known seasonal camps used for fishing in the autumn and for sealing during late winter. During the late 19 th and early 20 th century sealing in Lake Ladoga was concentrated in the northernmost areas of the lake, that is, in the area of the former parishes of Jaakkima and Sortavala (Lehtonen 1974: 136). There, a chain of islands created a zone perfectly suited for sealing. The islands both created conditions where the ice cover lasted longest, providing good conditions for seal hunting, and offered good campsites for 17

Fig. 10. Sites with housepits in the Rupunkangas area. 1) Pontuksenhauta 3, 2) Rupunkangas 4 (an oblong depression modified by man), 3) Rupunkangas 2, 4) Pontuksenhauta 2, 5) Pontuksenhauta 1, and 6) Rupunkangas 3. Redrawn by T. Mökkönen after Halinen & Mökkönen 2005; Mökkönen & Nordqvist 2005; Mökkönen et al. 2005. 18

use during hunting trips. In the northernmost part of Lake Ladoga one of the most popular campsites was Vossina Island, where the fishing cabins were used both during the salmon fishing season in the autumn and the sealing season during the winter (Lehtonen 1974: 136 7). Despite the environmental similarities between the regions around the Rupunkangas area during the Ancient Lake Ladoga Phase and the best sealing areas in the Lake Ladoga region some hundred years ago, the modest size of the osteological material only few bones of salmonid fishes and seals does not allow us to draw any definite conclusions about the function of the sites in the Rupunkangas area. However, at least the Rupunkangas 1A site may be tentatively interpreted as a seasonally occupied hunting and fishing camp with permanent structures that were utilized for a long time. However, this interpretation is not directly valid for all sites in the Rupunkangas area. THE OLDEST HOUSEPITS IN THE VICINITY OF THE FINNISH-RUSSIAN BORDER Housepits dating to the Stone Age are a recently discovered phenomenon on the Karelian Isthmus. The first housepits were found as late as 1999 5 (Lavento et al. 2001; see also Mökkönen et al. 2007: 117). On the Finnish side of the Finnish- Russian border a number of housepits were found in the area of Ancient Lake Saimaa in the surveys carried out in connection with a research project during 1990s (see Pesonen 2002 with references). However, due to the transgression of Ancient Lake Saimaa prior to the formation of the Vuoksi River ca. 4000 cal. BC, most of the Mesolithic sites in this region have been destroyed or covered by transgression layers. In other words, Mesolithic sites are rare in the Lake Saimaa area (see Koivikko 2000). This is not the case in the northern parts of the Lake Saimaa catchment, where a number of Mesolithic sites are located. Nonetheless, the known housepits with only one exception date to the Neolithic Stone Age and a few examples to the Early Metal Period. The only known site with a Mesolithic housepit near the Finnish-Russian boarder was discovered in 2004 at Rahakangas in Eno Parish, Finnish North Karelia. According to a radiocarbon date on burnt bone found in a round housepit 5 to 6 meters in diameter, this housepit dates to a period between 8800 8550 cal. BC (1 sigma) (Pesonen 2005: 4 5). This date is the second oldest radiocarbon date obtained from a dwelling site in Finland. Another site with Mesolithic housepits is located on the southern coast of Finland. More than ten housepits have been recorded at the Susikopinharju 1 site in Pyhtää Parish, and some of these have also been excavated. Based on the finds and shore displacement chronology, these round and roundish housepits date to the Late Mesolithic, most probably to the 7 th millennium cal. BC, although the elevation of the highest areas of the site would allow an even earlier date (Miettinen 1998: 19 22; Miettinen 1999). In the area of the Karelian Republic on the north north-eastern side of Lake Ladoga and around Lake Onega, the oldest known housepits appear during the 7 th millennium cal. BC (Filatova 1996; Zhul nikov 2003: 44). These oldest housepits are rectangular in shape and the frame construction is based on supporting posts. The round-shaped pithouses are a slightly younger phenomenon in the area, appearing from ca. 6000 cal. BC (Zhul nikov 2003: 43 4, 101 102, 115 20, Table 2). In the sphere of the Veretye culture to the east of Lake Onega, the dwelling site Veretye 1 contains several housepits (Oshibkina 2006: 8 10). The oldest date from the site is 9600 ± 80 BP (Le-1469, ca. 9200 8800 cal. BC), the youngest one being 7700 ± 80 BP (Le-1473, ca. 6600 6460 cal. BC). The oldest radiocarbon date dates to the end of Preboreal period, but the Veretye culture dates for the most part to the Boreal period (Oshibkina 2006: 27). Unfortunately, there are no radiocarbon dates connected to the housepits. The Russian chronology of the area of the Karelian Republic is based on radiocarbon dates on charcoal, as are also the dates obtained from the Rupunkangas 1A site. At Rupunkangas 1A the two oldest dates are older than the oldest radiocarbon dated housepits in the Karelian Republic. Thus, the initial building phase of the pithouses at Rupunkangas 1A is older than the oldest previously known housepits on the Karelian Isthmus and in the Karelian Republic. OTHER HOUSEPITS IN THE RUPUNKANGAS AREA Altogether, there are 17 housepits on the ancient 19

Table 3. The archaeological sites in the Rupunkangas area. Rupunkangas island (Table 3, Fig. 10). The largest concentration of housepits is at the Pontuksenhauta 1 site, where eight housepits are located on the top of a narrow esker formation. The largest housepit at Pontuksenhauta 1 is 10 x 7 m in size and 0.8 1 m in depth (see Lavento et al. 2001). Its shape is rectangular with entrances at both short ends, and there is a clear embankment around it. The other housepits at the Pontuksehauta 1 site are round, oval or rectangular and a bit smaller, the smallest round dwelling being only 2.5 m in diameter. 6 At the nearby site Pontuksenhauta 3, there are two rectangular housepits 8 x 6 m and 6 x 5 m in size. These roughly 50 cm deep housepits have been dug into a narrow terrace between two shore formations. Both dwellings were dug partly into the edge of the upper terrace. In the most southern part of the Rupunkangas area there are two housepits side by side at the Rupunkangas 3 site, which is radiocarbon dated to the beginning of the Boreal period (Hela 1165, 8740 ± 40 BP). One of the pits at the site is rectangular (10 x 6.5 m in size and 0.6 m in depth) and the other is round (4.5 m in diameter and 0.6 m in depth). These housepits are not located on the top of the terrace but in the background of the dwelling site, approximately 40 50 meters away from the shore terrace. The richest find area, observed on the mineral soil surface uncovered by a logging tractor, is located between the housepits and the terrace edge. In addition to housepits, the structures also include a large oblong depression 20 x 5 meters in size located at the Rupunkangas 4 site (see also Mökkönen et al. 2007: 119). 7 This depression is of natural origin but has been modified by man. Fire-cracked stones and quartzes line the embankments and at the bottom there is a lightly coloured cultural layer in which two quartz flakes were found while test pitting during the survey. However, most of the bottom area in this depression did not yield any signs of a cultural layer or any finds. Even though there is extensive variation as regards the shape, size and topographic location 20

Table 4. Phosphate values from the sites in the Rupunkangas area. of the housepits, there is also one common feature. A trait shared by all the housepits is the low number of finds, even though test pits were dug during surveys. None of the housepits yielded any datable finds during test pitting. In all, the number of finds directly connected with the housepits is insignificant, consisting of only a few quartz flakes and pieces of burnt bone. In addition to the low number of finds, the cultural layers are lacking in some of the housepits. In order to explore the cultural impact on the soil inside the housepits, sporadic phosphate analyses were carried out (Table 4) (Mökkönen et al. 2005; Mökkönen & Nordqvist 2005). According to the analysed samples from the Rupunkangas area and the nearby Autio 1 site, the natural background values in the area are quite high, from about 300 mg/kg to values as high as nearly 400 mg/kg. In all of the phosphate analyses carried out during the Kaukola-Räisälä Project, only three samples actually had a phosphate content lower than 300 mg/kg. In the phosphate analyses carried out in Finland in the Lake Saimaa area, the natural phosphate values have been somewhat lower, about 200 250 mg/kg (Matiskainen & Jussila 1984: 35; Lahelma 2006: 9; 2007: 54). In northern Finland the natural values are even lower, lying clearly below 200 mg/ kg (Carpelan & Lavento 1996: 99 102). At the Pontuksenhauta 1 site the largest and one of the small housepits with intensely coloured cultural layers produced phosphate values as high as over 800 mg/kg. 8 At the same site another small housepit without any signs of cultural layers or finds contained a natural phosphate concentration of 292 mg/kg. 9 These results are reasonable, but at the same time contrary results were obtained. At the other analysed sites high phosphate values were also obtained from samples taken from inside the structures, but in some cases there were no traces of cultural layers. At the Rupunkangas 3 site the rectangular housepit with a cultural layer contained clearly elevated phosphate values (853 mg/kg). In the adjacent round housepit at the same site the phosphate values were slightly elevated (422 mg/kg) regardless of the absence of cultural layers. At the Rupunkangas 4 site the samples taken from the bottom of the natural oblong depression partly modified by man contained high phosphate values (560 848 mg/kg) even though the cultural layer was missing. On the basis of phosphate analyses carried out in connection with the surveys, the presence of high phosphate values in the housepits and other depressions utilized by man is not directly dependent on the existence of a visible cultural layer. High phosphate values were observed connected both to heavily coloured cultural layers and to samples without any visible cultural content. A housepit without any cultural layer or finds or raised phosphate values as in the case of the housepit at Pontuksenhauta 1 is an interesting phenomenon, especially when such a feature is found in the same cluster as other housepits with clear cultural layers. The raised phosphate values, in some of the samples even higher than 800 mg/ kg, clearly indicate quite a wide range of human activity, which is not limited only to the areas with 21

artefacts and visible cultural layers. The differences between housepits at the same site as regards phosphate values and the presence of a cultural layer and finds might indicate differentiation in the activities performed in different pithouses. However, this suggestion is purely hypothetical, since it is not known whether the pithouses were contemporaneous or not. The reason for the great variation in the outward appearance of the housepits in the Rupunkangas area and in the characteristics of the layers inside them cannot be resolved here. The main question in this issue is the temporal relationship between the houses are they synchronous or not? Whatever the answer may be, at the moment it seems that the variation is more closely related to time than to variation between synchronous housepits. This is, however, only an assumption based on a notion of certain time-related changes in the shape and dimensions of housepits (see Mökkönen 2002; Okkonen 2003: 226 8; Núñez & Okkonen 2005: 29; cf. Pesonen 2002). 10 CONCLUSIONS At the Rupunkangas 1A site the location of the small-scale excavation area in the narrow strip of land between two sand pits made it possible to gain additional information about the stratigraphy of the site. It would not be possible to understand the context of the excavated cultural layers solely on the basis of the data obtained from the excavation area. The sectioning of the sand pit next to the excavation area clearly showed that the cultural layers at the site had formed in a man-made depression that cut into the natural layers of sandy till. The stratigraphical observations together with the observations made on the surface made it possible to interpret the depression as a housepit. The stratigraphy, finds, and radiocarbon dates together provided insights as concerns the chronological depth of the site. The four radiocarbon dates from the lowest blackish layers covered nearly the whole Mesolithic Stone Age. In the upper layers the ceramics dating to the Late Neolithic and the Early Metal Period provided evidence as to the youngest occupation phases at the site. On the basis of the oldest radiocarbon date (ca. 8000 7650 cal. BC) the initial occupation phase of the site is practically synchronous with the isolation of the Lake Ladoga Basin from the Baltic Sea Basin. This date also represents the first building phase of a pithouse at the site. The last occupation phase at the site is connected to Textile Ceramics and dates most probably to the period preceding the formation of the River Neva ca. 1300 cal. BC. Due to the poor preservation of the organic material and the partly mixed nature of the layers, no unambiguous strata that could be interpreted as floor layers or layers originating from the collapse of the structures were found. The interpretation that the remains represent several Mesolithic rebuilding phases is based on the radiocarbon dates and partly on stratigraphical observations. The gathered data do not allow conclusions to be drawn as concerns the origin of the Neolithic layers with Late Comb Ware, that is, are we looking at a rebuilding phase or a dump? On the basis of the stratigraphy the youngest rebuilding phase of the pithouses at the site is connected to the layers with Textile Pottery. The housepit visible in the surface elevation model is also connected to the last rebuilding phase. The oldest radiocarbon date (ca. 8000 7650 cal. BC) connected to the housepit at the Rupunkangas 1A site is considerably older than the oldest dates from housepits on the Karelian Isthmus or in the Karelian Republic. On the Karelian Isthmus there are no other sites with Mesolithic housepits, and in the Karelian Republic the oldest housepits date to 7 th millennium cal. BC. The oldest date from Rupunkangas 1A is, however, nearly a thousand years younger than the date obtained from the housepit at the Rahakangas site in Finnish North Karelia. Finds from all the sites indicate that the Rupunkangas area was used throughout its existence as an island, that is, ca. 8000 1300 cal. BC. The only temporal gap in the archaeological material is in the Early Neolithic period no finds can be connected to the period synchronous with Early Comb Ware. Nonetheless, the archaeological data from the Rupunkangas area and the absence of an alternative site of similar nature in the nearby archipelago points to a fairly continuous seasonal or permanent settlement in the Rupunkangas area. The continuation of the occupation through the millennia shows that the exploitation of resources in the archipelago of Ancient Lake Ladoga has remained, at least on the face of it, unchanged all through the Stone Age. Despite the strong connection between pithouses and an at least semi-sedentary settlement 22

pattern, pithouses located in a marginal area do not necessarily represent permanent residental sites. This notion, together with the environmentally marginal location of the Rupunkangas area in the archipelago far from the heartland of the hunter-gatherer settlement of the area and with a low number of finds in relation to the thickness of the cultural layers, lead to the interpretation of the excavated Rupunkangas 1A site as a regularly used seasonal camp for sealing and fishing. The occupation period has most likely been in the autumn and winter. This conforms well with the osteological material gathered from the sites in the Rupunkangas area, although it must be remembered that the number of identified bone fragments is low. The interpretation of the Rupunkangas 1A site as a seasonally occupied camp used regularly for the exploitation of aquatic resources is largely based on the low number and monotonous character of the finds and the environmental location of the site, which could be categorized as a marginal area. Especially the location of the site is seen to indicate why this particular site was occupied. The formation of the black thick cultural layers in the housepit is assumed to be connected to the abandonment processes rather than to the actual occupation phases of the site. The seasonal camp interpretation presented here is launched as a contradictory view with respect to the typical, overly direct and simplified argumentation where the occurrence of housepits is equated with more permanent, sedentary or semi-sedentary settlement at the site. Another, more traditional interpretation of the Rupunkangas 1A site might also be possible, but it appears that the available archaeological material alone would not support that interpretation, either. The conclusions drawn on the basis of archaeological data, which does not speak for itself, usually requires reasoning and the making of assumptions before the final statements. In the case of the Rupunkangas 1A site the interpretation either as a seasonally used camp or permanent dwelling site could not be read directly from the archaeological data the method of reasoning on the basis of the data points to certain conclusions. It is not possible to extend the interpretation of the Rupunkangas 1A site to the other sites in the Rupunkangas area. The low number of finds connected directly with the housepits is evident on every site, but at some of the sites, namely Rupunkangas 2 and Rupunkangas 3, the existence of well-developed cultural layers with a notable number of finds points to a fairly intensive occupation of the sites. Current archaeological data do not allow a comprehensive interpretation of the sites in Rupunkangas area. In order to gain a comprehensive understanding of the diverse sites in the Rupunkangas area, more detailed archaeological data, obtained through excavations and new radiocarbon dates, is needed. ACKNOWLEDGEMENTS The authors cooperated on the excavation at the Rupunkangas 1A site. Teemu Mökkönen, M.A. (University of Helsinki, Department of Archaeology), the principal author of this article, worked as director of the excavation and wrote the original English version of the excavation report. Kerkko Nordqvist, M.A. (University of Helsinki, Department of Archaeology) acted as cartographer and was responsible for the graphical recording of the excavation. Stanislav Bel skij, M.A. (Peter the Great Museum of Anthropology and Ethnography, Russian Academy of Sciences, Kunstkamera, St. Petersburg), the supervisor of the excavation, was responsible for the Russian version of the excavation report based on the English version. The excavation was conducted under Mr. Bel skij s archaeological fieldwork licence. The authors wish to thank the following persons for their participation in the fieldwork carried out in the Rupunkangas area during the Kaukola- Räisälä Project: Dmitriy Gerasimov, M.A. (St. Petersburg), Sergey Lisitsyn, Ph.D. (St. Petersburg), Boris Lych, M.A. (St. Petersburg), Maria Jushkova, M.A. (St. Petersburg), Professor Mika Lavento (Helsinki), Petri Halinen, Ph.D. (Helsinki), Oula Seitsonen, M.A. (Helsinki) and Sanna Seitsonen (née Puttonen), B.A. (Helsinki). The Kaukola-Räisälä Project was financed mainly by the University of Helsinki, but minor costs were also covered by the Russian co-operators. The writing of the article and the language revision was funded by the Finnish graduate school in archaeology. We would also like to thank the anonymous referees and Mikael A. Manninen, M.A., for commenting on the manuscript. The language has been revised by Jarmo Kankaanpää, Ph.D. 23

The calibration of radiocarbon dates in the article has been carried out using the OxCal Version 3.10 calibration program ( C. Bronk Ramsey 2005). NOTES 1 The pit structures, which are the remains of semi-subterranean structures used as dwellings, have many names. In Finnish archaeological literature published in English, terms like dwelling depression, pithouse and semi-subterranean house remain have been most popular. In this article the term housepit, adapted from North American archaeological literature, is used (in Finnish archaeology this term is also used in Kankaanpää 2002 and Hertell & Manninen 2006; see also Karjalainen 1996). The term pithouse refers to a whole pithouse and its standing wooden structures. When the structures have collapsed and only a pit remains, the preferred term is housepit (Hayden 1997). A housepit is an archaeological remain of a pithouse. 2 The dating of the formation of the Neva outlet is based on radiocarbon dates published by Saarnisto and Grönlund (1996). In this article, they give two reliable dates, namely 3020 ± 55 BP (Su-2271) and 3200 ± 80 BP (Su- 2275), the former of these being more suitable for dating the River Neva (Saarnisto & Grönlund 1996: 210). The calibration of the dates gives 1390 1200 cal. BC (67.5 % probability) for the former date and 1540 1390 cal. BC (60.1 % probability) for the latter date (OxCal Version 3.10 C. Bronk Ramsey 2005). Even though the approximate date of 1300 cal. BC for the formation of the River Neva is used here, it must be borne in mind that the frame of the exact date is much wider. 3 The shifting water level is due to the uneven speed of land uplift within the area. This means that the older the shore level, the more tilted it is at present. 4 Here, marginal areas are defined as areas located far from the heartland of the settlement, where only a few, often seasonally exploitable, resources are available. The study area, being located in the archipelago, was also environmentally harsh as compared to the mainland. 5 Possibly at least one housepit has been excavated during the active research period in the area in the early 20th century. The first Stone Age dwelling structure, the Räisälä hut, excavated and reconstructed by Pälsi (1918) at the Pitkäjärvi site in former Räisälä Parish, has upon re-examination proven to be a possible half of a housepit (Seitsonen 2006). 6 Finnish archaeological literature lacks an adequate definition of a housepit (or dwelling depression or semisubterranean house remain). As concerns the sizes of the pithouses, in a review article on housepits in Finland the smallest pits regarded as housepits and taken into consideration are ca. 4 meters in diameter (Pesonen 2002: 13). However, this is not a rule the limitation is made merely for practical purposes. In other areas the existence of housepits smaller than 4 meters in diameter is recognized. For example, the average diameter of the 41 excavated housepits in Wyoming (USA) was only 3.15 meters, the smallest diameter of a housepit being 1.7 meters (Smith 2003). In Sweden, the definition of the Neolithic housepits presented by Lundberg (1997: 13) argues for 2 3 meters in diameter as the minimum size of a housepit. 7 In the 2004 survey several round, 2.5 meter-wide pits were found at the site near the oblong depression. These were preliminary interpreted as cooking pits (see Mökkönen et al. 2007: 119). One of the pits excavated in 2005 turned out to be a traditional tar-burning pit (Mökkönen 2005b). The charred wood at the bottom of the pit was dated to 370 ± 40 BP (Hela-1181). A similar structure was also excavated at the Rupunkangas 1A site (see Fig. 5). 8 The phosphate samples were analysed at the laboratory of the University of Helsinki Department of Archaeology by Paula Kouki. Due to the limited resolution of the equipment used, the calibration of values higher than 800 mg/kg is only suggestive. In practice, this means that samples with a phosphate content of some 800 1000 mg/kg may give fairly similar values in the analyses (Paula Kouki, pers. comm.). 9 The absence of cultural layers and finds was easy to ascertain since a rectangular pit some 150 x 50 x 50 cm in size had been dug into the middle of the housepit for an unknown reason. 10 This, again, raises the question, how should a housepit be defined? Is the type of soil at the site where the house is erected a reasonable basis for classifying different types of semi-subterranean structures used as dwellings? According to present convention semi-subterranean structures erected on sand or gravel are called housepits while similar structures erected on rocky moraines are called settlement embankments (see Pesonen 2002: 13 14; Okkonen 2003: 28 30). 24

REFERENCES Unpublished sources Halinen, P., Lavento, M., Timofeev, V. & Saksa, A. 1999. The Archaeological Survey of Kaukola and Räisälä in Karelian Isthmus. The 1999 Report. Unpublished report at the Department of Archaeology, University of Helsinki. Halinen, P. & Mökkönen, T. 2004. The Archaeological Survey in Kaukola and Räisälä on the Karelian Isthmus. Kaukola-Räisälä -project: Report 2004. Unpublished report at the Department of Archaeology, University of Helsinki. Kotivuori, H. 2002. Alisen Kemijoen kivikautiset asumuspainanteet topografiseen havainnointiin ja aineistovertailuun perustuva asutuskuva. Licentiate s thesis. Department of Archaeology, School of Cultural Research, University of Turku. Mökkönen, T. 2005a. Kaukola Rupunkangas 1A. Excavation report 5 th 16 th July, 2005. Unpublished report at the Department of Archaeology, University of Helsinki. Mökkönen, T. 2005b. Kaukola Rupunkangas 4. Excavation report 26 th 28 th July, 2005. Unpublished report at the Department of Archaeology, University of Helsinki. Mökkönen, T. & Nordqvist, K. 2005. Kaukola Rupunkangas 3. Research report 10 th May and 12 th July, 2005. Unpublished report at the Department of Archaeology, University of Helsinki. Mökkönen, T., Seitsonen, O., Nordqvist, K. & Puttonen, S. 2005. The Archaeological Survey in Kaukola and Räisälä on the Karelian Isthmus. Kaukola-Räisälä project. Report 2005. Unpublished report at the Department of Archaeology, University of Helsinki. Puttonen, S. 2004. Osteological Analysis 05.11.2004. Survey of Kaukola and Räisälä 2004. Manuscript in the authors possession. Puttonen, S. 2005. Osteological Analysis 06.06.2005. Survey of Kaukola and Räisälä 2005. Manuscript in the authors possession. Seitsonen, O. & Nordqvist, K. 2006. Arkeologiset tutkimukset Pyhäjärvellä 2005 2006. Poster presented at the yearly fieldwork presentation in the National Museum of Finland, November 2006. Literature Ailio, J. 1909. Die Steinzeitlichen Wohnplatzfunde in Finnland I II. Helsingfors. Binford, L.R. 1980. Willow smoke and dogs tails: hunter-gatherer settlement systems and archaeological site formation. American Antiquity 45(1): 4 20. Binford, L.R. 1990. Mobility, housing, and environment: a comparative study. Journal of Anthropological Research 46(2): 119 52. Carpelan, C. 1999. Käännekohtia Suomen esihistoriassa aikavälillä 5100 1000 ekr. In P. Fogelberg (ed.), Pohjan poluilla: suomalaisten juuret nykytutkimuksen mukaan: 249 80. Bidrag till kännedom av Finlands natur och folk 153. Finska vetenskaps-societeten, Helsinki. Carpelan, C., in press. On the history and recent studies of the Antrea net find. To be published in Iskos 15. Carpelan, C. & Lavento, M. 1996. Soil phosphorous survey at subrecent Saami winter village sites near Inari, Finnish Lapland: a preliminary report. In V. Mejdahl & P. Siemn (eds.), Proceedings from the 6 th Nordic Conference on the Application of Scientific Methods in Archaeology, Esbjerg 1993: 97 107. Arkeologiske Rapporter nr. 1. Esbjerg Museum, Esbjerg. Carpelan, C. & Parpola, A. 2001. Emergence, contacts and dispersal of Proto-Indo-European, Proto-Uralic and Proto-Aryan in archaeological perspective. In C. Carpelan, A. Parpola & P. Koskikallio (eds.), Early Contacts between Uralic and Indo-European: Linguistic and Archaeological Considerations. Papers Presented at an International Symposium held at Tvärminne Research Station of the University of Helsinki, 8 10 January 1999: 55 150. Suomalaisugrilaisen seuran toimituksia 242. Suomalais-ugrilainen seura, Helsinki. Diehl, M.W. 1997. Changes in architecture and land use strategies in the American Southwest: upland Mogollon pithouse dwellers, A.D. 200 1000. Journal of Field Archaeology 24(2): 179 94. Engelstad, E. 1984. Diversity in arctic maritime adaptations: an example from the Late Stone Age of Arctic Norway. Acta Borealia 2(1): 3 24. Filatova, V.F. 1996. Mezolit. In M.G. Kosmenko & S.I. Kochkurkina (eds.), Arkheologiya Karelii: 36 61. Karel skii nauchnyi tsentr, RAN, Petrozavodsk. Fletcher, R. 1995. The Limits of Settlement Growth: A Theoretical Outline. Cambridge University Press, Cambridge. Forsberg, O. 2006. Jänisjoen reitin varhaisin asutus: inventointituloksia Laatokan pohjoispuolelta. Muinaistutkija 1/2006: 2 13. Gerasimov, D.V., Nordkvist, K., Seitsonen, O. & Subetto, D.A. 2007. Kompleksnoe izuchenie pamyatnikov kamennogo veka v mikroregione Otradnoe na Karel skom peresheike. Predvaritel nye itogi i perspektivy issledovanii. Radlovskie chteniya 2006. St. Petersburg. Gilman, P.A. 1987. Architecture as artefact: pit structures and pueblos in the American Southwest. American Antiquity 52(3): 538 64. Halen, O. 1994. Sedentariness during the Stone Age of Northern Sweden in the Light of the Alträsket site, c. 5000 B.C., and the Comb Ware Site Lillberget, c. 3900 B.C.: Source Critical Problems of Representativity in Archaeology. Acta Archaeologica Lundensia, Series in 4, No. 20. Almqvist & Wiksell, Stockholm. Halinen, P., Seitsonen, O., Seitsonen, S. & Nordqvist, K., in press: Excavations at the Juoksemajärvi Westend Stone Age dwelling site in 2002. To be published in Iskos 15. Hayden, B. 1997. The Pithouses of Keatley Creek: Complex Hunter-Gatherers of the Northwest Plateau. Harcourt Brace Collage Publishers, Fort Worth. Hayden, B. 2000a. Site formation processes at Keatley Creek. In B. Hayden (ed.), The Ancient Past of Keatley Creek, Volume 1: Taphonomy: 299 336. Archaeology Press, Burnaby. 25

Hayden, B. 2000b. Dating deposits at Keatley Creek. In B. Hayden (ed.), The Ancient Past of Keatley Creek, Volume 1: Taphonomy: 35 40. Archaeology Press, Burnaby. Helskog, K. 1984. The younger Stone Age settlements in Varanger, North Norway. Acta Borealia 1(1): 39 70. Hertell, E. & Manninen, M. 2006. House pit formation processes: a preliminary assessment of pit 4 at Rävåsen, southern Ostrobothnia, Finland. In V.-P. Herva (ed.), People, Material Culture and Environment in the North: Proceedings of the 22nd Nordic Archaeological Conference, University of Oulu, 18-23 August 2004: 183 97. Studia humaniora ouluensia 1. University of Oulu, Oulu. Hood, B. 1995. Circumpolar comparison revisited: hunter-gatherer complexity in the north Norwegian Stone Age and the Labrador maritime archaic. Arctic Anthropology 32(2): 75 105. Huurre, M. 2003. Viipurin läänin kivikausi. In M. Saarnisto (ed.), Viipurin läänin historia 1: Karjalan synty: 151 244. Karjalan kirjapaino, Lappeenranta. Johnson, L.L. & Wilmerding, E.G. 2001. Bringing the house down: modelling construction and deconstruction of Aleut semisubterranean houses. In D.E. Dumond (ed.), Archaeology in the Aleut Zone of Alaska: Some Recent Research: 127 149. University of Oregon Anthropological Papers 58. University of Oregon, Eugene. Kankaanpää, J. 2002. The house pits at Kauvonkangas, Tervola. In H. Ranta (ed.), Huts and Houses: Stone Age and Early Metal Age Buildings in Finland: 65 78. National Board of Antiquities, Helsinki. Karjalainen, T. 1996. Pithouse in Outokumpu Sätös excavated in 1992 1994. In T. Kirkinen (ed.), Pithouses and Potmakers in Eastern Finland: Reports of the Ancient Lake Saimaa Project: 71 88. Helsinki Papers in Archaeology 9. University of Helsinki, Helsinki. Karjalainen, T. 1999. Sedentariness and dating stone age houses and sites. In M. Huurre (ed.), Dig It All: Papers Dedicated to Ari Siiriäinen: 185 90. The Archaeological Society of Finland & The Finnish Antiquarian Society, Helsinki. Katiskoski, K. 2002. The semisubterranean dwelling at Kärmelahti in Puumala, Savo Province, eastern Finland. In H. Ranta (ed.), Huts and Houses: Stone Age and Early Metal Age Buildings in Finland: 171 200. National Board of Antiquities, Helsinki. Kent, S. 1992. Studying variability in the archaeological record: an ethnoarchaeological model for distinguishing mobility patterns. American Antiquity 57(4): 635 60. Koivikko, M. 2000. Hukkunut kivikausi vedenalaiset asuinpaikat, uusi muinaisjäännöstyyppi. Muinaistutkija 1/2000: 4 11. Kotivuori, H. 1993. Kivikauden asumuksia Peräpohjolassa vertailua ja rakenteellisia tulkintoja. In R. Huopanen (ed.), Selviytyjät: näyttely pohjoisen ihmisen sitkeydestä: 119 60. Lapin maakuntamuseon julkaisuja 7. Lapin maakuntamuseo, Rovaniemi. Lahelma, A. 2006. Excavating art: a ritual deposit associated with the rock painting of Valkeisaari, Eastern Finland. Fennoscandia archaeologica 23: 3 23. Lahelma, A. 2007. Uhritulilla Valkeisaaressa? Kalliomaalauksen edustalla järjestettyjen kaivausten tuloksia. Aurinkopeura 3: 49 82. LaMotta, V.M. & Schiffer, M.B. 1999. Formation processes of the housefloor assemblages. In P.M. Allison (ed.), Archaeology of Household Activities: 19 28. Routledge, London. Lavento, M. 2001. Textile Ceramics in Finland and on the Karelian Isthmus: Nine Variations and Fugue on a Theme of C.F. Meinander. Suomen Muinaismuistoyhdistyksen Aikakauskirja 109. Finnish Antiquarian Society, Helsinki. Lavento, M., Halinen, P., Timofeev, V., Gerasimov, D. & Saksa, A. 2001. An archaeological field survey of Stone Age and Early Metal Period settlement at Kaukola (Sevastyanovo) and Räisälä (Melnikovo) on the Karelian Isthmus in 1999. Fennoscandia archaeologica 18: 3 25 (including Appendix 1: sites, in Fennoscandia archaeologica 19: 67 73). Lavento, M., Halinen, P. & Mökkönen, T. 2006. Subsistence strategies and changes of communities between 9000 1 BC: an archaeological intensiveinvestigation in the western part of Lake Ladoga, on the Karelian Isthmus (Kaukola-Räisälä project). In V.-P. Herva (ed.), People, Material Culture and Environment in the North: Proceedings of the 22nd Nordic Archaeological Conference, University of Oulu, 18 23 August 2004: 120 30. Studia humaniora ouluensia 1. University of Oulu, Oulu. Lehtonen, J.U.E. 1974. Laatokan hylkeenpyynti. Suomen Museo 1974: 134 46. Leskinen, S. 2002. The Late Neolithic House at Rusavierto. In H. Ranta (ed.), Huts and Houses: Stone Age and Early Metal Age Buildings in Finland: 147 70. National Board of Antiquities, Helsinki. Lundberg, Å. 1997. Vinterbyar ett bandsamhälles territorier i Norrlands inland, 4500 2500 f.kr. Studia Archaeologica Universitatis Umensis 8. University of Umeå, Umeå. Matiskainen, H. 1989. The paleoenvironment of Askola, southern Finland: Mesolithic settlement and subsistence 10000 6000 b.p. In Studies on the Chronology, Material Culture and Subsistence Economy of the Finnish Mesolithic, 10 000 6000 b.p. (Iskos 8): 1 97. Finnish Antiquarian Society, Helsinki. Matiskainen, H. 1996. Discrepancies in deglaciation chronology and the appearance of man in Finland. In L. Larsson (ed.), The Earliest Settlement of Scandinavia and Its Relationship with Neighbouring Areas: 249 62. Acta Archaeologica Lundensia, Ser. 8:o, no. 24. Almquist & Wiksell, Stockholm. Matiskainen, H. & Jussila, T. 1984. Naarajärven kampakeraaminen asumus. Suomen Museo 91: 17 52. Miettinen, A., Sarmaja-Korjonen, K., Sonninen, E., Jungner, H., Lempiäinen, T., Ylikoski, K. & Carpelan, C., in press. The palaeoenvironment of the Antrea net find. To be published in Iskos 15. Miettinen, T. 1998. Kymenlaakson esihistoriaa. Kymenlaakson maakuntamuseon julkaisuja 26. Kymenlaakson maakuntamuseo, Kotka. Miettinen, T. 1999. Keskustelua: mitä Susikopinharjulla tapahtui? Muinaistutkija 2/1999: 35 8. Muurimäki, E. 2007. Stone Age village in Saarijärvi an overview. In J. Vilkuna, J.-P. Taavitsainen & V. Heikkinen (eds.), IV Mittnordiska Arkeologidagar 26

Saarijärvi 14. 16. juni 2007: 91 112. Keski-Suomi 20. Keski-Suomen museoyhdistys, Jyväskylä. Mökkönen, T. 2002. Chronological variation in the locations of hunter-gatherer occupation sites vis-ávis the environment. In H. Ranta (ed.), Huts and Houses: Stone Age and Early Metal Age Buildings in Finland: 53 64. National Board of Antiquities, Helsinki. Mökkönen, T., Lavento, M. & Halinen, P. 2007. Neolithic sites in various environments of the Karelian Isthmus: Kaukola-Räisälä -project s survey 2004. In A.N. Kirpichnikov, E.N. Nosov & A.I. Saksa (eds.), Slavs, Finns and Ougriens: The Zones of Contacts and Cooperation. The papers of Russian Finnish symposium on the Problems of Archaeology and History, Puškinskije Gory 2004, October, 7 10: 112 29. Institute for the History of Material Culture Russian Academy of Sciences, Saint-Petersburg. Mökkönen, Belsky & Nordqvist in press. Protochnoe IV (fin. Rupunkangas 1) novaja stratificirovannaja stojanka épohi mezolita rannego metalla na Karel skom peresheike. (English summary: Protochnoe IV (fin. Rupunkangas 1) A multilayer dwelling site from Early Metal period to Mesolithic Stone Age on Karelian Isthmus). To be published in the series of Tver archaeological conference Tverskoj arheologicheskij sbornik 2006, vol 7. Núñez, M. & Okkonen, J. 1999. Environmental background for the rise and fall of villages and megastructures in North Ostrobothnia 4000 2000 cal BC. In H. Ranta (ed.), Huts and Houses: Stone Age and Early Metal Age Buildings in Finland: 105 15. National Board of Antiquities, Helsinki. Núñez, M. & Okkonen, J. 2005. Humanizing of north Ostrobotnian landscapes during the 4 th and 3 rd millennia BC. Journal of Nordic Archaeological Science 15: 25 38. Ojanlatva, E. & Alakärppä, J. 2002. Interpretation of the Peurasuo House in Oulu. In H. Ranta (ed.), Huts and Houses: Stone Age and Early Metal Age Buildings in Finland: 109 21. National Board of Antiquities, Helsinki. Okkonen, J. 2003. Jättiläisen hautoja ja hirveitä kiviröykkiöitä Pohjanmaan muinaisten kivirakennelmien arkeologiaa. Acta Universitatis Ouluensis Series B Humaniora 52. Oulun yliopisto, Oulu. Oshibkina, S.V. 2006. Mezolit Vostochnogo Prionezh ya. Kul tura Veretye. Rossiiskaya Akademiya Nauk, Moskva. Pesonen, P. 2002. Semisubterranean houses in Finland a review. In H. Ranta (ed.), Huts and Houses: Stone Age and Early Metal Age Buildings in Finland: 9 41. National Board of Antiquities, Helsinki. Pesonen, P. 2005. Sarvingin salaisuus Enon Rahakankaan varhaismesoliittinen ajoitus. Muinaistutkija 2/2005: 2 13. Pesonen, P. 2006. One house two households? An investigation of a Late Subneolithic pithouse in Kuorikkikangas site, Posio, southern Lapland. In V.-P. Herva (ed.), People, Material Culture and Environment in the North: Proceedings of the 22nd Nordic Archaeological Conference, University of Oulu, 18 23 August 2004: 198 213. Studia humaniora ouluensia 1. University of Oulu, Oulu. Prentiss, W.C., Lenert, M., Foor, T.A., Goodale, N.B. & Schlegel, T. 2003. Calibrated radiocarbon dating at Keatley Creek: the chronology of occupation at a complex hunter-gatherer village. American Antiquity 68(4): 719 35. Pälsi, S. 1915. Riukjärven ja Piiskunsalmen kivikautiset asuinpaikat Kaukolassa. Suomen Muinaismuistoyhdistyksen Aikakauskirja 28(1): 1 192. Pälsi, S. 1918. Kaivaus Pitkäjärven kivikautisella asuinpaikalla Räisälässä v. 1915. Suomen Museo 25: 25 34. Pälsi, S. 1920. Ein steinzeitlicher Moorfund bei Korpilahti im Kirchspiel Antrea, Län Viborg. Suomen Muinaismuistoyhdistyksen Aikakauskirja 28(2): 1 22. Renouf, M.A.P. & Murray, M.S. 1999. Two winter dwellings at Phillip s Garden, a Dorset site in Northwestern Newfoundland. Arctic Anthropology 36(1 2): 118 32. Saarnisto, M. 2003. Karjalan geologia: Karjalan luonnonmaiseman synty. In M. Saarnisto (ed.), Viipurin läänin historia 1: Karjalan synty: 21 80. Karjalan kirjapaino, Lappeenranta. Saarnisto, M. & Grönlund, E. 1996. Shore displacement of Lake Ladoga new data from Kilpolansaari. Hydrobiologia 322: 205 15. Saarnisto, M. & Siiriäinen, A. 1970. Laatokan transgressioraja. Suomen Museo 77: 10 22. Seitsonen, O. 2006. Räisälä Pitkäjärvi revisited new interpretations of the dwelling remains. In T. Mökkönen & P. Pesonen (eds.), Arkeologipäivät 2005: arkeologia ja kulttuuri & uutta kivikauden tutkimuksessa: 138 45. The Archaeological Society of Finland, Helsinki. Schanche, K. 1993. Var det store landsbyboplatser i Finnmark i yngre steinalder? Debattinnlegg med utgangspunkt i bosetningsspor fra ca. 2000 f.kr. In K.A. Bergsvik, S. Nygaard & A.J. Nærøy (eds.), Steinalderkonferansen i Bergen 1993: 174 86. Arkeologiske Skrifter 8. Universitetet i Bergen, Bergen. Schiffer, M.B. 1987. Formation Processes of the Archaeological Record. University of New Mexico Press, Albuquerque. Schlanger, S.H. 1988. Patterns of population movement and long-term population growth in Southwest Colorado. American Antiquity 54(4): 773 93. Schlanger, S.H. & Wilshusen, R.H. 1993. Local abandonments and regional conditions in the North American Southwest. In C.M. Cameron & S.A. Tomka (eds.), Abandonment of Settlements and Regions: Ethnoarchaeological and Archaeological Approaches: 85 98. Press Syndicate of the University of Cambridge, Cambridge. Schulz, H.-P. 1996. Pioneerit pohjoisessa: Suomen varhaismesoliittinen asutus arkeologisen aineiston valossa. Suomen Museo 103: 5 45. Siiriäinen, A. 1974. Studies relating to shore displacement and Stone Age chronology in Finland. Finskt Museum 80: 5 22. Sirelius, U.T. 1921. Suomen kansanomaista kulttuuria II. Helsinki. Smith, C.S. 2003. Hunter-gatherer mobility, storage, and houses in a marginal environment: an example from the mid-holocene of Wyoming. Journal of Anthropological Archaeology 22: 162 89. Smith, C.S. & McNees, L.M. 1999. Facilities and 27

hunter-gatherer long-term land use patterns: an example from Southwest Wyoming. American Antiquity 64(1): 117 36. Takala, H. 2004a. The Ristola Site in Lahti and the Earliest Postglacial Settlement of South Finland. Lahti City Museum, Lahti. Takala, H. 2004b. Archaeological research in the former jurisdictional district of Äyräpää and excavations at the Telkkälä site in Muolaa, Russia. In P. Uino (ed.), Fenno-Ugri et Slavi 2002, Dating and Chronology: 117 23. Museoviraston Arkeologian osaston julkaisuja 10. National Board of Antiquities, Helsinki. Takala, H. & Sirviö, T. 2003. Telkkälä Muolaa a multiperiod dwelling site on the Karelian Isthmus. Fennoscandia archaeologica 20: 55 77. Uino, P. 1997. Ancient Karelia: Archaeological Studies Muinais-Karjala: arkeologisia tutkimuksia. Suomen Muinaismuistoyhdistyksen Aikakauskirja 104. Suomen Muinaismuistoyhdistys, Helsinki. Uino, P. 2003. Karjalan arkeologiaa 150 vuotta. In M. Saarnisto (ed.), Viipurin läänin historia 1: Karjalan synty: 117 50. Karjalan kirjapaino, Lappeenranta. Vaara, R. 2000. Yli-Iin Kierikin kivikautinen kylä asumuskonsruktiot 1998 1999. Muinaistutkija 2/ 2000: 2 12. Wandsnider, L. 1992. The spatial dimension of time. In J. Rossignol & L. Wandsnider (eds.), Space, Time, and Archaeological Landscapes: 257 82. Plenum Press, New York. Wilshusen, R.W. 1986. The relationship between abandonment mode and ritual use in Pueblo I Anasazi protokivas. Journal of Field Archaeology 13(2): 245 54. Ylimaunu, J. 2000. Itämeren hylkeenpyynti ja ihminenhylje -suhde. Suomalaisen Kirjallisuuden Seuran Toimituksia 773. Suomalaisen Kirjallisuuden Seura, Helsinki. Zhul nikov, A.M. 2003. Drevnie zhilishcha Karelii. Karel skii gosudarstvennyi kraevedcheskii muzeii, Petrozavodsk. Zhul nikov, A.M. 2005. Poseleniya epokhi rannego metalla Yogo-Zapadnogo Pribelomor ya. Petrozavodskii gosudarstvennyi universitet & Karel skii gosudarstvennyi kraevedchskii muzei, Petrozavodsk. Personal comments Halinen, Petri (Ph.D.). 5 th June 2006. Personal communication about the radiocarbon dates made in Kaukola-Räisälä Project by the Department of Archaeology, University of Helsinki. Kouki, Paula (Lic. Phil.) <Paula.Kouki@helsinki.fi>. 17 th November 2005. Tuloksia. A personal e-mail about results of the phosphate analyses including some advice concerning interpretation. 28

Paper III Mökkönen, T. 2008. A Review of Neolithic multi-room housepits as seen from the Meskäärtty site in Virolahti parish, extreme south-eastern Finland. Estonian Journal of Archaeology, Vol. 12, No. 2, 114 151. Publisher: The Estonian Academy of Sciences, Tallinn. ISSN 1736-7484 (electronic) ISSN 1406-2933 (print)

Paper IV Mökkönen, T. 2009. Neolithic housepits in the River Vuoksi Valley, Karelian Isthmus, Russia chronological changes in size and location. Fennoscandia Archaeologica XXVI, 133 161. [Errata & Corrigenda. Fennoscandia Archaeologica XXVII, 108 109.] Publisher: The Archaeological Society of Finland, Helsinki ISSN 0781-7126

Fennoscandia archaeologica XXVI (2009) Teemu Mökkönen NEOLITHIC HOUSEPITS IN THE VUOKSI RIVER VALLEY, KARELIAN ISTHMUS, RUSSIA CHRONOLOGICAL CHANGES IN SIZE AND LOCATION Abstract Since the discovery of the first housepits on the Karelian Isthmus in 1999, intensive surveys in the research area of the Vuoksi River Valley have revealed 82 housepits dispersed among 24 dwelling sites. Most of the housepits date to the Neolithic Stone Age. During the time the housepits were in use, the sites were located on the shores of Ancient Lake Ladoga. The first part of this article deals with pattern recognition: chronological variation in housepit size and shape, and in the placement of pithouses. The clearest change in tradition in pithouse placement is observed in the Typical Comb Ware Period, when pithouses appeared in large numbers in the archipelago. This change is accompanied by an unprecedented increase in pithouse size and numbers. The second part of this article attempts to infer what motivated the observed changes. This paper argues that there was an increase in sedentism during the first half of the 4 th millennium cal BC, in the Middle Neolithic. This is reflected in the development of pithouses and pithouse sites, in shifts in site placement to settings favouring other than winter-only habitation, and in logistical mobility based on water transportation. The article is based on survey data mostly gathered during the Kaukola-Räisälä Project (2004 2005). Keywords: Neolithic, Stone Age, housepits, ancient Lake Ladoga, Karelian Isthmus, Finland, Russia, hunter-gatherers, survey data, sedentism, water transport Teemu Mökkönen, Institute for Cultural Research, Archaeology, P.O. Box 59, FI-00014 University of Helsinki, Finland, E-mail: teemu.mokkonen@helsinki.fi INTRODUCTION This article is the first comprehensive presentation of the Stone Age housepits discovered over the last ten years in the Kaukola Räisälä region in the northeastern part of the Karelian Isthmus, Russia (Fig. 1). Archaeologically, this is the most intensively studied area of the Karelian Isthmus, with a history of research extending back more than a hundred years. 1 Regardless of the long research history, the majority of the data used in this study was gathered during the Kaukola Räisälä Project 2, which included several surveys and small-scale excavations. As a result of the surveys, the number of known Stone Age and Early Metal Period sites doubled. 3 Sites are referred to here by their Finnish names 4, a convention also followed in previous articles (see Lavento et al. 2001; 2006; Mökkönen et al. 2006). The first housepits in the research area were discovered in 1999 (Lavento et al. 2001). At the moment, known housepits number 82, divided among 24 dwelling sites 5 (Appendix 1). Most of the housepits date to the Neolithic 6 Period, with a few exceptions (see Mökkönen et al. 2007). During the Stone Age these sites were located by Ancient Lake Ladoga (ca. 7800/7000 1350 cal BC), the water level of which lay at ca. 21 meters above sea level, i.e., about 16 meters higher than at present. The central question of this paper deals with chronological variation in the housepits. The 133

LIMITATIONS OF CHRONOLOGICAL TOOLS Fig. 1. The research area is located in the River Vuoksi Valley, Karelian Isthmus, Russia. The maximum size of the research area is 30 x 40 kilometers. absence of excavated housepits has directed the analysis towards information available through archaeological surveys. In boreal forest that has never been agricultural land, housepits are still clearly observable. Therefore, it is possible to measure the size and depth and examine the shape and alignment of the housepits without excavation. Other issues also examined here are the distribution of housepits with respect to environmental zones, the housepit sites immediate environment, and the pottery types associated with housepits. The development of housepits and changes in site location provide information, for example, about the degree of sedentariness and the mobility strategies of the groups that inhabited the shores of Ancient Lake Ladoga. This article commences with a discussion of the problems with chronological accuracy regarding the Neolithic material found on the Karelian Isthmus and the general development of housepits with emphasis on the areas in the vicinity of the study area. The article then focuses on the housepits in the lower Vuoksi River Valley and describes the observed changes: the chronological trends in housepit shape and size as well as changes in the housepit sites environmental and topographic setting. Finally, the observations are discussed with emphasis on questions relating to the degree of sedentism, mobility, and the housepit sites suitability for year-round habitation. When the number of excavated sites is low, the presumed chronological distribution of the sites is usually not based on radiocarbon dates. Chronological tools often employed in such cases include shoreline displacement chronology and pottery typology. Land uplift is an active natural force on the Karelian Isthmus. However, sites in the research area cannot be dated accurately by shoreline displacement chronology. Likewise, the pottery typology of the area is still poorly studied and presents a number of unique characteristics. The limitations of these tools are presented in the following. Shoreline displacement The Ancient Lake Ladoga Phase covers the span from the isolation of the lake from the Baltic Sea Basin (ca. 7800 7000 cal BC) to the formation of the River Neva (ca. 1350 cal BC). During this period the lower Vuoksi River Valley was a large bay of Ancient Lake Ladoga, reaching far into the interior of the Karelian Isthmus (see Fig. 10). The research area is located approximately on the same land uplift isobase as the former outlet channel at the Heinijoki threshold east of the town of Vyborg (Saarnisto & Grönlund 1996; Saarnisto 2003; 2008). The Heinijoki threshold, which was active during the whole Ancient Lake Ladoga period, split the ancient lake into two areas: (1) the northwestern area with a higher land uplift rate and regressive water level, and (2) the south-eastern area with a lower land uplift rate and transgressive water level. The water level in the research area, which is located on the same land uplift isobase as the threshold, was nearly stable. Around 4000 cal BC, the formation of the Vuoksi River, a new outflow channel of Lake Saimaa running into Lake Ladoga, accelerated the rise of the water level in the latter, including, of course, the research area. The maximum water level was reached just before the formation of the current outflow channel, the River Neva (ca. 1350 cal BC), the opening of which caused a ca. 10-meter drop in the water level. Nevertheless, the fluctuation of the water level in the research area during the Ancient Lake Ladoga Phase was minor, within a limit of two meters 7 (Saarnisto & Siiriäinen 1970). 134

In addition to slight water level fluctuations, there is another factor that complicates the dating of dwelling sites on the basis of their elevation. During the Stone Age there was a clear tendency to occupy locations with a steep shore profile (Mökkönen et al. 2006: 116). These sites were not affected at all by small water level fluctuations and could consequently have been occupied throughout the Ancient Lake Ladoga Phase. Pottery typologies A detailed pottery typology for the Karelian Isthmus does not exist as of yet. On the basis of typologies created for nearby areas, it can be assumed that the Early Neolithic on the Karelian Isthmus was rather similar as regards dating and ceramic types (see Piezonka 2008). Later, during the late Middle Neolithic and the Late Neolithic 8, regional variation in ceramics increases (Fig. 2). This is evident when comparing the chronologies of areas adjacent to the Karelian Isthmus (see Carpelan 1999; Zhul nikov 1999; 2003; 2005; Lang & Kriiska 2001; Pesonen 2004). Therefore, it is probable that none of the pottery chronologies used in the nearby areas will automatically work on the Karelian Isthmus. Discrepancies in dating certain types of ceramics are suggested by comparisons of AMS-dates from the Karelian Isthmus with Finnish dates. For example, Kierikki Ware found in Finland dates to ca. 3350 2900 cal BC (Pesonen 2004) while an AMS-dated piece of Kierikki Ware from the Johannes Väntsi site, located by the seashore on the Karelian Isthmus, dates notably older, ca. 3770 3530 cal BC (1 sigma) 9 (Huurre 2003: 198 9, 512). As an inverse example, Late Comb Ware from the Viipuri Häyrynmäki site, also once located by the sea, dates to 3500 3030 cal BC (1 sigma) (Pesonen 2004: 91 2). This is nearly 300 years younger than the youngest Late Comb Ware dates obtained from Finland (Pesonen 2004). Recent dates of similar age from south-eastern Finland (Mökkönen 2008: 122 4) demonstrate that Late Comb Ware stays longer in use in the eastern Gulf of Finland than in most other parts of Finland (for Estonian dates see Kriiska 2001; Lang & Kriiska 2001: 92). To sum up the limitations of the chronological tools: The use of shoreline displacement for dating is hazardous because of the minor water level fluctuation and the ancient population s habit of occupying locations with steep shore profiles. At the moment, our knowledge of Middle and Late Neolithic ceramics on the Karelian Isthmus is insufficient for understanding the chronological variation in all its details. There are numerous similarities with the Stone Age ceramics found within the current borders of Finland, but on the other hand, similarities with materials found in nearby Russian areas are highly likely, although at the moment this is poorly understood. I suppose that the Finnish pottery typology and chronology, for the lack of something better, can be used on the Karelian Isthmus as a directional device, yet keeping in mind the limitations of its accuracy. CHRONOLOGICAL TRENDS IN NORTHERN EUROPEAN HOUSEPITS Another approach to understanding the chronology is to analyse the sizes and shapes of the housepits. A certain developmental trend may be observed taking place quite simultaneously in the Northern European forest belt. A preference for larger, deeper, and more oblong housepits (a.k.a. semi-subterranean houses) appears during the late 4 th millennium cal BC, continues universally during the 3 rd millennium cal BC and sporadically in some areas up to the early 2 nd millennium cal BC (Norberg 2008: 159 60). In Finland, the larger and deeper housepits with surrounding embankments and with entrances or antechambers visible on the surface predominantly date from ca. 3500 cal BC to the beginning of the Bronze Age ca. 1800 cal BC (Halinen et al. 2002; 2003; Katiskoski 2002; Mökkönen 2002; 2008; Ojanlatva & Alakärppä 2002; Núñez & Okkonen 2005; Vaneeckhout 2008; 2009). Multi-room housepits, including terrace houses, appear during the 4 th quarter of the 4 th millennium cal BC (Mökkönen 2008). Housepits of smaller size, shallower depth, and nearly equal length and width have a longer period of occurrence. In Finland they date from the beginning of the Typical Comb Ware Period ca. 4000 cal BC to the Early Metal Period (1800 cal BC 50 cal AD) (Mökkönen 2002: 58; Pesonen 2002: 29; Halinen et al. 2003). So far, known housepits from the Mesolithic Period are so few in Finland (see Núñez & Uino 1997; Pesonen 2002), that it is not possible to draw conclusions as to their chronological variation. 135

Fig. 2. A suggestive chronological diagram of the pottery styles in Finland and the Republic of Karelia, Russia. Stone Age ceramics in grey. Early Metal period/bronze Age Early Roman Iron Age ceramics in white. A lighter grey colour towards the end of the column indicates a paucity of dates. Abbreviations: Finland (after Asplund 1997; 2004; Carpelan 1999; Edgren 1999; Lavento 2001; Pesonen 2004) CW1 = Early Comb Ware (aka Sperrings), SÄR1 = Säräisniemi 1 Ware, EAW = Early Asbestos Ware, JÄK = Jäkärlä Ware, CW2 = Typical Comb Ware, CW3 = Late Comb Ware, KIE = Kierikki Ware, PÖL JYS = Pöljä and Jysmä Wares, PYH = Pyheensilta Ware, Corded W = Corded Ware, MZC = Middle Zone ceramics, KIU = Kiukainen ware, TXT = Textile pottery, PAI = Paimio ware, MORBY = Morby Ware, SÄR2 = Sär[äisniemi]2 ceramics. Republic of Karelia Left columns after Kosmenko (2004): Sperrings = Early Comb Ware, PIT-COMB = Pit-Comb Ware, COMB-PIT = Comb-Pit Ceramics, RHOMB-PIT = Rhombic Pit Ceramics, CLASSIC = Classic ceramics (organic and asbestos tempered), TXT = Net ( Textile ) Ware, EIA = Early Iron Age, Net Ware-Ananino mixed type ceramics Right columns after Zhulnikov (1999): C-P & R-P = Comb-Pit and Rhomb-Pit Ceramics, VOI = Voinavolok XXVII ceramics, TXT = Textile pottery 136

In the Republic of Karelia, which lies northeast of the Karelian Isthmus in Russia, the oldest pithouses also date to the Mesolithic Period. These pithouses were either rectangular or round in their ground plan, with a structure supported by upright posts. Around Lake Onega, the first rectangular pithouses with a frame of horizontal logs are associated with Pit-Comb Ware. They date to the first half of the 5 th millennium cal BC. The first interconnected housepits also appear at this time (Zhul nikov 2003: 101 2). In the Republic of Karelia, as in Finland, pithouses were most numerous as well as largest in size during the latter part of the Middle Neolithic. According to the chronology established for the Republic of Karelia, this peak corresponds to the beginning of the Eneolithic Period, ca. 3300 2500 cal BC. Later, but still before the beginning of the Bronze Age ca. 1800 cal BC, the number of pithouses declined and their size decreased. No housepits dating to the Bronze Age are known from this area. The chronologies reflecting the development of housepits are not fully comparable. The Finnish chronology is based on several sources (conventional radiocarbon dates, AMS dates, shore displacement chronology, and other dating methods, e.g. optically stimulated luminescence and thermoluminesence dating) while the Russian chronology relies mainly on conventional radiocarbon dates on charcoal. Regardless of the differences in dating methods, it appears that the building of semi-subterranean houses is an older phenomenon in the Republic of Karelia than in Finland. However, in the Finnish inland Lake District the almost total absence of housepits Fig. 3. Two housepits from the Kaukola Räisälä region. In boreal forest that has never been cultivated the housepits are clearly visible on the ground: a) Roundish depression at the Räisälä Peltola B site (housepit no. 1, 8.3x9.6 m), and b) nearly rectangular housepit with surrounding embankment and an entrance at the Räisälä Mäenrinne 1 site (housepit no. 2, 10x8.5 m, including entrance). Photos: S. Seitsonen (a) and T. Mökkönen (b). 137

Fig. 4. Some housepit sites in the Kaukola Räisälä region and the associated ceramics. Sites: A Juoksemajärvi Westend (below) and Juoksemajärvi, B Juoksemajärvi West, C Mäenala, D Mömmönsalmi 1, E Repokorpi, F Peltola A, B, and C (the clusters are named from south to north), G Sylijärvi SW2, and H Seppälä 4. All sites are located in the former municipality Räisälä. Ceramics: CW1 = Early Comb Ware, CW2 = Typical Comb Ware, PCW = Pit-Comb Ware, Kie = Kierikki Ware. Sites A and B are located in Environmental Zone 3. Other sites are located in Environmental Zone 2. Redrawn by the author after originals by V. Deckwirth & N. Närväinen, P. Halinen, K. Nordqvist, O. Seitsonen, and S. Seitsonen in reports Halien et al. 1999, Halinen & Mökkönen 2004 and Mökkönen et al. 2005. 138

urements made from the dip point have indicated (e.g., Alexander 2000: 38; Halinen et al. 2002; Kankaanpää 2002). The errors between measurements made on the surface and those made during excavation range between 5 and 30 percent in scale. Hence, the size of the pit measured on the surface provides a rough approximation of the pithouse s original size. Variation in size Fig. 5. Site classification according to housepit sizes. older than Typical Comb Ware may at least partly be due to natural causes, such as transgressive lakes. 10 In the Republic of Karelia the largest housepits date to ca. 3300 2500 cal BC (Zhul nikov 2003: 101). This is fairly synchronous with the occurrence of the largest housepits in Finland (Núñez & Okkonen 2005; Mökkönen 2008; Vaneeckhout 2009). Therefore, as a hypothesis, it is assumed that similar developments took place quite contemporaneously also on the Karelian Isthmus. HOUSEPIT SIZE VARIATION IN THE VUOKSI RIVER VALLEY In boreal forest that has never been cultivated or otherwise subjected to earth moving or ploughing, Stone Age pit structures are still clearly discernible on the surface. In such circumstances, size as well as certain structural features like embankments and corridors may be defined without any excavations (Figs. 3 and 4). Consequently, the housepits discussed in this study are examined on the basis of size measurements and structural details derived mainly from survey data. In order to study chronological changes in size, the housepit sites are divided into seven size classes (Fig. 5). The housepits are measured from the dip point of the housepit s edge, which has been shown to correspond closely to the original house wall (Pesonen 2002: 27). The ratio of pit area to original floor area naturally varies with the type of soil into which the pithouse was excavated and with the depth of the original pit, but as a rule the floor areas of the excavated housepits have turned out to be somewhat smaller than the meas- The smallest housepits included in the data are under 20 m 2 and the largest approximately 100 m 2 in size (Fig. 6). The average housepit size is 42.4 m 2, while the median is slightly smaller at 35 m 2. The longest housepit, from the Räisälä Sylijärvi SW2 site, measures 19.4 meters in length and 4.5 meters in width (see Fig 4G). Plotting the size class against the number of housepits at individual sites gives interesting results (Table 1, Fig. 7). There is a tendency for the co-occurrence of a low number of housepits and minimal variation in housepit size at the same site (size classes 1, 2 and 3). The largest number of housepits is found among the sites in size classes 5 and 7, which also represent maximal variation in pit size. Among these sites the number of housepits peaks at the Räisälä Valkialampi site with 12 housepits 11. A closer examination of the sites with four or more housepits reveals that most of the housepits are relatively small (Fig. 8). In three cases the largest housepits measure over 80 m 2 at the Räisälä Peltola A site ca. 100 m2, at the Räisälä Fig. 6. Housepit sizes on the Kaukola Räisälä region. 139

Fig. 7. The number of housepits per site divided into Size Classes. Peltola C site ca. 82 m 2, and at the Räisälä Valkialampi site ca. 100 m 2. The latter is associated with a fragment of a completely polished slate arrowhead dating to the late Middle Neolithic Late Neolithic (Halinen et al. 1999). In five out of the eight cases, the sites include one or two housepits that are clearly larger than the others (Fig. 8). However, at two sites the size range is much more limited. At the Räisälä Juoksemajärvi Westend site all the housepits are smaller than 30 m 2, while at the Räisälä Peltola B site all the housepits are between 60 and 80 m 2 in size. Chronological trends in size and shape As noted above, pottery typology is the only relatively practicable device available for chronological studies in the research area. This sec- Table 1. Housepits of different sizes from the research area. # Size m 2 Sites HPs HPs/site Shapes Details Ceramics Other dating 1 <25 6 25 % 13 15 % 2.6 (1/4) ov, rd - CW1, CW2, CPW Mesolithic (finds + C14) 2 25 50 3 5 1.7 (1/2) ov, re, rd emb CW2, KIE, KIE/PÖL - 13 % 6 % 3 >50 6 11 1.8 (1/4) ob, ov ent CW2, CW3, ORG, TXT - 25 % 13 % 4 <50 - - - - - - - 5 >25 5 30 6.0 (2/11) ov, rd, re ent CW2, CW3, unidentified 1) - 21 % 36 % 6 <25+>50 1 4 % 2 7 % 2 (-/-) re, rd emb - Mesolithic (finds + C14) 7 all sizes 3 13 % 23 27 % 7.7 (3/12) ov, re, rd emb, ent CW2, KIE Mesolithic (?) #= Size Class; HPs= housepits; HPs/Site (minimum/maximum); Shapes: ob= oblong, ov= oval, re= rectangular, rd= round; Details (structural details): emb= embankment, ent= entrances; Ceramics: CW1= Early Comb Ware, CPW= Comb-Pit Ware, ORG= Organic tempered ware, TXT= Textile ceramics, 1) coarse tempered ware (CW) 140

tion discusses different kinds of housepits and housepit sites and their association with different ceramics types. Looking at the ceramics from the sites in the different size classes (Table 1), the first thing to note is that Typical Comb Ware is found in sites with both larger and smaller housepits. More detailed information is provided by a scatter diagram showing the size distribution of individual housepits and pottery types associated with them. This diagram reveals that the smallest housepits are, indeed, associated with Early Comb Ware (Fig. 9, Table 2). It also shows that the housepits associated with Typical Comb Ware have a wide size range. A parallel, even slightly wider size range is to be found in the housepits containing Late Comb Ware, with which the largest housepits are associated. The late Middle Neolithic/Late Neolithic dating for the largest housepits would be even more evident if the 10 x 10 m housepit with a polished slate arrow-head from the Räisälä Valkialampi site were included in the scatter diagram. Not only the size but also the shape of the housepits changed over time. Table 2 illustrates the shift towards a larger average housepit size as Fig. 8. Size distribution of housepit sizes at individual sites with four or more housepits. Sites are grouped according to the find material. 141

Table 2. Sizes and shapes of the housepits associated with different ceramic assemblages. The housepits associated with Early Comb Ware are those at the Räisälä Juoksemajärvi Westend site. The Mesolithic examples from the Rupunkangas 3 site date to ca. 8200 7600 cal BC (Mökkönen et al. 2007). Average L/W Range L/W Average m 2 Range m 2 N Mesolithic 1.27 1.00 1.50 42.63 20 61 2 Early Comb Ware 1.18 1.00 1.45 18.80 16 23 4 Typical Comb Ware 1.24 1.00 1.50 47.32 27 80 21 Late Comb Ware 1) 1.33 1.07 1.85 51.41 31 78 7 Typical Comb Ware + Pitted Ware 1.50 1.07 2.00 46.83 18 70 3 Typical Comb Ware + Kierikki 1.24 1.00 1.50 41.00 24 56 4 Late Neolithic/Textile ceramics (?) 2) 1.57-56.40-1 Polished slate arrow head 1.00-100.00-1 Abbreviations: L= length, W= width. 1)- 2) ) houses with entrances included, slightly exaggerating the maximum area: 1) 2 examples, 2) 1 example. Fig. 9. Distribution of housepit sizes associated with different pottery styles. The sites included in the scatter plot diagram in Kaukola municipality are following: Rupunkangas 1 and Rupunkangas 2. Site included from Kirvu parish: Harjula. Sites included from Räisälä municipality: Juoksemajärvi West, Juoksemajärvi Westend, Kankaala, Mäenala, Mäenrinne 1, Mömmönsalmi 1, Peltola B, Peltola C, Repokorpi and Seppälä 4. well as towards a more oblong shape that takes place over the course of the Neolithic. However, the longitudinal growth as well as the areal increase connected with Late Comb Ware as seen in table 2 is in part a consequence of the occasionally appearing entrances/antechambers. A partly excavated housepit from the same site is associated with Early Comb Ware or coarse Typical Comb Ware. The Räisälä Juoksemajärvi Westend site is in Size Class 1. Another excavated site with Typical Comb Ware is the Räisälä Peltola C site, which dates to 3750 3650 cal BC (Halinen & Mökkönen, in this volume). The housepits there are larger, varying from 20 40 m² (5 cases) to over 80 m² (1 case). These are the only two excavated sites but, in any case, the results are in accordance with the observations drawn from the survey data. The data shows that the development of pithouses on the Karelian Isthmus is rather analogous to the development observed in Finland and the Republic of Karelia. Housepits associated with Early Comb Ware are lacking in Finland, but this could be at least partly caused by the lacustrine transgression of Ancient Lake Saimaa. The largest housepits most probably occur in all areas fairly 142

concurrently, i.e., roughly from 3300 to 2500 cal BC. The increase in pithouse size, however, began in the research area already at the sites associated with Typical Comb Ware (ca. 4000 3400 cal BC). The sites with the highest number of housepits are also those with Typical Comb Ware. HOUSEPITS AND THE ENVIRONMENT As described above, there is a certain regularity regarding the housepits size and shape and the ceramics associated with them. This chapter focuses on site location with respect to the environment. The housepit sites are analysed with the help of wider environmental zones and a topographic shelter index, which describes the sites degree of vulnerability to winds (see also Halinen & Mökkönen, in this volume). The housepits located in one cluster in the outer archipelago in the Rupunkangas area are excluded in detailed analyses. These sites are exceptional in environmental sense and are interpereted as seasonally occupied hunting stations (Mökkönen et al. 2007, see also Halinen & Mökkönen, in this volume). The geographical distribution of housepits over environmental zones The environmental zones used in this study are defined on the basis of changes in the ratio between land and water areas in the hydrological reconstruction of the period preceding the formation of the River Neva (ca. 1350 cal BC). The environmental zones are the following: Zone 1 Outer archipelago and open water area, Zone 2 Inner archipelago and the mouths of bays, and Zone 3 Shores of narrow fjord-like bays and inland areas. 12 These zones are roughly equivalent to geographical and ecological zones (Fig. 10). The overall distribution of housepits within environmental zones indicates that most of the housepits were located in the inner archipelago and around the mouths of bays (Zone 2). However, there are some well-defined housepit clusters (Fig. 10). At first glance, it appears that the densest clusters consist of housepits of all sizes, but closer scrutiny shows that there is a noticeable change in site location with relation to size class. Beginning with Size Class 1 (<25 m 2 ), the only cluster in the inland zone (Zone 3) at the head of a bay in Lake Juoksemajärvi has only housepits of less than 25 m 2 size (Fig. 10A). As far as waterways are concerned, this cluster was located at a dead end. On the other hand, these sites are situated adjacent to a nearly ten kilometre long moraine esker, which is the easiest natural route through the forests for one preferring to use land routes. The sites with housepits of widely varying size, i.e., Size Classes 5 and 7, lie among the clusters in the Räisälä inner archipelago and in the Papinkangas area (Zone 2). These clusters lay in the inner archipelago characterized by narrow stretches of water speckled by islands of various sizes. Particularly the Räisälä inner archipelago is characterised by varying small-scale topography. During the Ancient Lake Ladoga Phase these sites were located by waterways with easy access in all directions. These sites are clearly attainable by water, whether by boat during ice-free periods or by sledge when the lake was frozen over. Also, suitable land routes provided by moraine eskers are found nearby. The sites with only large housepits, i.e., Size Class 3 with housepits over 50 m 2 in size, have a different distribution than other housepit sites (Fig. 10A). They are located on the outer fringes of the densest housepit clusters or in areas devoid of other housepits, mainly in Zone 2. The general distribution of the large housepits is environmentally, in wider sense, more terrestrially oriented than most of the housepit sites. During the Ancient Lake Ladoga Phase these sites were located by waterways in areas with a monotypic topography, often at the heads of capes by narrow straits. These particular locations were also of great strategic significance if one wished to keep an eye on the best water routes of the area. Site placement and protection against the wind The topographic shelter index describes a site s topographic location in terms of how well-sheltered or vulnerable to the winds the locus is (see Halinen & Mökkönen, in this volume). This index is composed of three variables regarding the degree of vulnerability from the direction of (1) the background of the site, (2) the shoreline, i.e., the rate of exposure to the open sea, and (3) the immediate environment as a whole, i.e., the topographic location of the site as a protected or vulnerable locus. Each variable is evaluated on a three-digit scale where the number 3 stands for maximal and number 1 for minimal 143

Table 3. Topographic shelter index values at the sites with and without housepits. Environmental zones: Zone 1: Outer archipelago and open water area, Zone 2: Inner archipelago and the mouth of bays, and Zone 3: Shores of narrow fjord-like bays and inland. N min max mean median All sites 135 1.3 3.0 2.05 1.82 Without housepits 111 1.3 3.0 2.02 1.82 With housepits 24 1.3 3.0 2.32 2.12 With housepits in Zone 1 5 1.3 2.3 1.73 1.33 With housepits in Zone 2 14 1.7 3.0 2.19 2.15 With housepits in Zone 3 5 1.7 3.0 2.38 2.30 Fig. 10. The distribution of housepits over environmental zones: A. The distribution of different sized housepits, B. Current water system at the research area, C. Number of housepits per site. Areas cited in the text: 1. Lake Juoksemajärvi area, 2. Rupunkangas area, 3. Räisälä inner archipelago, 4. Riukjärvi-Piiskunsalmi area, and 5. Papinkangas area. 144

shelter. The higher the index value, the better sheltered the site. 13 In general, housepit sites are better sheltered than archaeological sites without housepits (Table 3, see Halinen & Mökkönen, in this volume). As an exception, the housepit sites located in the outer archipelago (Zone 1) are located in very vulnerable places. The sites in the Rupunkangas area, presumed to be permanent campsites used for exploiting seasonal resources of the outer archipelago (Mökkönen et al. 2007), are not included in this section. The variation in the degree of vulnerability to winds of housepit sites is considerable (Tables 3 and 4). The sites in Size Class 1 (<25 m 2 ) have remarkably high values in the topographic shelter index. High values are observed also at the sites with only large housepits (Size Class 3, >50 m 2 ), while the sites with housepits of various sizes (size classes 5 and 7) have shelter index values in approximately the same range as the values of sites without housepits. Comparing size classes 1 and 3 with the other sites with housepits and sites without housepits, it is clear that the sites with only small and those with only large housepits are located in well-sheltered places. Summary of observations Before discussing the phenomena relating to housepits, the observations concerning the material need to be summed up. In short, the observations are the following: Pithouse size tends to grow over the course of the Neolithic in the Vuoksi River Valley. However, smaller pithouses are present all the time. The hypothesis that the evolution of pithouses in the study area is similar to and coeval with pithouse evolution in the Republic of Karelia and Finland is partly supported by the data. At least the largest housepits are associated with Middle Neolithic (4000 2300 cal BC) material. However, the expansion of the pithouse sizes begins already during the early Middle Neolithic on sites associated with Typical Comb Ware and dating to ca. 4000 3400 cal BC. The sites associated with Typical Comb Ware also have the largest number of housepits per site. The smallest housepits are located in Environmental Zone 3, in highly sheltered loci with better connections to land routes than to water routes. These sites are associated with Early Comb Ware, Pitted Ware, and Typical Comb Ware. With respect to other areas, both the number of housepits per site and the variation in the size of housepits within a site reach their maximum in the Räisälä inner archipelago. There, the sites are clearly attainable by water and poorly sheltered. The environment of the area was a patchy one a labyrinth of land and water with varying topography. With respect to site location and immediate environment, these sites have much in common with sites without housepits, which were most probably used for non-winter habitation. Most of the sites are associated with Typical Comb Ware, although Kierikki Ware, Late Comb Ware and Pit-Comb/Comb-pit Ware are also present at some sites. The sites with exclusively large housepits were located by the most strategic waterways, often by narrow straits. These sites are relatively well-sheltered. Regionally, these sites are located differently from the other housepits. The materials associated with these sites are very similar to those associated with smaller housepits, i.e., Typical Comb Ware, Late Comb Ware and Kierikki Ware. DISCUSSION The changes in site placement associated with different size classes of housepits provide grounds for arguing that there is also a change in the conditions defining the nature of occupation. These changes could have something to do with changes in subsistence strategies, technological innovations, social organization, as well as in other cultural concepts defining the settlement (see, e.g., Rafferty 1985; Gilman 1987; Binford 1990; Kelly 1992; Ames 1994; 1996; 2004). This section discusses the possible explanations for the recognized changes in the housepit sites. Firstly, the limitations of the data are discussed. Then aspects of the ethnographic data on housepits concerning settlement patterns and subsistence are presented and the causes and effects of sedentism are briefly discussed. The discussion continues with an examination of logistical mobility, which is important for the question of the degree of sedentism. The housepit sites degree of vulnerability to wind has advantages and disadvantages with respect to the season of occupation. This is discussed regarding the sites themselves and the pithouses constructed in certain environments. Lastly, the questions of what the observed 145

Table 4. Topographic shelter index values of housepit sites in various size classes. The sites located on the former outer archipelago on the Rupunkangas area are excluded. Number 3 stands for the best shelter against wind and number 1 for the worst. Immediate environment BG WA TL Avg. Sites Size class 1 1 3 3 2.30 Juoksemajärvi West 2 3 1 2.00 Juoksemajärvi 3 3 3 3.00 Juoksemajärvi Westend 1 3 3 2.30 Siirilahti 2 3 3 2.70 Portinharju 3 3 3 3.00 Mäenrinne 2 Avg. 2.00 3.00 2.67 2.55 Median 2.00 3.00 3.00 2.50 Size class 2 3 2 2 2.30 Mäntylinna Avg. 3.00 2.00 2.00 2.30 Median - - - - Size class 3 3 3 1 2.30 Harjula 2 3 2 2.30 Sylijärvi SW 2 1 3 1 1.70 Repokorpi 2 3 1 2.00 Peltola B 2 2 3 2.30 Kankaala Avg. 2.00 2.80 1.60 2.12 Median 2.00 3.00 1.00 2.30 Size class 5 1 3 1 1.70 Mäenala 3 3 3 3.00 Mäenrinne 1 2 3 1 2.00 Peltola A 2 3 1 2.00 Peltola C 1 2 2 1.70 Mömmönsalmi 1 Avg. 1.80 2.80 1.60 2.08 Median 2.00 3.00 1.00 2.00 Size class 7 2 2 3 2.30 Valkialampi 1 3 1 1.70 Seppälä 4 Avg. 1.50 2.50 2.00 2.00 Median 1.50 2.50 2.00 2.00 Immediate environment: BG= background, WA= water areas, TL= topographic location changes indicate and what might have caused the changes are presented briefly. Limitations of the data The survey data has certain limitations. The distribution of known housepits is, naturally, affected by the obliteration of sites by recent agriculture. This is especially true of the northern part of the research area. However, the presence of housepit concentrations is not merely an artefact of preservation, nor of the survey methods used (see Lavento et al. 2006; Mökkönen et al. 2006). Consequently, I believe that the housepit clusters do in fact provide a rough approximation of past reality. The long occupation periods of the sites and the re-use of the pit structures over a long timespan (see Mökkönen et al. 2007) make it difficult to understand the chronology of the area. Another weakness of the data is that the possible contemporaneity of the housepits has not been established: were the largest and the smallest pithouses at one site contemporaneous, and how many pithouses in one cluster were occupied at the same time? Likewise, the chronological correlation of the housepits and the ceramics found at individual sites during survey is not self-evident. None of the data is totally objective. The measurements and classifications could have been done differently. In the data the measurements of the housepits are almost exclusively the work of a field crew of less than ten persons in the Kaukola Räisälä Project and are fully comparable. On the other hand, the topographic shelter index (see also Halinen & Mökkönen, in this volume) used in this study is highly subjective since the used variables are not based on actual measurements. 146

The long occupation periods of the sites and the re-use of structures pose problems when trying to associate finds with visible structures without a larger excavation. Therefore, it is rather unexpected that the data exhibits so obvious clustering. This is not a coincidence, since the various combinations of certain kinds of housepit sites and pottery types have different geographical distributions and display a preference for different environmental settings. Hence, it is highly probable that the smallest housepits are, indeed, associated with older pottery types (Early Comb Ware, Typical Comb Ware and Pitted Ware), while the sites with larger housepits lack the Early Neolithic component. The final weakness of the data is that this study focuses only on housepit sites. Due to difficulties with dating, the sites without housepits could not be used. Housepits, subsistence, and settlement patterns According to the Ethnographic Atlas by Murdock (1967) there are three conditions that are nearly always present when semi-subterranean pit structures are used as dwellings. These are (1) non-tropical, i.e., cold climate during the season of use, (2) reliance on stored food while the dwelling is inhabited and (3) permanent winter sites or a more sedentary settlement pattern15 (Gilman 1987: 541 3). Basically, pithouses are winter dwellings because the structures tend to be damp and susceptible to vermin infestation during warmer periods (Gilman 1987: 542 3). In the present research area soil moisture cannot have been a severe problem since the pithouses were mostly built on moraine soils which serve as natural drains for the dwellings. Following ethnographic data, hunter-gatherers building semi-subterranean dwelling structures practise residential mobility as follows: 2 % fully nomadic, 49 % semi-nomadic, 29 % semi-sedentary, and 20 % fully sedentary (Binford 1990: 124). Murdock s Ethnographic atlas (1967) shows that 77 % of pithouse dwellers are exclusively hunter-gatherers, although the remaining 23 % includes societies that practise either intensive or small-scale cultivation (Gilman 1987: 545 6). As concerns the Karelian Isthmus, it can be proposed that pithouses were inhabited at least during winter and that the settlement systems to which the housepits belong ranged between semisedentary and sedentary. This assumption, which is in accordance with the interpretations drawn from archaeological data in the nearby areas (e.g. Lundberg 1997; Katiskoski 2002; Kotivuori 2002; Norberg 2008: 177), will serve as a basis for this discussion. As indicated by the ethnographic sources mentioned above, pithouse dwellers have mostly been hunter-gatherers, but there are also examples of pithouse use by agriculturalists. Pithouses have been used at least for winter habitation, although a longer occupation period is also possible. As noted above, the mere presence of housepits does not allow us to draw any far-reaching conclusions concerning settlement patterns or subsistence (e.g. Binford 1990; Ames 1994: 219). Instead, changes in site location can be used to explain the nature of the variation. Sedentism The most commonly used definition of sedentism focuses on year-round habitation at one dwelling site by at least part of the population (see Rafferty 1985: 115; Kelly 1992; 2007: 148 9 with references). Still, sedentism is often thought to be a relative rather than an absolute condition, which means that settlement can become more sedentary, i.e., less mobile than before (Kelly 1992). Many archaeologists, including myself, use the categorization modified by Murdock (1967) in the Ethnographic Atlas. In this categorization societies are divided into fully nomadic, seminomadic, semi-sedentary, and fully sedentary. This is a more fine-grained categorization than the bipolar classification into mobile or sedentary. However, it is not easy to define archaeologically whether a society is semi- or fully sedentary. Changes in mobility are, undoubtedly, the results of long-term processes (Binford 1980; Marshall 2006: 158), but the finer details of the change process are often difficult to distinguish in the archaeological record. Sedentariness is a result of decreased residential mobility. Stationary year-round habitation at one dwelling site, however, does not decrease the overall mobility but reorganizes it into increasing logistical mobility (Binford 1980; Kelly 2007: 149). Sedentism has various causes as well as various consequences, but it is not always easy to point out which category a phenomenon belongs 147

to. Causes of sedentism include abundant food resources that encourage reduced mobility (the pull hypothesis), subsistence stress leading to an intensification of subsistence efforts and resulting in sedentism (the push hypothesis), group packing caused by population growth, situations where the cost of moving is high compared to the cost of remaining in the current camp, and a domino effect where one sedentary society encourages neighbouring groups to become sedentary (Rafferty 1985; Kelly 1992; 2007: 152, 160). Sedentism is followed by several effects: population growth, territoriality, more tightly controlled social boundaries, the intensification of long distance trade, and an increased reliance on stored food entailing an increased investment of labour (Kelly 1992; 2007: 152, 311 3). Numerous archaeological indicators have been used to define sedentism. Changes in the settlement pattern, new kinds of settlements, the presence of substantial houses (often rectangular), pottery, heavy artefacts, luxury goods, agriculture, cemeteries, ceremonial structures and storage are the most common ones (Rafferty 1985; Kelly 1992; 2007: 152, 160). However, the core elements in identifying sedentism are associated with the nature of the dwellings and the dwelling sites (Rafferty 1985: 128; Marshall 2006: 157). Proximal resources and mobility patterns The way people have settled the landscape relates to a number of cultural and ecological factors. The factors effectively setting the limits for cultural variation are the environmental conditions and the abundance and seasonality of food resources, the effects of which can, however, be reduced by storage capability and solutions relating to logistical mobility. In an aquatic environment the logistical mobility provided by boats also has a wide-ranging effect on the exploitable resources and thereby also on the areas to be settled (see e.g. Binford 1990; Ames 2002). Site placement in relation to larger environmental zones highlights which resources are the easiest to exploit from the site. In this respect the smallest Early Neolithic housepits, located in Environmental Zone 3, are ideally located for utilizing terrestrial resources by land routes. Yet, surprisingly, 30% of the identified mammal bones at the Räisälä Juoksemajärvi Westend site were seal (Halinen et al. 2008), which indicates that the winter resources located in the archipelago were also utilized at the site, most probably with the help of sledges. The densest clusters of housepit sites are, however, located in the archipelago. Sites situated in the inner archipelago and at the mouths of bays are located in an aquatic environment. It is important to notice that these sites are obviously meant to be reached by boats, which indicates logistical mobility based on water transport. As compared with pedestrian hunter-gatherers, the availability of water transport makes it possible to take advantage of resources spread over larger areas and to carry larger bulk over longer distances (Binford 1990; Ames 2002; 2003). Groups with logistical transportation based on boats tend to be more sedentary, and because of the effective transportation, the dwelling sites do not need to be located adjacent to certain individual resources, but rather in a central position with respect to all exploitable resources (Ames 2002). Other traits often connected with aquatically oriented hunter-gatherers as opposed to terrestrially oriented ones include a potential for higher population density, larger camps, longer residential moves, long distance trade, and the possibility of developing a more complex society (Binford 1990; Ames 1994; 2002; 2003). Of the said traits connected with aquatically oriented hunter-gathers, two the development of larger camps and the central position of the sites (most probably sedentary ones) with respect to all resources are also apparent in the primary data of this study. In the material from the Kaukola Räisälä region, it is not only the housepits that are distributed in the archipelago during the period when Typical Comb Ware was in use. A similar trend is also observed in the geographic distribution of Typical Comb Ware and other pottery styles of comparable or younger ages. These are evenly distributed in various dwelling sites over the archipelago, while the older pottery styles have a more terrestrially oriented distribution (Halinen & Mökkönen, in this volume). This supports the hypothesis that logistical mobility based on boats arose at the time Typical Comb Ware was in use. The existence of water transport together with the village-like clustering of housepits in the inner archipelago most probably marks a decrease in residential mobility and an increase in logistical 148

mobility. As will be demonstrated below, these housepits sites also have other features suggesting occupation was not limited to the winter season. This can be interpreted as a sign of increasing sedentariness. On the windy capes or by the sheltering hills A site s environment gives a hint of the possible occupation season. The larger environmental zone in which the site is situated indicates both the resources most easily exploited from the site and the prevailing transportation used. The immediate environment of the site provides more site-specific information about conditions that were vital and significant for the people who once erected their dwelling in the location. One such condition is the degree of protection against wind, which is described here by the Topographic Shelter Index. There are advantages and disadvantages as regards a site s degree of protection against winds. In the case of winter-only sites, a well-sheltered locus may be presumed (see Lundberg 1997: 111 2). On the other hand, in the case of summer sites, other aspects such as the amount of insects and protection against forest fires direct the selection of dwelling sites (Jordan 2001:93). The topographic shelter index shows that the most terrestrially oriented sites are also the most sheltered ones. The sites located by Lake Juoksemajärvi are situated on NE-E facing shores. There are high shading eskers in the background, and therefore the sites do not receive much drying sunlight. Most probably the pit structures at the site were extremely vulnerable to dampness when the ground was not frozen. Thus, these sites, with the smallest housepits associated with Early Comb Ware, are perfect for winter habitation (see also Halinen et al. 2008: 259) but not very suitable for use during the more humid seasons. The larger sites, as regards both the number and the size of the housepits, are located in more poorly sheltered locations that seem to be more suitable for summer than winter habitation. It is noteworthy that windy locations have some advantages for a housepit dweller planning to live on the site also during warmer seasons. During other seasons than winter, pithouses tend to be damp, which is probably the main reason for not using pithouses as summer dwellings (Gilman 1987: 542 3). On windy locations the drying effect of wind may have kept the pit structures drier for longer periods and enabled longer occupation of the pithouses (see below). Also the storage of foods benefits from drying breezes, and during summer the stronger winds provide relief from the swarms of mosquitoes. The housepits located in the archipelago and by the water routes are typically located in a patchy environment or at the heads of capes in an otherwise monotypic environment. These sites are consequently in places that are not susceptible to forest fires, and additional protection against fire could have been easily gained by cutting down a few trees or by digging a trench. Thermal regulation of housepits The thermal regulation of a dwelling structure is always a compromise between thermal efficiency and thermal adjustability in relation to outdoor climatic conditions, since both of these cannot be maximized at the same time (Wilkins 2009). A pithouse with heavyweight structures, for example, a cladding of sod or earth, is thermally effective but poorly adjustable. There are, however, several technical options that can render the thermal regulation of a structure more effective. A small pithouse with only one door is good for keeping heat inside (thermally very effective), but at the same time it lacks the capability to react to variation in outdoor temperatures (poorly controllable). The best solution to finding a balance in this contradiction is to increase structural complexity. In a dwelling with heavyweight structure, this balance is acquired by increasing the number of closable openings, increasing the number of spaces at different levels, and creating separate means of heating and ventilation for each space. This increases the thermal flexibility and adjustability within the structure. Such an arrangement allows the occupant to regulate the thermal microclimate of the dwelling (Wilkins 2009). The housepit data used in this study includes both small and large housepits. The smallest ones undoubtedly represented high thermal efficiency and poor thermal control, while the largest ones especially those with antechambers/entrances most probably offered the occupants better thermal control. As concerns the smaller housepits (less than ca. 50 m² in size), controlling the pithouse s inner temperature with respect to outdoor 149

temperature was an impossible task, with few options for ventilation. A structure with limited means of thermal regulation can be made more or less adjustable by a judicious choice of location (see Wilkins 2007). Better thermal control of such a structure can be acquired through optimising ventilation, which can be achieved by placing the structure in a windy location. The adjustable ventilation allows the occupants to control the moisture balance of the structure. Combining the ideas of thermal regulation and the evolution of the housepits in the Kaukola Räisälä area produces interesting results. As interpreted above, the small pithouses associated with Early Neolithic ceramics and located in Environmental Zone 3 were situated in places very suitable for winter habitation. They were structures with high thermal efficiency and poor thermal control located in highly sheltered locations, which makes them ideal winter dwellings. The relocating of pithouses into the archipelago at the time Typical Comb Ware was in use makes sense from the point of thermal regulation. The locating of pithouses in windy places rendered them more thermally controllable through ventilation, and as a consequence the housepits microclimate (including moisture balance) could be made more suitable also for non-winter habitation. The unsheltered windy location may have rendered these types of dwelling structures more functional for longer or even year-round occupation. The sites with only large housepits (Size Class 3, > 50 m²) are located more terrestrially and in more sheltered places than the sites with both large and small housepits. Could the largest pithouses have been more amenable to thermal control than the small ones, and could these structures have been comfortably used year round without a strong wind to maintain thermal control through ventilation during non-freezing seasons? Is this the reason for the more sheltered and terrestrially oriented location of the sites with only large structures? What do the changes in housepits indicate A change in architecture is always a response to new requirements. The changes might be driven by modifications taking place in the subsistence base, residential mobility, or social relations (e.g. McGuire & Schiffer 1983). Whatever the reason for changes in architecture, the change often occurs with some time lag in relation to the emergence of the causes (McGuire & Schiffer 1983; Rafferty 1985: 130; Ames 2003: 64). The relation between the shape of a dwelling s ground plan and residential mobility, again following ethnographic data, shows rather clearly that circular shapes are more typical of nomadic groups and rectangular forms characterize the ground plans of buildings used by more sedentary groups (McGuire & Schiffer 1983: 284; Binford 1990: 122 3). In cases where settlement has become more sedentary, the change in house shape has usually occurred, if it has occurred at all, with some time lag after the advent of sedentariness (McGuire & Schiffer 1983; Rafferty 1985: 130). The sizes of individual houses, together with the overall distribution of house sizes, are data that enable us to draw conclusions about the society that built the structures. Beyond the nearly selfevident positive correlation between the size of the houses and the probable number of habitants, the co-existence of both larger and smaller houses is seen as evidence of an uneven distribution of wealth or social hierarchy (McGuire & Schiffer 1983: 282 3, 289 90; Fitzhugh 2003: 31 2; Prentiss et al. 2003; Ames 2004: 62 4). This interpretation naturally presupposes that the larger and smaller structures were in simultaneous use, a stipulation that cannot be confirmed through survey data. Although the presence of substantial houses does not directly indicate the termination of residential movements, in the case of the study area the larger pithouses, being built of non-transportable materials (most probably logs and sod/earth) with a larger labour investment than the smallest pithouses, are arguable related to low residential mobility (see, e.g., Gilman 1987). It has been suggested that rectangular dwellings are more flexible regarding variation in space as compared to dwellings with circular ground plans. They are easier to divide internally and to extend by adding new rooms (McGuire & Schiffer 1983: 285 6). In the Kaukola Räisälä region elongated rectangular ground plans increase together with larger size. There are several reasons for the adoption of larger sized dwellings. Larger houses with larger households are a response to 150

increased demands for labour, expanded storage, and a growing number of tasks required by food production (Ames 1996: 132 3; 2003: 27 8), at least on the Northwest Coast of North America. In the American Southwest the appearance of large pithouses is seen as a response to economic intensification and a manifestation of individual households as autonomous units of production (Wills 2001). What caused the changes? I have interpreted the changes in housepits and housepit sites as markers of an increasing degree of sedentism during the Neolithic. The increase of sedentism is likely to be coupled with several other changes. In his monograph on hunter-gatherer lifeways, Kelly writes: I take a reduction in residential mobility, eventually resulting in sedentism, to be the significant kick that sets dramatic socio-political changes in motion (Kelly 2007: 310). In Kelly s scenario, environments with a constant and reliable resource base encourage the development of sedentism, which is in turn followed by population growth, territoriality, more tightly controlled social boundaries, and an increased use of stored food. The last of these also requires more labour in order to succeed (Kelly 2007: 151, 310 3). Kelly s scenario of sedentism is shortened above into a list. This article argues for a growing degree of sedentism during the period when winter dwellings were erected in the inner archipelago at locations favouring longer than winter-only habitation. Therefore, it is worth revisiting the common elements in the data and Kelly s scenario. In the Kaukola Räisälä region the relocating of housepits in the archipelago in locations more suitable for summer than winter habitation marks the beginning of more permanent settlement as seen in the archaeological record. This shift is followed by an increase in the size and number of housepits, and also in their rectangular ground plans. Although there is no direct evidence for the increased storage of food, the growing size of the houses probably indicates larger households and, at the same time, an increase in labour and storage capacity. The location of sites with only largest-size housepits by the most strategic waterways may result from a growing need to oversee the traffic and, at the same time, to restrict the access of other groups to resources considered the local group s private property. This might be one manifestation of growing territoriality. Is this just a general development of habitation (see Kelly 2007: 159 60, 259, 313 9), or do the changes also allow other lines of interpretation? The Neolithic cultures in Finland, on the Karelian Isthmus, and in the Eastern European coniferous forest zone are usually labelled Subneolithic, i.e., cultures that have other traits in common with real Neolithic cultures but have not practised agriculture. However, recent studies in Estonia have proved that cereal cultivation was known there already in the beginning of (sub)neolithic Stone Age and that it became more common at the time Typical Comb Ware was in use (Kriiska 2003; Poska et al. 2004). Similarly, on the eastern shore of Lake Onega the first signs of cereal cultivation date to 3800 3700 cal BC (Vuorela et al. 2001). These signs are associated with Comb-Pit and Rhomb-Pit Ceramics and correspond chronologically to Typical Comb Ware. Although there is no evidence of Stone Age cereal cultivation from the Karelian Isthmus so far, the evidence from Lake Onega and Estonia raises the possibility, if not even the probability, that agricultural practices were known also on the Karelian Isthmus during the Typical Comb Ware period (see Mökkönen in press). The idea of cereal cultivation being involved in the process of the changes in housepits and housepit sites that commenced during the Typical Comb Ware Period is not impossible. Similar changes in site location have been recorded in Eastern Sweden. There, in southern Norrland, the relocation of sites from riverine estuaries to the archipelago began around 3400 cal BC. This process was simultaneous with the spread of agriculture, and therefore the changes in sites and site locations have been connected with the adoption of agriculture (Björk 2003; Björk & Larsson 2007). I suppose that the domino effect, in which one group adopting sedentism also encourages its neighbours to become sedentary (Kelly 2007: 152), could be one cause for the increase of sedentism in the research area. At the time Typical Comb Ware was in use, the long-distance trade networks supplying, e.g., Baltic Amber to the East European forest zone were activated for 151

the first time (Edgren 1999: 68 70; Carpelan 1999; Zvelebil 2006; Zhul nikov 2008). Societies already practising agriculture also took part in this network. Therefore, it is possible that the introduction of agriculture and the domino effect causing sedentism might have coincided in the sphere of this network system. SUMMARY AND CONCLUSIONS A number of distinct chronological changes may be observed taking place in housepits and housepit sites. The chronological development of housepits in the research area and some of the interpretations are summarized in Table 5. The housepits associated with the Early Neolithic (5100 4000 cal BC) are small in size. In the beginning of the Middle Neolithic (4000 2300 cal BC), characterized by Typical Comb Ware (4000 3400 cal BC), the size of pithouses grew. The increase in size mainly affected the largest pithouses, while the smallest ones remained roughly the same size as before. Concurrently, the number of housepits per site increased. The later development of pithouses during the Middle Neolithic and Late Neolithic is not very easy to follow. Larger housepits that occur together with smaller ones date most likely to the later part of the Middle Neolithic, i.e., presumably starting from ca. 3700 cal BC. The data does not allow more accurate dating. The occurrence of larger solitary pithouses is, likewise, a phenomenon that is difficult to date. I presume that they are younger than 3300 cal BC. This assumption is based on the general occurrence of the largest housepits in northern Fennoscandia (Norberg 2008), which mainly dates after cultural contact following the arrival of the Corded Ware Culture or other agrarian cultures (Mökkönen 2008). The limited number of finds associated with the solitary large housepits in the research area, however, also supports this date. The environmental location of housepit sites changed quite simultaneously with the changes observed in size and clustering. The small Early Neolithic housepits have terrestrially oriented catchment areas and their locations are extremely well-sheltered against winds, but the sites do not receive much sunlight. Controlling the moisture balance of pithouses in such locations in nonfreezing temperatures is challenging. Sites like this are ideal for winter-only habitation. In the beginning of the Middle Neolithic, housepit sites associated with Typical Comb Ware Table 5. The chronological development of housepits on the research area summarized. cal BC Period Ceramics Observations on housepits Interpretation 5100 4000 Early Neolithic Early Comb Ware Pit-Comb Ware Small housepits on highly sheltered locations by the mainland. Poor dwelling sites Winter-only habitation 4000 2300 Middle Neolithic Typical Comb Ware Late Comb Ware Comb-Pit Ceramics Asbestos tempered wares (Kierikki-Pöljä) organic tempered wares (not well-studided) 2300 1800 Late Neolithic Textile pottery organic tempered wares during moist seasons. 1st half: Housepit clusters are appearing to the archipelago, to the places vulnerable to the winds. The size of the largest housepits grows, and the number of housepits per site increases. These sites are aquatically oriented and made to be reached by boats. Good sites concerning fire safety and the control of houses moisture balance. 2nd half: The growing of housepit size peaks, the number of housepits decreases. Large solitary housepits are occuring at the sites, which are by the most prominent water routes but, at the same time, more terrestrially oriented than before. Not known Year-round habitation at the villages, logistical mobility based on boats Year-round habitation, in a villages, but also on large multifamily houses Not known 152

(4000 3400 cal BC) spread over the archipelago. At this time the number of housepits per site increased, and the size of the housepits started to grow. The sites in the archipelago are clearly intended to be reached by a boat, which indicates logistical mobility based on water transport. These sites are poorly sheltered against winds, a feature that favours non-winter habitation due to, among others, better control of the pithouses thermal and moisture balance through maximum ventilation. These sites are also protected from fire, which makes them safe investments. During the later Middle Neolithic, the large solitary housepits, which are probably younger than 3300 cal BC, are found in more terrestrially oriented sites. The larger pithouses, which often had entrances/antechambers, were thermally more easily controllable and therefore habitable in non-freezing temperatures without the maximal ventilation generated by a windy location. I suppose that Early Neolithic pithouses were built in winter villages and used as winter-only dwellings. In the beginning of the Middle Neolithic, settlement became more sedentary and even fully sedentary habitation may have been possible. During this time the winter dwellings, i.e., pithouses, began to be built in the archipelago at location that were more similar to summer dwelling sites. The other concurrent changes the growth of housepit size, the increasing number of housepits per site, logistical mobility obviously based on water transport, and other characteristics of the sites as well as cultural aspects favouring year-round habitation all speak for a growing degree of sedentism. I presume these sites were small villages inhabited over most of the year or even year-round. It would appear to me that that the degree of sedentism did not change during the latter part of the Middle Neolithic, at the time when the large and solitary housepits emerge. The number of housepits per site decreases and the habitation increasingly concentrates in single large, probably multifamily houses instead of several smaller pithouses. For me, the occurrence of large solitary houses is very similar as concerns both the concept of the dwelling structure and the dwelling site itself to the dispersed single farmstead settlements of the Corded Ware culture. In my opinion there are not one but several reasons for the changes seen in housepits and housepit sites. I am positive that the beginning of agriculture probably plays a role in this change, whether through contacts with agricultural groups encouraging the spread of sedentism or in the form of actual early small-scale agriculture. The growing degree of sedentism, presumably even a fully sedentary settlement pattern, arises during the Typical Comb Ware Period. This is indicated also by other cultural phenomena such as the intensification of long-distance trade during the early Middle Neolithic (Carpelan 1999; Edgren 1999: 68 70; Zvelebil 2006; Zhul nikov 2008), the emergence of a number of cemeteries with red ochre graves (e.g. Halinen 1999) and the appearance of village-like pithouse clusters. The question of the specific dates of these changes in settlement is impossible to answer without more excavated material and radiocarbon dates. Similar changes have been observed in a case-study carried out in eastern Finland. In the Lake Saimaa area around one hundred kilometres north of the Kaukola Räisälä region, the placement of housepits changes during the Typical Comb Ware period (ca. 4000 3400 cal BC). The older pithouses were built at the heads of bays in highly sheltered loci, while the younger pithouses were built on windy capes and islands. Afterwards, these very same unsheltered sites were frequently occupied by people using asbestostempered pottery (Kierikki and Pöljä Wares) and also building larger rectangular pithouses (Mökkönen 2002). My argument concerning the growing degree of sedentism is based on the changes observed in housepits and in the environmental settings of the housepit sites. Interestingly, the changes in the archaeological record following the colonization of the inner archipelago include traits shared with Kelly s (2007: 310 3) scenario concerning the development of sedentism. However, it is not known what really caused the change towards residentially more stationary settlement. The cause of sedentism could have been, as in Kelly s scenario, merely normal development within a society or a domino effect resulting from contacts with sedentary societies also practising agriculture. Whatever the cause, the change in housepit shape and size takes place quite simultaneous over large areas of Northern Fennoscandia. Likewise, the rather contemporaneous change in placing pithouses in locations vulnerable to winds takes place in at least two major lake regions. This 153

article puts forward a few ideas concerning this change in settlement, but at present this process is still poorly understood. ACKNOWLEDGEMENTS The archaeological data used in this article was mostly gathered during survey and excavation expeditions carried out in connection with the Kaukola Räisälä Project (2004 2005), financed by a Chancellor s research grant (University of Helsinki, Finland) with a number of minor expenses covered by co-operating partner institutions in St. Petersburg and the University of Helsinki Department of Archaeology. The writing of the article was funded by the Finnish Graduate School in Archaeology. I wish to thank the following persons for participating in Kaukola Räisälä Project: Mr. Dmitriy Gerasimov, M.A. (St. Petersburg), Mr. Stanislav Beslkij, M.A. (St. Petersburg), Dr. Sergey Lisitsyn (St. Petersburg), Mr. Boris Lych, M.A. (St. Petersburg), Ms. Maria Juškova, M.A. (St. Petersburg), Professor Mika Lavento (Helsinki), Dr. Petri Halinen (Helsinki), Mr. Kerkko Nordqist, M.A. (Helsinki), Mr. Oula Seitsonen, M.A. (Helsinki), Ms. Sanna Seitsonen (née Puttonen), B.A. (Helsinki) and Professor Aivar Kriiska (Tartu). Furthermore, I wish to thank Mr. Vadim Adel, M.A., for commenting on my chronological scheme for ceramics, Professor Peter Jordan for commenting on my presentation and manuscript on the subject in a post-graduate student meeting in Helsinki, February 2009, and Dr. Jarmo Kankaanpää for revising my English language and commenting on the manuscript. NOTES 1 For more information on early research see Uino 1997; 2003; Lavento et al. 2001; Huurre 2003; Nordqvist & Lavento 2008; Nordqvist et al. 2009; and Halinen & Mökkönen, in this volume; on recent research see Lavento et al. 2001; 2006; Gerasimov 2006; Mökkönen et al. 2006; Lavento 2008; Nordqvist et al. 2009; and Halinen & Mökkönen, in this volume. 2 The Kaukola-Räisälä Project is named after two former Finnish municipalities forming the core of the research area (now Russian Sevast janovo and Mel nikovo). The official name of the project is Subsistence strategies and changes of communities between 9000 1 BC: an archaeological intensive-investigation in the western part of Lake Ladoga, Karelian Isthmus. The project was carried out as a cooperative Finnish-Russian- Estonian effort. The field work surveys and small scale excavations was carried out in 2004 2005. The project was lead by professor Mika Lavento (University of Helsinki). The English versions of the survey and excavation reports are available at the University of Helsinki Department of Archaeology. 3 Some of the results have already been published. A number of these publications deal with the preliminary results of the surveys (Gerasimov et al. 2006; Lavento et al. 2006; Mökkönen et al. 2006) and others present the excavation results (Gerasimov et al. 2006; Mökkönen et al. 2007). See also Halinen & Mökkönen, in this volume). 4 The Finnish names are used because the most detailed and accurate maps available for use in the surveys were Finnish topographic maps from the 1930 s (in 1:20 000 scale). Russian names for sites discovered before the 21 st century may be found in Nordqvist et al. 2008 and for the sites discovered during 2004 2005 in Gerasimov 2006. 5 There are two sites, namely Räisälä Pitkäjärvi (see Seitsonen 2005) and Kaukola Kyöstälänharju, excavated during early 20 th century both of which are having possible or probable housepits. Due to inexact nature of the data these sites are excluded from this study. 6 It must be noted that in the prehistory of the East European coniferous forest zone, the Neolithic Stone Age is defined by the occurrence of pottery rather than agriculture and is therefore often labelled Subneolithic. 7 In the most southern part of the research area the water level has fluctuated more than in the northern part. 8 In this article the Finnish periodization of the Neolithic is used: Early Neolithic (5100 4000 cal BC), Middle Neolithic (4000 2300 cal BC) and Late Neolithic (2300 1800 cal BC) (see, e.g., Carpelan 2002). 9 This particular date Hela-465, 4870 ± 85 BP was published by Huurre (2003: 198 9, 512). This date is incorrectly referred in another article (Timofeev et al. 2004) to a sherd of Typical Comb Ware instead of Kierikki Ware. Another date of nearly comparative age derives from Kierikki Ware from the Inari Saamen museo Vuopaja site in North Finland (see Pesonen 2004). 10 For example, in the Lake Saimaa area the older housepits could be covered by flooded sediments during the transgressive Ancient Lake Saimaa prior to 154

the formation of the Vuoksi River, the latest outflow channel, ca. 4000 cal BC, or destroyed by cultivation (for discussion on the absence of housepits in certain areas see Pesonen 2002). 11 In this study the Räisälä Peltola site, with 15 housepits in total, is divided into three separate sites following the three housepit clusters (areas A, B, and C). These clusters are located side by side, but separated from each other by hillocks. The reason for this division, in addition to topographic features separating the clusters, is the fact that while two of the clusters were associated with Typical Comb Ware, one yielded different, notably coarse tempered ware. For a general map, see Fig. 4F. 12 The term Pitted Ware used here refers to both Pit- Comb Ware and Comb-Pit Ceramics. 13 This division follows a typological division of ecological and geographic zones used in the archipelago of the Gulf of Finland. Put simply, it relies on the ratio between land and water areas (see Hanhijärvi & Yliskylä-Peuralahti 2006: 9 11). In Zone 2 the inner archipelago and the mouths of bays the amount of water and land areas is roughly equal. In Zone 1 the outer archipelago and open water area water areas dominate. And, conversely, in the Zone 3 there is clearly more land than water (see also Halinen & Mökkönen, in this volume). 14 As an example, a site located on the point of a cape without any topographic shelter in the background and surrounded by vast areas of open water is located in a poorly sheltered locus and therefore has a topographic shelter index value of 1. 15 There are, however, a few exceptions. These are housepit sites located in ecologically and environmentally marginal environments, where the sites were used repeatedly as seasonal campsites (e.g. Renouf & Murray 1999; Smith 2003;). In this respect, the housepit sites in the outer archipelago posses a different character as compared to other housepit sites (see Mökkönen et al. 2007), and therefore the housepits clustered in the Rupunkangas area in Zone 1 seen as exceptions to the rule are excluded from this analysis. Even though the interpretation presented here concerning the duration of the occupation period of housepit sites in the outer archipelago were to prove invalid, these sites are, in terms of their environmental location, clearly something different from most housepit sites located in other environmental zones. REFERENCES Unpublished sources University of Helsinki, Department of Archaeology, Helsinki Halinen, P., Lavento, M., Timofeev, V. & Saksa, A. 1999. The Archaeological Survey of Kaukola and Räisälä in Karelian Isthmus. The 1999 Report. Report. Halinen, P. & Mökkönen, T. 2004. The archaeological Survey in Kaukola and Räisälä on the Karelian Isthmus. Kaukola Räisälä project. Report 2004. Report. Kotivuori, H. 2002. Alisen Kemijoen kivikautiset asumuspainanteet topografiseen havainnointiin ja aineistovertailuun perustuva asutuskuva. Licentiate thesis. Mökkönen, T., Seitsonen, O., Nordqvist, K. & Puttonen, S. 2005: The Archaeological Survey in Kaukola and Räisälä on the Karelian Isthmus. Kaukola Räisälä -project. Report 2005. Report. Literature Alexander, D. 2000. Pithouses on the Interior Plateau of British Columbia: ethnographic evidence and interpretation of the Keatley Creek Site. In B. Hayden (ed.), The Ancient Past of Keatley Creek. Volume II: Socioeconomy: 29 66. Archaeology Press, Burnaby, B.C. Ames, K.M. 1994. The Northwest Coast: complex hunter-gatherers, ecology, and social evolution. Annual Review of Anthropology 23: 209 29. Ames, K.M. 1996. Life in the Big House: household labor and dwelling size on the Northwest Coast. In G. Coupland & E.B. Banning (eds.), People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures: 131 50. Monographs in World archaeology 27. Prehistory Press, Wisconsin. Ames, K.M. 2002. Going by boat. The foragercollector continuum at sea. In B. Fitzhugh & J. Habu (eds.), Beyond Foraging and Collecting. Evolutionary Change in Hunter-gatherer Settlement Systems: 19 52. Kluwer/Plenum Press, New York. Ames, K.M. 2003. The Northwest Coast. Evolutionary Anthropology 12: 19 33. Ames, K.M. 2004. Tempo and scale in the evolution of social complexity in Western North America: four case studies. In T.R. Pauketat & D.D. 155

Loren (eds.), North American Archaeology: 56 78. Blackwell Publishing, Oxford. Asplund, H. 1997. Kemiön suurpitäjän esihistoria. Kemiön suurpitäjän historia 1: 213 82. Sagalundin museon kuntayhtymä, Kemiö. Asplund, H. 2004. Problems of pre-roman Iron Age radiocarbon dating an example from SW Finland. In P. Uino (ed.), Fenno-Ugri et Slavi 2002. Dating and Chronology: 9 14. Museoviraston arkeologian osaston julkaisuja 10. Museovirasto, Helsinki. Binford, L.R. 1980. Willow smoke and dogs tails: hunter-gatherer settlement systems and archaeological site formation. American Antiquity 45(1): 4 20. Binford, L.R. 1990. Mobility, housing, and environment: a comparative study. Journal of Anthropological Research 46(2): 119 52. Björk, N. 2003. The Neolithic coastal settlements cosmology and ideology in a Baltic Sea perspective. In C. Samuelsson & N. Ytterberg (eds.), Uniting Sea. Stone Age Societies in the Baltic Sea region: 20 42. Occasional Papers in Archaeology 33. Uppsala universitet, Uppsala. Björk, N. & Larsson, F. 2007. Det Neolitiska samhället i Östrä Mellansverige. In N. Stenbäck (ed.), Stenåldern i Uppland, Uppdragsarkeologi och eftertanke. Vol. 1: Arkeologi E4 Uppland -studier: 67 96. Riksantivarieämbetet, Uppsala. Carpelan, C. 1999. Käännekohtia Suomen esihistoriassa aikavälillä 5100 1000 ekr. In P. Fogelberg (ed.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan: 249 80. Bidrag till kännedom av Finlands natur och folk 153. Finska vetenskaps-societeten, Helsinki. Carpelan, C. 2002. Esihistorian vuosiluvut, ajoitukset ja kronologia. In R. Grünthal (ed.), Ennen muinoin: 18 27. Tietolipas 180. Suomalaisen Kirjallisuuden Seura, Helsinki. Edgren, T. 1999. Käännekohtia Suomen kivikaudessa. In P. Fogelberg (ed.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan: 281 93. Bidrag till kännedom av Finlands natur och folk 135. Finska vetenskaps-societeten, Helsinki. Fitzhugh, B. 2003. The evolution of complex hunter-gatherers in the Kodiak archipelago. In J. Habu, J.M. Savelle, S. Koyama & H. Hongo (eds.), Hunter-Gatherers of the North Pacific Rim: 13 48. Senri Ethnological Studies 63. National Museum of Ethnology, Osaka. Gerasimov, D.V. 2006. Kamennyi vek Karelskogo peresheika v arkheologicheskom sobranii. MAZ RAN. Svod Arkheologicheskikh istochnikov Kunstkamery, Bypusk 1: 109 88. Kunstkamera, Sankt- Peterburg. Gerasimov D.V., Lavento M., Myokkonen T., Nordkvist K., Puttonen S., Seitsonen O. & Khalinen, P. 2006. Arkheologicheskie izyskaniya goda 2005 po kamennomu veku epokhe rannego metalla v severo-vostochnoi chasti Karel skogo peresheika. Radlovskie chteniya 2006: 247 53. Gilman, P. A. 1987. Architecture as artefact: pit structures and pueblos in the American Southwest. American Antiquity 52(3): 538 64. Halinen, P. 1999. Burial practices and the structure of societies during the Stone Age in Finland. In M. Huurre (ed.), Dig it all. Papers dedicated to Ari Siiriäinen: 173 9. The Finnish Antiquarian Society & The Archaeological Society of Finland, Helsinki. Halinen, P., Joensuu, J., Lavento, M. & Martio, L. 2002. House pit studies at Martinniemi in Kerimäki. In H. Ranta (ed.). Huts and Houses. Stone Age and Early Metal Age Buildings in Finland: 201 10. Finnish National Board of Antiquities, Helsinki. Halinen, P., Joensuu, J., Lavento, M., Martio, L. & Mökkönen, T. 2003. Kerimäen Raikuun Martinniemen tutkimukset 1998 1999. Sihti 5: 9 26. Halinen, P., Seitsonen, O., Seitsonen, S. & Nordqvist, K. 2008. Excavations at the Juoksemanjärvi Westend Stone Age dwelling site in 2002. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003: 235 65. Iskos 16. The Finnish Antiquarian Society, Helsinki. Halinen, P. & Mökkönen, T., in this volume. Between lake and sea Stone Age settlement by ancient Lake Ladoga on the Karelian Isthmus. Fennoscandia archaeologica XXVI. Hanhijärvi, J. & Yliskylä-Peuralahti, J. 2006. Suomen rannikkostrategia. Suomen ympäristö 10/2006. Ympäristöministeriö, Helsinki. Huurre, M. 2003. Viipurin läänin kivikausi. In M. Saarnisto (ed.), Viipurin läänin historia 1. Karjalan synty: 151 244. Karjalan kirjapaino, Lappeenranta. Jordan, P. 2001. Cultural landscapes in colonial Siberia: Khanty settlements of the sacred, the living and the dead. Landscapes 2(2): 83 105. Kankaanpää, J. 2002. The house pits at Kauvonkangas, Tervola. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Build- 156

ings in Finland: 65 77. Finnish National Board of Antiquities, Helsinki. Katiskoski, K. 2002. The semisubterranean dwelling at the Kärmelahti site in Puumala, Savo province, Eastern Finland. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland: 171 200. Finnish National Board of Antiquities, Helsinki. Kelly, R.L. 1992. Mobility/sedentism: concepts, archaeological measures, and effects. Annual Review of Anthropology 21: 43 66. Kelly, R.L. 2007. The Foraging Spectrum: Diversity in Hunter-gatherer Lifeways. Percheron Press, New York. Kosmenko, M.G. 2004: The chronology of the Stone Iron Ages of the Karelian Republic. In P. Uino (ed.), Fenno-Ugri et Slavi 2002. Dating and Chronology: 46 55. Museoviraston arkeologian osaston julkaisuja 10. Museovirasto, Helsinki. Kriiska, A. 2001. Stone Age Settlement and Economic Processes in the Estonian Coastal Area and Islands. University of Helsinki, Helsinki. Available at <http://ethesis.helsinki.fi/julkaisut/ hum/kultt/vk/kriiska/tekstid/01.html> Kriiska, A. 2003. From hunter-fisher-gatherer to farmer Changes in the Neolithic economy and settlement on Estonian territory. Archeologia Lituana 4: 11 26. Lang, V. & Kriiska, A. 2001. Eesti esiaja periodiseering ja kronoloogia. Eesti Arheoloogia Ajakiri 5/2: 83 109. Lavento, M. 2001. Textile Ceramics in Finland and on the Karelian Isthmus. Nine Variations and Fugue on a Theme of C.F. Meinander. Suomen Muinaismuistoyhdistyksen Aikakauskirja 109. The Finnish Antiquarian Society, Helsinki. Lavento, M. 2008. Archaeological research in the Saimaa district and in the Karelian Isthmus in 1992 1999. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003. Iskos 16: 26 45. The Finnish Antiquarian Society, Helsinki. Lavento, M., Halinen, P., Timofeev, V., Gerasimov, D. & Saksa, A. 2001. An Archaeological field survey of Stone Age and Early Metal Period settlement at Kaukola (Sevastyanovo) and Räisälä (Melnikovo) on the Karelian Isthmus in 1999. Fennoscandia archaeologica XVIII: 3 25. Errata: Appendix 1 (sites) in Fennoscandia archaeologica XIX: 67 73. Lavento, M., Halinen, P. & Mökkönen, T. 2006. Subsistence strategies and changes of communities between 9000 1 BC: an archaeological intensiveinvestigation in the western part of Lake Ladoga, on the Karelian Isthmus (Kaukola Räisälä project). In V.-P. Herva (ed.), People, Material Culture and Environment in the North. Proceedings of the 22nd Nordic Archaeological Conference, University of Oulu, 18-23 August 2004: 120 30. Studia Humaniora Ouluensia 1. University of Oulu, Oulu. Lundberg, Å. 1997. Vinterbyar. Ett Bandsamhälles territorier i Norrlands inland, 4500 2500 f. Kr. Studia Archaeologica Universitatis Umensis 8. Umeå universitet, Umeå. Marshall, Y. 2006. Introduction: adopting a sedentary lifeway. Y. Marshall (ed.), Sedentism in Nonagricultural Societies. World Archaeology 38(2): 153 63. Routledge, Oxon. McGuire, R. & Schiffer, M. 1983. A theory of architectural design. Journal of Anthropological Archaeology 2: 277 303. Mökkönen, T, 2002. Chronological variation in the locations of hunter-gatherer occupation sites vis-à-vis the environment. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland: 53 64. Finnish National Board of Antiquities, Helsinki. Mökkönen, T. 2008. A Review of Neolithic multi-room housepits as seen from the Meskäärtty site in Virolahti parish, extreme south-eastern Finland. Estonian Journal of Archaeology 12(2): 114 51. Mökkönen, T. in press. Kivikautinen maanviljely Suomessa (with English summary Neolithic cereal cultivation in Finland ). To be published in Suomen Museo 2009. Mökkönen, T., Lavento, M. & Halinen, P. 2006. Neolithic sites in various environments of the Karelian Isthmus. Kaukola Räisälä -project s survey 2004. In A.N. Kirpichnikov, E.N. Nosov & A.I. Saksa (eds.), Slavs, Finns and Ougriens. The Zones of Contacts and Cooperation. The Papers of Russian-Finnish Symposium on the Problems of Archaeology and History, Puškinskije Gory 2004, October, 7.-10.: 112 29. Institute for the History of Material Culture Russian Academy of Sciences, Saint-Petersburg. Mökkönen, T., Nordqvist, K. & Belskij, S. 2007. The Rupunkangas 1A site in the archipelago of ancient Lake Ladoga: a housepit with several rebuilding phases. Fennoscandia archaeologica XXIV: 3 28. Murdock, G.P. 1967. Ethnographic atlas. University of Pittsburgh Press, Pittsburgh. Norberg, E. 2008. Boplatsvallen som bostad i 157

Norrbottens kustland 5000 till 2000 före tideräckning. En studie av kontinuitet och förändringar. Studia Archaeologica Universitatis Umensis 23. Umeå univeristet, Umeå. Nordqvist, K. & Lavento, M. 2008. Archaeological survey in Kaukola and Räisälä in 1999 and a study of environmental settings of the Stone Age dwelling sites in the area. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003: 140 63. Iskos 16. The Finnish Antiquarian Society, Helsinki. Nordqvist, K., Seitsonen, O. & Uino, P. 2008. Appendix 1. Stone Age and Early Metal Period sites in the studied municipalities. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003: 291 328. Iskos 16. The Finnish Antiquarian Society, Helsinki. Nordqvist, K., Seitsonen, O. & Lavento, M. 2009. Waterways and the Stone Age and Early Metal Period studies on the Karelian Isthmus The pre- World War II studies and research carried out by the University of Helsinki in 1998 2006. Quaternary International 203: 25 32. Núñez, M. & Okkonen, J. 2005. Humanizing of north Ostrobothnian landscape during the 4th and 3rd millennia BC. Journal of Nordic Archaeological Science 15: 25 38. Núñez, M. & Uino, P. 1997. Dwellings and related structures in prehistoric mainland Finland. In O. Kyhlberg (ed.), Hus och tomt i Norden under förhistorisk tid: 133 52. Bebyggelsehistorisk tidskrift 33. Stockholms universitet, Stockholm. Ojanlatva, E. & Alakärppä, J. 2002. Interpretation of the Peurasuo house pit in Oulu. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland: 109 22. Finnish National Board of Antiquities, Helsinki. Pesonen, P. 2002. Semisubterranean houses in Finland a review. In H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland: 9 41. Finnish National Board of Antiquities, Helsinki. Pesonen, P. 2004. Neolithic pots and ceramics chronology AMS-datings of Middle and Late Neolithic ceramics in Finland. In P. Uino (ed.), Fenno-Ugri et Slavi 2002. Dating and Chronology: 87 97. Museoviraston arkeologian osaston julkaisuja 10. Museovirasto, Helsinki. Pienzonka, H. 2008. Neue AMS-Daten zur frühneolithischen Keramikentwicklung in der nordosteuropäischen Waldzone. Estonian Journal of Archaeology 12(2): 67 113. Poska, A., Saarse, L. & Veski, S. 2004. Reflections of pre- and early-agrarian human impact in the pollen diagrams of Estonia. Palaeogeography, Palaeoclimatology, Palaeoecology 209: 37 50. Prentiss, W.C., Lenert, M., Foor, T.A., Goodale, N.B. & Schlegel, T. 2003. Calibrated radiocarbon dating at Keatley Creek: the chronology of occupation at a complex hunter-gatherer village. American Antiquity 68(4): 719 735. Rafferty, J.E. 1985. The archaeological record on sedentariness: recognition, development, and implications. In M.B. Schiffer (ed.), Advances in Archaeological Method and Theory 8: 113 56. Academic Press, New York. Renouf, M.A.P. & Murray, M.S. 1999. Two winter dwellings at Phillip s Garden, a Dorset site in Northwestern Newfoundland. Arctic Anthropology 36(1 2): 118 32. Saarnisto, M. 2003. Karjalan geologia. Karjalan luonnonmaiseman synty. In M. Saarnisto (ed.), Viipurin läänin historia 1. Karjalan synty: 21 80. Karjalan kirjapaino, Lappeenranta. Saarnisto, M. 2008. Emergence history of the Karelian Isthmus. In K. Nordqvist (ed.), Karelian Isthmus Stone Age studies in 1998 2003: 128 39. Iskos 16. The Finnish Antiquarian Society, Helsinki. Saarnisto, M. & Grönlund, E. 1996. Shoreline displacement of Lake Ladoga new data from Kilpolansaari. Hydrobiologia 322: 205 15. Saarnisto, M. & Siiriäinen, A. 1970. Laatokan transgressioraja (Referat: Die Transgressionsgrenze des Ladogasees). Suomen Museo 77: 10 22. Seitsonen, O. 2005. Räisälä Pitkäjärvi revisited new interpretation of the dwelling remains. In P. Pesonen & T. Mökkönen (eds.), Arkeologipäivät 2005. Arkeologia ja kulttuuri & Uutta kivikauden tutkimuksessa: 138 45. Suomen Arkeologinen Seura, Helsinki. Smith, C. S. 2003. Hunter-gatherer mobility, storage, and houses in a marginal environment: an example from the mid-holocene of Wyoming. Journal of Anthropological Archaeology 22: 162 89. Timofeev, V.I., Zaitseva, G.I., Lavento, M., Dolukhanov, P. & Halinen, P. 2004. The radiocarbon datings of the Stone Age Early Metal Period on the Karelian Isthmus. Geochronometria 23: 93 9. Uino, P. 1997. Ancient Karelia. Archaeological Studies Muinais-Karjala. Arkeologisia tutkimuksia. Suomen Muinaismuistoyhdistyksen 158

Aikakauskirja 104. The Finnish Antiquarian Society, Helsinki. Uino, P. 2003. Karjalan arkeologiaa 150 vuotta. In M. Saarnisto (ed.), Viipurin läänin historia 1. Karjalan synty: 117 50. Karjalan kirjapaino, Lappeenranta. Vaneeckhout, S. 2008. Sedentism on the Finnish Northwest Coast: shoreline reduction and reduced mobility. Fennoscandia archaeologica XXV: 61 72. Vaneeckhout, S. 2009. 2500 years of social evolution in a northern Finnish Stone Age village. In T. Mökkönen & S.-L. Seppälä (eds.), Arkeologipäivät 2008. Sosiaaliarkeologiaa yhteisöt arkeologisen aineiston takana & Tutkimushistorian painolasti: 63 71. Suomen Arkeologinen Seura, Helsinki. Vuorela, I., Saarnisto, M., Lempiäinen, T. & Taavitsainen, J.-P. 2001. Stone Age to recent landuse history at Pegrema, northern Lake Onega, Russian Karelia. Vegetation History and Archaeobotany 10: 121 38. Wilkins, H. 2007. An investigation of the adaptive opportunity of rudimentary structures based on field experiments. Building & Environment 42: 3883 93. Wilkins, H. 2009. Transitional change in protobuildings: a quantitative study of thermal behaviour and its relationship with social functionality. Journal of Archaeological Science 36: 150 6. Wills, W.H. 2001. Pithouse architecture and the economics of household formation in the prehistoric American Southwest. Human Ecology 4: 477 500. Zhul nikov, A.M. 1999. Eneolit Karelii (pamyatniki s poristoi i asbestovoi keramikoi) (with summary: Aeneolithic Age of Karelia [the sites with organic and asbestos-tempered ceramics]). Karelskii nauchnyi tsentr RAN, Petrozavodsk. Zhul nikov, A.M. 2003. Drevnie zhilishcha Karelii. Karel skii gosudarstvennyi kraevedcheskii muzei, Petrozavodsk. Zhul nikov, A.M. 2005. Poseleniya epokhi rannego metalla Yugo-Zapadnogo Pri-belomor ya. Petrozavodskii gosudarstvennyi universitet & Karel skii gosudarstvennyi kraevedcheskii muzei. Petrozavodsk. Zhul nikov, A. 2008. Exchange of amber in northern Europe in the III millennium as a factor of social interactions. Estonian Journal of Archaeology 12(1): 3 15. Zvelebil, M. 2006: Mobility, contact, and exchange in the Baltic Sea basin 6000 2000 BC. Journal of Anthropological Archaeology 25: 178 92. 159

APPENDIX 1 Ceramics Municipality/site Survey 1) HPs site N Shp SC 1) # LxW D Features Kaukola Pontuksenhauta 1 1999 - - 8 re, ov/ rd ENA, EMB Kaukola Rupunkangas 1 2004 2) CW3, TXT, ORG 160 CW3, TXT, ORG, LNeo, ASB 7 1 2 3 4 5 6 7 8 7x10 7x8 6x7 5x5 4.5x4.5 5x5.5 4x4 2.5x2.5 0.8 1 0.8 1 0.6 0.8 ca. 0.6 0.4 0.35 0.3 0.35 1 ov 3 1 9,4x6 0.4 m ENA, EMB? Kaukola Rupunkangas 2 2004 - CW2, KIE, KIE/PÖL 2 rd 2 1 2 6x6 6x6 Kaukola Rupunkangas 3 2004 - - 2 rd, 6 12 4.5x4.5 re 6.5x10 Kaukola Pontuksenhauta 3 2004 - - 2 re 2 1 8x6 2 6x5 Kirvu Harjula 2004 3) - CW3 1 ov 3 1 6.5x9 w/ entrance 6.5x12 Räisälä Juoksemajärvi Westend Räisälä Juoksemajärvi Westend Räisälä Juoksemajärvi Westend 1999> CW1, CW2 CW1, CW2 4 ov 1 1 2 3 4 1999 - CW2, PitC 2 oval 1 1 2 1999 4) CW1, 4 rd, CW2? ov CW1, CW2?, VOL? 1 1 2 3 4 5 4x4 4.5 5 4x4 4.5 5 4x4 4.5 3x3 3x6 3x6 4x5.5 4x4 4x4.5 4x4.5 0.4 0.15 0.6 0.6 0.5 0.5 NDA - NDA - NDA - 0.2 0.3 NDA NDA NDA Räisälä Sylijärvi SW2 1999 - - 1 ob 3 1 4.5x19 ca. 0.4 - Räisälä Valkialampi 1999 - - 12 rd, ov 7 1 2 3 12 10x10 5x5 5x5; others ø 5 4 NDA - Räisälä Siirlahti 1999 - - 1 ov 1 5x4 m 0.4 - Räisälä Mäenala 1999 CW2 CW2 11 rd, - ov 5 1 2 3 4 5 6 7 8 9 10 11 5x7.5 5x7.5 5x6 7.5x6 5x7 6x9 5x6.5 7.5x6 5x6 6x7 4.5x6 all ca. 0.2 0.4 Räisälä Portinharju 1999 - CW? 1 ov 1 4x3,5 NDA - Räisälä Mäenrinne 1 2004 CW3 CW3 6 rd, ov, re 5 1 2 3 4 5 6 5x7 8.5x10.3 w/entrance 3x3.8 5.6x6 5x6.2 7.7x9.7 4x6 0.2 0.5 entrance 0.2 0.2 0.3 0.3 0.3 Räisälä Mäenrinne 2 2004 - - 1 rd 1 4.5x4 ca. 0.3 - Räisälä Peltola A 2004 - UnID d coarse ware (CW) - 5 ov 5 1 2 3 4 5 Räisälä Peltola B 2004 CW2 CW2 4 rd 3 1 2 3 4 5x6.2 5.2x6 5.8x6. 5.1x5.5 10x10 8.3x9.6 7.8x8 8x9 7.4x8.3 Räisälä Peltola C 2004 3) CW2 CW2 6 rd 5 8.2x10 5x5 7.5x7.5 5.5x6.5 5.2x5.2 0.3 0.2 0.2 0.2 0.45 0.45 0.3 0.35 0.3 0.3 0.2 0.4 0.2 0.3 EMB EMB ENA, NDA - ENA, EMB - -

Ceramics Municipality/site Survey 1) HPs site N Shp SC 1) # 5.2x5.2 LxW 0.3 D Features Räisälä Repokorpi 2004 CW2 CW2 1 ov 3 7x10.5 ca. 0.3 - Räisälä Seppälä 4 2004 CW2, CW2, KIE 3 ov 7 1 6x8 NDA - KIE 2 3 7x8 4x6 NDA NDA Räisälä Mömmönsalmi 1 2004 CW2 CW2 2 ov 5 5x6.5 ca. 0.3 EMB 7x9 ca. 0.6 Räisälä Mäntylinna 2005 - - 1 ov 2 5.5x8 NDA - Räisälä Kankaala 2005 CW2/ PitC CW2/ PitC 3 ov 3 1 2 3 NDA 7x10 7x7.5 NDA 0.6 0.4 - Survey: 1) the year the site was found, 2) excavated by Mökkönen in 2005, 3) excavated by Halinen in 2005, 4) excavated by Halinen, Lavento & Timofeev in 2002; Ceramics: HPs= housepits, CW1= Early Comb Ware, CW2= Typical Comb Ware, VOL= Volosovo pottery, PitC= Pit-Comb Ware, CW3= Late Comb Ware, KIE= Kierikki Ware, PÖL= Pöljä Ware, ASB= asbestos tempered ware, ORG= organic tempered ware, LNeo= Late Neolithic pottery, TXT= Textile pottery. Shapes (Shp): ob= oblong, ov= oval, re= rectangular, rd= round; Size Class (SC)= see fig. 5; Number of housepits (#); Length (L), width (W) and depth (D); Features: ENA= entrances/antechambers, EMB= embankment NDA= no data available 161

Paper V Mökkönen, T. 2010. Kivikautinen maanviljely Suomessa (with English summary Neolithic cereal cultivation in Finland ). Suomen Museo 2009, 5 38. Publisher: The Finnish Antiquarian Society, Helsinki ISBN 978-951-9057-77-4 ISSN 0355-1806

Teemu Mökkönen Kivikautinen maanviljely Suomessa TIIVISTELMÄ Viimeisimpien maanviljelyn alkua käsittelevien julkaisujen mukaan Suomen vanhimmat viljojen siitepölyt ajoittuvat kivikauden lopulle noin 2500 1800 cal BC. 2000-luvulla Suomen lähialueilta on saatu runsaasti aiempaa vanhempia viljan siitepölyjen ajoituksia. Sen sijaan Suomessa vanhimman kivikautisen viljelyn tutkimustilanne ei ole juuri muuttunut sitten 1990-luvun. Tämän artikkelin ensimmäisenä tavoitteena on esitellä vanhimmat viljelyn merkit Suomesta ja Suomen lähiympäristöstä. Toisena tavoitteena on esittää näkemys siitä, miten vanhoja maanviljelyn merkkejä Suomesta on odotettavissa. Tätä näkemystä perustellaan olemassa olevien siitepölytutkimusten lisäksi myös arkeologisessa aineistossa näkyvillä muutoksilla yhteisöjen tavoissa asua ja toimia ympäristössään. Kivikautisen maanviljelyn leviämistä pohjoiseen käsitellään useimmiten yhä 1980-luvulla luotujen mallien pohjalta. Baltian pohjoisosien tutkimustilanne on kuitenkin muuttunut merkittävästi viimeisen kymmenen vuoden aikana. Artikkelissa pohditaan myös maanviljelyn leviämistä kuvaavien mallien soveltuvuutta nykyiseen tutkimustilanteeseen. Johdanto Suomen arkeologiaan on iskostunut ajatus nuorakeraamisesta kulttuurista, joka leviää Suomeen maanviljelyn kannalta otollisimmille alueille. Nuorakeraamista kulttuuria on pidetty myös sinä kulttuurivaiheena, jonka aikana maatalous levisi ensimmäistä kertaa Suomeen (esim. Carpelan 1999). Viimeisimmät tutkimustulokset kuitenkin osoittavat, että maanviljely oli tunnettua Suomen lähialueilla Baltian maissa ja Luoteis-Venäjällä jo huomattavasti ennen nuorakeraamisten kulttuurien vaikutusaaltoa. Suomen (sub)neoliittisia kulttuureita on pidetty puhtaasti metsästäjä-keräilijäkulttuureina. Vasta kivikauden lopun Kiukaisten kulttuuri (2300 1600 cal BC) on nähty maataloutta harjoittavana kulttuurivaiheena. Näin siitä huolimatta, että suurin osa kivikauden lopulle ajoittuvista viljan siitepölyistä on löydetty sisämaasta Kiukaisten kulttuurin esiintymisalueen ulkopuolelta. Siitepölytutkimuksissa yksittäisen viljan siitepölyn esiintymiseen suhtaudutaan suurella varovaisuudella, vaikka diagrammissa olisi havaittavissa sen kanssa Suomen Museo 2009 5

samanaikaisia raivaukseen viittaavia merkkejä. Koska varhainen viljely on todennäköisesti ollut pienimuotoista, muun kasvillisuuden sekaan tehdyillä raivioilla harjoitettua viljelyä ja siinä käytetyt itsepölytteiset viljalajit ovat heikosti siitepölyissä havaittavia, varhaisen maatalouden tutkimuksessa pitäisi ottaa paremmin huomioon pienimuotoisen viljelyn merkit. Uudet Itämeren itäpuolelta tehdyt siitepölytutkimukset osoittavat maatalouden olleen tunnettua alueen (sub)neoliittisten kulttuurien piirissä. Virossa maanviljelyn katsotaan olevan yleisesti tunnettua jo tyypillisen kampakeramiikan aikana. Suomesta yhtä vanhat yleisesti hyväksytyt viljan siitepölylöydöt puuttuvat, vaikka siitepölytutkimuksissa on löydetty tähän aikaan ajoittuvia metsänraivausvaiheita. Tässä artikkelissa esitetään, että pienimuotoista maanviljelyä on mahdollisesti ollut myös Suomessa jo tyypillisen kampakeramiikan (4000 3400 cal BC) aikana. Tätä perustellaan siitepölytutkimuksilla, tyypillisen kampakeramiikan aikana tapahtuneilla muutoksilla tavassa asua sekä ajan kontaktiverkostolla, jossa oli jo mukana maataloutta harjoittavia yhteisöjä aivan Suomen lähialueillakin. Artikkelissa esitellään aluksi Suomen ja Itämeren itäpuolen vanhimman maanviljelyn ajoituksia. Seuraavat kaksi lukua käsittelevät esiteltyjen ajoitusten vaikutusta maatalouden leviämisestä käytävään keskusteluun. Ennen keskusteluosiota artikkelissa esitellään tyypillisen kampakeramiikan aikana tapahtuvat muutokset tavassa asua, jotka tulkitaan merkeiksi aiempaa paikallaan pysyvämmästä asutuksesta. On mahdollista, että näiden muutosten aikana varhainen maanviljely on ollut jo tunnettua. Keskusteluosiossa käsitellään siitepölyanalyysien virhelähteitä, eroja pienimuotoisen ja sitä intensiivisemmän viljelyn jättämissä merkeissä sekä viljelyn mahdollisuuksia nuorakeraamisen kulttuurin pohjoispuolella, kuusen leviämisen myötä havumetsävyöhykkeeksi muuttuvalla alueella. Artikkelissa esitetään, että pienimuotoisen maatalouden omaksuminen jo tyypillisen kampakeramiikan aikana on mahdollista. Mahdollisuus tätä vanhemmasta viljelystäkään ei liene poissuljettu. Varhaisessa viljelyssä on mahdollisesti käytetty pähkinäpensas-lehmusalueille tehtyjä pienialaisia raivioita, joista jää vain heikkoja merkkejä siitepölyihin. Intensiivisemmän maiseman muokkauksen katson alkaneen vasta nuorakeraamisen kulttuurin myötä. Maanviljelyn alku Suomessa Suomen vanhimmat konkreettiset viljelyn merkit ajoittuvat kivikauden lopulle. Turun Niuskalasta löydetyt Suomen vanhimmat makrofossiilit ovat kuoretonta ohraa. Jyvät on radiohiiliajoitettu aivan kivi- ja pronssikauden vaihteeseen, 1700 1300 cal BC (Vuorela & Lempiäinen 1988, 40 41; Lempiäinen 1999). Suomen vanhimmat viljan siitepölyt ovat näitä vanhempia. Selkeästi kivikauden lopulle, aikavälille 2500 1800 cal BC ajoitettuja viljojen siitepölyjä oli 1990-luvun puolivälissä tiedossa alle kymmeneltä paikalta (Vuorela & Hicks 1996; ks. myös Vuorela 1998; 1999a; 1999b). 6

Kiista Suomen nuorakeraamisen kulttuurin (3200 2350 cal BC) osuudesta maanviljelyn leviämiseen on yhä olemassa. En aio referoida aiheeseen liittyvää keskustelua enempää kuin toteamalla, että konkreettiset nuorakeraamista maanviljelyä osoittavat todisteet viljan siitepölyt, viljanjyvät ja -jyvien painaumat keramiikassa puuttuvat yhä Suomesta. Olemassa olevat aihetodisteet asutuksen sijainti, erikoistuneen metsästysesineistön puuttuminen ja kaikkien muiden maiden nuorakeraamisten kulttuurien maanviljely ovat jakaneet arkeologit Suomen nuorakeraamisen kulttuurin maanviljelyn puolustajiin (Salo 1997; Carpelan 1999; Nunez 1999; Siiriäinen 2003; Núñez 2004) ja vastustajiin (Edgren 1984; 1999; Meinander 1984; Zvelebil & Rowley-Conwy 1984). Suomen siitepölyaineistoissa on vuonna 1980 julkaistuja kiistanalaisia, muita tuloksia vanhempia viljan siitepölyjä. Ensimmäiset viljan siitepölyt ajoittuvat Raahen Järvelänjärvellä ja Utajärven Ahmasjärvellä jo tyypillistä kampakeramiikkaa vanhempaan aikaan (Reynaud & Hjelmroos 1980) 1 (taulukko 1, kuva 1). Suomalaiset arkeologit (Huurre 1983, 219 221; Meinander 1984; Nunez 1999; Núñez 2004, 362 363) ja siitepölytutkijat (Vuorela & Hicks 1996) ovat suhtautuneet kriittisesti näin vanhoihin ajoituksiin. Pohjois-Pohjanmaan ja Lapin esihistoriassa Huurre (1983, 220) toteaa, ettei kampakeraamisen ajan asuinpaikkojen valinnassa ole havaittavissa maanviljelyyn viittaavia muutoksia eikä kampakeraamisesta kulttuurista löydy selkeästi maanviljelyyn erikoistunutta välineistöä. Vaikuttaa siltä, että tutkijat epäilevät Oulunjokilaakson ja Yli-Tornion vanhojen tuloksien syyksi jotakin ajoitusta vanhentavaa tekijää (Vuorela & Hicks 1996; Nunez 1999; Vuorela 2002). Osassa julkaisuja Oulunjokilaakson varhaisiin viljelyn merkkeihin suhtaudutaan kuitenkin vähemmän varautuneesti ja tulosten varmistaminen nähdään tulevaisuuden haasteena (Vuorela 1999b, 146; 2002, 87; ks. myös Welinder 1998b, 171 172). Kirjallisuudesta ei löydy tarkempaa keskustelua siitä, miksi vuonna 1980 julkaistut varhaiset tulokset on hylätty virheellisinä. Vuonna 1996 julkaistussa Suomen varhaisimpia viljan siitepölyjä käsittelevässä kokooma-artikkelissaan Vuorela ja Hicks toteavat tekstissä Reynaudin ja Hjelmroosin (1980) tuloksista seuraavasti: the evidence from some of the sites has been omitted here either because there is a strong cause to believe that the dating is too old or else the pollen is thought to represent long-distance transported Cerealia pollen and thus does not represent the local situation (Franzén and Hjelmroos, pers. comm..) (Vuorela & Hicks 1996, 246). Saman artikkelin lopussa olevassa taulukossa Reynaudin ja Hjelmroosin kohteet on koottu huomattavasti ehdottomamman otsikon alle: Sites where all the dates in the profile are old compared with other dated events in the area. 1 Vuonna 1980 Faravid 4:ssä julkaistu artikkeli on suomennos samana vuonna julkaistusta artikkelista Reynaud, C. & Hjelmroos, M. 1980: Pollen evidence and radiocarbon dating of human activity within the natural forest vegetation of the Pohjanmaa region (northern Finland). Candonella 35. 257 304. 7

Kuva 1. Kivikautiset viljan siitepölyt Suomesta ja lähiympäristöstä. Suomen kohteiden numerointi viittaa artikkelin taulukkoon 1. Viron Akalin tulos Poska & Saarse (2006). Muut Viron tulokset Kriiska (2003) ja Poska et al. (2004). Venäjän Pegreman tulos Vuorela et al. (2001) ja Novgorodin tulos Königsson et al. (1997). Katkoviiva esittää nuorakeraamisen kulttuurin levinnän pohjoisrajaa. Fig. 1. Stone Age cereal pollen finds from Finland and neighbouring areas. The numbering of the Finnish sites refers to Table 1. Of the Estonian pollen finds, Akali has been published by Poska and Saarse (2006); for other results from Estonia see Kriiska (2003) and Poska et al. (2004). Of the Russian pollen finds, Pegrema has been published by Vuorela et al. (2001) and Novgorod Radbelik by Köningsson et al. (1997). The northern limit of the Corded Ware culture is marked by a dashed line. 8

Taulukko 1. Suomen vanhimmat viljan siitepölyt. Numerointi viittaa kuvaan 1 (sivulla 8). Table 1. The oldest cereal pollen finds from Finland. The numbering refers to Fig. 1 (on p. 8). 9

Pohjois-Pohjanmaan vanhoista viljan siitepölyistä käytävässä keskustelussa on huomionarvoista, että Reynaudin ja Hjelmroosin esittämä kuusen leviämisen ajoitus on yhteneväinen muiden tutkimusten kanssa, kuten he itsekin toteavat (Reynaud & Hjelmroos (1980, 65). Nämä tulokset ovat myös mukana kuusen leviämistä esittävissä tuoreessa tutkimuksessa (Giesecke & Bennett 2004), eivätkä kyseiset tulokset poikkea muusta aineistosta. Tämä on mielestäni vahva argumentti ainakin ajoitustulosten oikeellisuudelle. Vaikuttaakin siltä, että vanhat viljelyn merkit on jätetty huomioimatta lähinnä siksi, että löytöaikanaan ne olivat täysin poikkeavia vallitsevaan tutkimustilanteeseen nähden (ks. Welinder 1998b, 171 172). Näin ollen ajoitusvirhettä todennäköisemmäksi virhelähteeksi nousee siitepölyjen virheellinen määritys. Näiden tutkimustulosten todennäköistä todistusarvoa nostavat kuitenkin muut varhaisiin viljan siitepölyihin ajallisesti yhdistyvät havainnot: metsäpalot (raivaaminen tulella) ja laidunnukseen viittaavien kasvien siitepölymäärien kohoaminen. Pohjois-Suomesta on muitakin vanhoja ajoituksia maatalouden alulle. Kainuusta, Puolangan Vasikkasuolta vanhimmat selkeät merkit Cerealia-tyypin siitepölyistä ja raivauksesta ajoittuvat noin 2300 1880 cal BC. Siitepölydiagrammissa tulkintaa maanviljelyyn liittyvästä raivausvaiheesta tukevat viljan kanssa samassa horisontissa kuolleenkourasaniaisen (Pteridium), pillikkeen (Lamiaceae), suolaheinän (Rumex), peltohatikan (Spergula arvensis) ja nokkosen (Urtica) runsas siitepölystö sekä kasvanut hiilihiukkasten määrä. Vasikkasuolta on toinen tätäkin vanhempi viljan siitepöly, joka ajoittuu kuusen leviämisen kanssa samanaikaiseksi, eli noin 2900 2800 cal BC. Tässä yhteydessä vastaavat viljelyn merkit ovat myös näkyvissä, mutta huomattavasti heikompina (Vuorela & Kankainen 1991; Vuorela 1998; 1999a-b; 2002). Irmeli Vuorela kommentoi Puolangan Vasikkasuon noin 2300 1880 cal BC ajoittuvaa viljelyvaihetta raportissaan seuraavasti: Tähän mennessä tunnetun maanviljelyhistorian valossa vaihe edustaa ennen todistamatonta maanviljelyn tulosuuntaa maassamme. Ilmeistä on, että viljelytoiminta, joka eteni maahamme lounaasta ja etelästä ja saavutti Lounais-Suomen rannikon n. vuonna 2000 e.kr. ei saavuttanut Kainuuta. (Vuorela & Kankainen 1991, 19). Myöhemmissä julkaisuissa Vuorela esittää Puolagan viljelyn tulosuunnaksi idän tai kaakon (Vuorela 1998, 179) tai pitää sekä itäistä että läntistä tulosuuntaa yhtä mahdollisina (Vuorela 1999a, 344). Vasikkasuon vanhinta, noin 2900 2800 cal BC ajoittuvaa viljansiitepölyä Vuorela kommentoi vuonna 2002: vaihetta ei silti ole uskallettu tulkita viljelyraivioksi, siksi sensaatiomainen sen ajankohta olisi. (Vuorela 2002, 87). Tämän hetken yleiskuvan mukaisesti Suomen vanhimmat viljan siitepölyt löytyvät Sisä-Suomesta. Ne ajoittuvat kivikauden lopulle, aikavälille 2500 2000 cal BC. Tätä vanhempia yksittäisen viljan siitepölylöytöjä sellaisista 10

siitepölydiagrammeista, joiden pätevyyttä ei ole epäilty, tunnetaan ainoastaan Puolangan Vasikkasuolta. Tosin tätäkään ei ole uskallettu tulkita varmaksi merkiksi viljan viljelystä huolimatta siitä, että vanhimman yksittäisen viljan siitepölyn kanssa esiintyy samanaikaisesti myös muita raivaukseen viittaavia merkkejä. Maanviljelyn alku Suomen lähiympäristössä nykytutkimuksen mukaan 2000-luvun alku on tuonut mukanaan uusia tuloksia, jotka siirtävät maanviljelyn alkua varhaisemmaksi (ks. kuva 1). Itä-Virossa Akalin kivikautisen asutuksen tuntumassa vanhimmat vehnän (Triticum t.) ja hampun (Cannabis t.) siitepölyt ajoittuvat jo mesoliittisen kivikauden lopulle 5600 cal BC. Hieman myöhemmin samassa diagrammissa, aikavälillä 4900 3200 cal BC, kaura (Avena t.) ja ohra (Hordeum t.) esiintyvät hampun ja vehnän rinnalla, mutta samalla siitepölyjen esiintyminen muuttuu katkonaisemmaksi (Poska & Saarse 2006). Kaiken kaikkiaan Virossa on nuorakeraamista kulttuuria vanhempia viljelyn merkkejä niin runsaasti, että siellä maanviljelyn katsotaan olevan tunnettua jo tyypillisen kampakeramiikan alusta saakka eli noin 4000 cal BC lähtien (Kriiska 2003, 15; Poska et al. 2004). Päätoimisen maanviljelyn ja karjanhoidon alku yhdistettynä yksittäistaloasumiseen alkaa Virossa kuitenkin vasta nuorakeraamisen kulttuurin aikana (Kriiska 2003, 15). Uusia, entistä varhaisempia maanviljelyn alun ajoitustuloksia on saatu myös Venäjältä. Äänisen luonteisrannalla Pegreman kivikautisen asuinpaikka-alueen läheisyydessä viljan (Triticum t.) siitepölyjen esiintyminen alkaa n. 3800 3700 cal BC ja selkeä maiseman avautuminen on nähtävissä 3500 3400 cal BC (Vuorela et al. 2001). Näin ollen Äänisen alueella maanviljelyn alku liittyy rombikuoppa- ja kuoppakampakeramiikkaan, joka vastaa ajallisesti jotakuinkin Suomen tyypillistä ja myöhäiskampakeramiikkaa. Edellä mainitut Itä-Viron vanhimmat viljojen siitepölyjen ajoitukset vaikuttavat todella vanhoilta, kun niitä verrataan Itämeren rantojen muihin ajoituksiin. Näistä mesoliittisen kivikauden lopulle ja aivan neoliittisen kivikauden alkuun ajoitettuja viljan siitepölyjä onkin epäilty mahdollisesti virheellisesti määritetyiksi (Behre 2007, 208). Aiemmat Itämeren piirin vanhimmat, Puolasta löydetyt maanviljelyn merkit ajoittuivat n. 4350 cal BC (Raska-Jasiewiczowa & Latałowa 1996 Polska & Saarse 2006 mukaan). Tämän jälkeen viljelyn merkit lisääntyvät, ja Baltian maissa viljan viljely on yleistä jo 4300 4000 cal BC (Stančikaitė et al. 2009, 11 viitteineen). Idempänä, Venäjällä Ilmajärven alueella (Ilmen) vanhimmat tunnetut viljelyn merkit ajoittuvat n. 3200 cal BC (Königsson et al. 1997). Etelämpänä Valko-Venäjällä vanhin viljely liittyy nauhapikarikulttuuriin ja ajoittuu n. 5700 5000 cal BC (Zernitskaya & Mikhailov 2008). Myös Itämeren länsipuolella vanhin maanviljely on vanhempaa kuin Suomessa. Keski-Ruotsissa Mälarin ja Bergslagenin alueen vanhimmat vehnän ja ohran 11

siemenet liittyvät suppilopikarikulttuuriin ja ajoittuvat välille 3800 3000 cal BC (Hallgren 2008, 117 118). Hieman pohjoisempana, Itä-Ruotsin Upplannissa vanhimmat havainnot sekä ohran että vehnän siitepölyistä alkavat vasta keskineoliittisella kivikaudella (3300 2800 cal BC) ja vielä pohjoisempana Länsipohjassa (ruots. Västerbotten) hieman myöhemmin 2800 2300 cal BC (Welinder 1998a, 181; Karlsson 2007). Samaan ajankohtaan ajoittuvat ensimmäiset merkit laidunnuksesta. Mielenkiintoisena yksityiskohtana Ruotsin pohjoisista maanviljelyyn liittyvistä löydöistä haluan nostaa esiin ohranjyvät, jotka löydettiin 2800 2400 cal BC ajoittuvien suurien asuinpaikkavallien (30 12 m x 10 8 m) kaivauksissa Norlannin Åkermanlandista (Holback et al. 2004). Maanviljelyn leviäminen Baltian pohjoisosiin ja Suomeen Suomen varhaisimman maanviljelyn tutkimus on auttamattomasti jäänyt 1990-luvulle. Itämeren piirin varhaisimman maanviljelyn kuva on 2000-luvun aikana mullistunut verrattuna 1990-luvun lopun tilanteeseen. Tällä on seurauksia myös aiheesta käytävään keskusteluun. Uusien ajoitusten myötä ensimmäisen selkeän maatalouskulttuurin tässä tapauksessa nuorakeramiikan ja ensimmäisten viljakasvien viljelyn välinen ajallinen etäisyys on kasvanut Baltian pohjoisosissa huomattavaksi: vanhimmat viljelyn merkit ajoittuvat jo 2400 vuotta ja useat nuoremmatkin jäljet reilusti yli 300 vuotta nuorakeraamista kulttuuria vanhemmiksi (ks. mm. Kriiska 2003; Poska & Saarse 2006; Poska et al. 2004). Näin ollen nuorakeramiikka ei tuonut pohjoiseen maataloutta vaan ainoastaan uuden, entistä intensiivisemmän tavan toteuttaa sitä (ks. Kriiska 2003, 22). Vuonna 2000 julkaistussa, Itämeren alueen viljelyä aikavälillä 4000 2200 cal BC kuvaavassa kartassa maanviljelysyhteisöjä esiintyy Puolasta Tanskan kautta aina Keski-Ruotsin eteläosiin levittäytyvällä vyöhykkeellä. Korvautumisasteella olevia yhteisöjä (eli yhteisöjä, joissa maanviljely on jo tunnettua, mutta se ei vielä ole toimeentulopohjan kannalta merkittävässä asemassa) löytyy aivan Baltian eteläosista sekä Ruotsista ja Norjasta kapealta rannikkovyöhykkeeltä kohti pohjoista aina Torniota vastaavalle korkeudelle saakka (ks. Zvelebil & Lillie 2000, Fig. 3.10; ks myös Zvelebil 1998, Fig. 1.3). Verrattaessa tätä vuonna 2004 julkaistuun karttaan Euroopan vanhimmista viljan siitepölyistä (Poska et al. 2004, Fig. 6.) Itämeren ympäristön tilanne on muuttunut erityisesti itäpuolen uusien tulosten osalta. Nykyisin aikavälille 4900 3800 cal BC ajoittuvia siitepölyjä löytyy Itämeren eteläosien lisäksi myös kaikista Baltian maista ja Valko-Venäjältä. Nämä ajoitukset eivät ole enää kovinkaan paljon nuorempia kuin Suomen ja Baltian (sub) neoliittisten kulttuurien vanhimmat ajoitukset (ks. Carpelan 1999; 2002). Vuonna 2003 Aivar Kriiska esitti maatalouden saapuneen Viroon joko pohjoisesta Keski-Euroopasta tai vesiyhteyksien välityksellä Etelä-Skandinaviasta. Tuolloin kaikki Viron vanhimmat viljan siitepölyt löytyivät rannikolta ja ajoittui- 12

vat vanhimmillaan 4300 cal BC. Sittemmin tutkimus on muuttanut tätä kuvaa. Nyt Itä-Virosta, Akalin asuinpaikan lähettyviltä Peipsijärven länsirannalta tunnetut, noin 5600 cal BC ajoittuvat viljan viljelyn merkit ovat huomattavasti vanhempia kuin Itämeren rannikkoalueiden maanviljely (Poska & Saarse 2006). Sen sijaan yhtä vanhoja viljelyn merkkejä tunnetaan etelämpää Valko-Venäjän suunnalta (Zernitskaya & Mikhailov 2008). Ottaen huomioon Itä-Viron ja Valko-Venäjän vanhimmat ajoitukset, maanviljely on saattanut levitä alueelle Unkarin suunnalta 5700 5200 cal BC pohjoiskoilliseen suuntautuvan nopean maanviljelyn leviämisaallon seurauksena (Bocquet-Appel et al. 2009). En tosin tiedä, onko Baltian itäosien arkeologisessa aineistossa viitteitä tämän suuntaisista kontakteista. Maanviljelyn leviämistä Eurooppaan käsittelevät tutkimukset painottuvat usein Kreikan ja Keski-Euroopan välisen akselin ympärille; entisten itäblokin maiden tutkimustuloksia ei juuri ole mukana (ks. esim. Gkiasta et al. 2003; Bocquet-Appel et al. 2009). Tämä johtuu todennäköisesti siitä, että itäisen ja läntisen Euroopan neoliittiset kivikaudet ovat keskenään erilaisia. Siinä missä Balkanin kautta Keski-Eurooppaan leviävät neoliittiset kulttuurit omaksuvat maatalouden, Itä-Euroopan (ja Suomen) neoliittinen kivikausi pitää sisällään kaikki muut molemmille yhteiset elementit paitsi maanviljelyn. 2 Nykytutkimuksen valossa Itä-Euroopan (sub)neoliittisten kulttuurien yhteydestä tunnetaan kuitenkin jo runsaasti maanviljelyn merkkejä. Edellä esitin yhtenä mahdollisuutena, että maanviljely leviää Valko-Venäjän nauhapikarikulttuurin yhteydestä pohjoiseen. Toisena vaihtoehtona varhaisen maanviljelyn leviämiselle Pohjois-Itämeren piiriin on Mustanmeren pohjoispuolitse kulkeva itäinen leviämisreitti sama reitti, jota pitkin keramiikan valmistustaito leviää pohjoisella metsävyöhykkeellä ja saapuu Suomeen samaan aikaan kuin maanviljelyskulttuurit Keski-Eurooppaan (Carpelan 1999, kuva 1). Vielä ei ole täysin selvää, missä vaiheessa maanviljely tulee Itä-Euroopan metsävyöhykkeen (sub)neoliittisiin kulttuureihin. Nykyisten siitepölytutkimusten perusteella viljely on ollut tunnettua Itämeren itäreunalla jo ennen nuorakeramiikan mukana leviävää maatalouden vaikutusaaltoa, joka on Balkanin kautta Keski-Eurooppaan ja siitä edelleen pohjoiseen Itämeren piiriin maataloutta levittävän aallon viimeinen häntä. Baltian pohjoisosien osalta on syytä kysyä, miten varhainen maanviljely on levinnyt alueelle. Onko maatalouden tietotaito mukana jo itäisessä vaikutusallossa, joka tuo keramiikan alueelle (ks. Davison et al. 2009), vai tuleeko maanviljely itäiseen neolitikumiin mukaan vasta, kun idän (sub)neoliittiset kulttuurit kohtaavat Balkanilta Keski-Eurooppaan suuntautuvan maataloutta levittävän vaikutusaallon 5500 5000 cal BC tietämillä Karpaattien vuoriston ja Volgan välisellä metsä-/ aroalueella (Zvelebil & Lillie 2000)? Mikäli maatalous omaksutaan metsäalueen (sub)neoliittisiin kulttuureihin, on täysin mahdollista, että maatalouden idea 2 Itä-Euroopan keraamisella kivikaudella, josta puuttuu maanviljely, on monta nimitystä: Boreal Neolithic, Subneolithic, Forest Neolithic, Paraneolithic ja joskus jopa Ceramic/Pottery Mesolithic (esim. Werbart 1998). 13

leviää myös metsäalueella keramiikan leviämisen mukana. Sikäli ei ole täysin mahdotonta, että maanviljely olisi tunnettua Suomessa jo Sperrings 1 -keramiikan käyttöaikana. Leviämisreittiin liittyvistä epävarmuuksista huolimatta on selvää, että maanviljely oli tunnettua Itämeren itäpuolen (sub)neoliittisten kulttuurien piirissä jo ennen Keski-Euroopan kautta nuorakeraamisen kulttuurin muodossa tulevaa vaikutusaaltoa. Maanviljelyn leviämistä kuvaavat mallit Uudet tulokset eivät enää mahdu vanhojen mallien raameihin, vaan ne vaativat seurakseen uutta tulkintaa. 1980- ja 1990-luvuilla esitetyt maanviljelyn leviämistä kuvaavat mallit pohjautuvat tilanteeseen, jossa Baltian pohjoisosien maatalouden alun katsottiin alkavan jotakuinkin samanaikaisesti nuorakeraamisten kulttuurien saapumisen kanssa. Pohjois-Euroopan tilannetta kuvaavat mallit on tehty maataloutta nopeasti levittävien kulttuurien nauhapikari-, suppilopikari- ja nuorakeraamisen kulttuurin näkökulmasta. Nämä mallit eivät anna luontevaa mahdollisuutta keskustella nuorakeramiikkaa varhaisemmasta maanviljelyksestä. On muistettava, että keskusteltaessa maanviljelyn alkamisesta, viljelyn osuus toimeentulopohjasta ei voi olla kynnyskysymyksenä. Myös pyynnillä toimentulonsa perustan hankkiva yhteisö, joka harjoittaa pienimuotoista maataloutta, harjoittaa maanviljelyyn kuuluvia prosesseja siinä missä intensiivisempääkin viljelyä harjoittavat kulttuurit. 1990-luvulla siirtymistä maanviljelyyn kuvattiin kolmiasteisena prosessina maanviljelys- ja metsästäjä-keräilijäkulttuurien kontaktivyöhykkeellä: 1 saatavuusaste, 2 korvautumisaste, jossa metsästyskulttuuri siirtyy toimeentulossa enemmän maanviljelyyn 3 vakiintumisaste, jossa maanviljely on pääelinkeinona (Zvelebil & Rowley-Conwy 1984; Zvelebil & Dolukhanov 1991; Zvelebil 1998; Zvelebil & Lillie 2000). Tuolloin koko Suomen (sub)neoliittisen kivikauden katsottiin olevan saatavuusasteella, ja korvautumisaste saavutettiin vasta pronssikaudella (Zvelebil 1998; Zvelebil & Lillie 2000). Valter Lang (1999) on kritisoinut edellä esitettyä mallia liian epätarkaksi, jotta sitä voisi käyttää kuvaamaan Baltian ja Suomen siirtymistä maatalouteen. Lang esittää tilalle seuraavaa nelitasoista mallia: 1 metsästäjä-keräilijöiden kasvien käytön tehostuminen, 2 metsästäjä-keräilijöiden ensimmäiset kokemukset maanviljelystä ja karjanhoidosta, 3 asteittainen siirtyminen maanviljelytalouteen 4 intensiivisen maatalouden alku. 14

Mielestäni 1980- ja 1990-luvuilla esitetyt maatalouden alkua kuvaavat mallit ovat ongelmallisia, sillä niissä alkupisteenä on puhdas metsästäjä-keräilijäkulttuuri ja lopputuotoksena intensiivinen maatalouskulttuuri. Tällaiset mallit tuntuvat varmasti toimivilta tilanteessa, jossa ajallinen etäisyys maanviljelyn ensimmäisten merkkien ja selkeän maatalouskulttuurin välillä on suhteellisen lyhyt. Näin oli vielä 1990-luvun lopulla, jolloin Baltian alueen vanhimmat viljelyn merkit ajoittuivat 3500 3100 cal BC eli vanhimmillaan noin kolmesataa vuotta alueen ensimmäistä selvää maatalouskulttuuria nuorakeraamista kulttuuria vanhemmaksi (Rimantienė 1992; Lang 1999). Malleissa, jotka kuvaavat Itämeren itärannan metsästäjä-keräilijöiden siirtymistä maanviljelyyn, on muitakin ongelmia kuin niiden rakentuminen kuvaamaan tutkimustilannetta, jossa maanviljely leviää suhteellisen nopeasti Keski-Euroopasta pohjoiseen nuorakeraamisen kulttuurin mukana. Toinen ongelma liittyy maanviljelyn leviämistapaan. 1990-luvulla luoduissa malleissa maanviljelyn leviäminen nähdään eräänlaisten kulttuuristen kehitysvaiheiden vyöhykkeinä, joiden sijainti ja ajoitus muuttuvat maanviljelijöiden ja metsästäjä-keräilijöiden kontaktin mukana. Suurin ongelma tässäkin on mallin evolutiivinen lähtökohta erilaiset maatalouskulttuuriin johtavat kehitysasteet esitetään kronologiakaavioissa seuraten kokonaisia kulttuureita ja karttaesityksissä vastaavat vaiheet kuvataan vyöhykkeinä (ks. Pluciennik 2008). Varhainen maanviljely ei kuitenkaan leviä vyöhykkeinä vaan pesäkkeisesti ja monimuotoisesti varioiden (Borić 2005; Kotsakis 2005; Robb & Miracle 2007; Bocquet-Appel et al. 2009). Siksi viljelyn varhaisen vaiheen toimeentulopohjan ja asutusmallin variaatio on olettavasti ollut suurta; saman materiaalisen kulttuurikompleksin sisällä on sekä maanviljelyä harjoittavia yhteisöjä että yhteisöjä, joiden toimeentulopohjassa ei vielä ole mukana maatalouselinkeinoja (Rowley- Conwy 2004; Robb & Miracle 2007). Näin ollen saman arkeologisen kulttuurin sisällä voi olla eri toimeentulostrategioita noudattavia yhteisöjä. Muutoksia tavassa asua Neoliittisen kivikauden alkuun kuuluvia kulttuurisia muutoksia kuvataan usein termillä neolithic package. Tähän neoliittiseen kokonaispakettiin sisältyy keramiikan, hiottujen kiviesineiden ja maatalouden (sekä viljely että kotieläimet) lisäksi myös kyläasutus (esim. Robb & Miracle 2007). Itä-Euroopan (sub)neoliittisten kulttuurien on katsottu täyttävän neoliittisen kivikauden piirteet maataloutta lukuun ottamatta (esim. Werbart 1998; Davidson et al. 2009). Kylämäinen asutus ilmenee Suomen aineistossa viimeistään tyypillisen kampakeramiikan aikaisissa (4000 3400 cal BC) asumuspainannekeskittymissä. Tyypillisen kampakeramiikan aikana tapahtuu kuitenkin myös muutoksia siinä, millaisille paikoille asumuspainanteet rakennetaan. Lisäksi kivikauden lopulla asumusten koko kasvaa. Seuraavaksi esittelen lyhyesti tätä kehitystä ja pohdin sen merkitystä maatalouden vaatimien toimintojen näkökulmasta. 15

Muutoksia asumuspainanneasuinpaikkojen sijainnissa Maanviljely on toimintaa, joka vaikuttaa toimeentulon lisäksi myös yhteisön muuhun elämään. Maanviljelytuotteet vaativat mm. varastointia, joka on helpointa toteuttaa suhteellisen paikallaan pysyvässä asutusmallissa. Tyypillisen kampakeramiikan aikana asumuspainanteiden sijoittumisessa tapahtuva muutos merkitsee aiempaa pysyvämmän asutuksen syntyä. Samaan aikaan syntyvät myös kylämäiset asumuspainannekeskittymät. Mielenkiintoinen (sub)neoliittisiin asumuspainanteisiin liittyvä muutos on havaittavissa Saimaalla (Mökkönen 2002) ja Muinais-Laatokan rannoilla Karjalankannaksella (Mökkönen 2009). Molemmissa tapauksissa asumuspainanteiden sijaintiympäristö muuttuu tyypillisen kampakeramiikan (n. 4000 3400 cal BC) aikana suojaisista lahdenperukoista saariin ja tuulisiin niemenkärkiin. Muutos vanhemmista suojaisista paikoista nuorempiin, tuulisiin ja suojattomiin paikkoihin on asuinpaikan välittömän ympäristön kannalta radikaali. Käytännössä tämä muutos tarkoittaa ainakin talviasumuksina käytettyjen asumuspainanteiden rakentamista paikoille, jotka lähtökohtaisesti soveltuvat paremmin kesä- kuin talviasumiseen. Oletan tämän muutoksen merkitsevän asutuksen muuttumista paikallaan pysyvämmäksi (Mökkönen 2009). Suurempien asumuspainannekeskittymien, mahdollisten kylien, synty tapahtuu tämän muutoksen kanssa samanaikaisesti. Keskisen kivikauden aikana tapahtuva muutos asumuspainannekohteiden sijoittumisessa näkyy eri alueilla eri tavalla. Sisä-Suomessa Muinais-Saimaan mosaiikkimaisessa ympäristössä asutuksen siirtyminen suojaisesta paikasta suojattomaan merkitsee usein konkreettisena etäisyytenä vain lyhyttä siirtymää lahden pohjukasta viereisen niemen kärkeen (Mökkönen 2002). Laajemmassa mielessä, esimerkiksi asuinpaikan resurssialueen suhteen, tämä muutos on merkityksetön. Sen sijaan esimerkiksi ruuan varastoinnin suhteen tuulisemmat, paremmin kuivana pysyvät asuinpaikat ovat ihanteellisia pitempiaikaista kuin vain yhden talven säilöntää varten. Muinais-Laatokan rantamilla muutos asumuspainanneasuinpaikkojen sijoittumisessa näkyy selvemmin kuin Muinais-Saimaalla. Entisten Kaukolan ja Räisälän kuntien alueella Muinais-Laatokan lahdet ovat pitkiä ja kapeita, minkä seurauksena varsinkin Räisälän alueella tyypillisen kampakeramiikan aikana tapahtunut asumuspainanteiden siirtyminen sisäsaaristoon on selvästi nähtävissä (Halinen & Mökkönen 2009; Mökkönen 2009). Pohjois-Pohjanmaalla Iijoen kivikautisen jokivarsi- ja jokisuuasutuksen asumuspainannekohteiden (ks. Vaneeckhout 2008a-b) sijaintiympäristössä ei ole havaittavissa samanlaisia muutoksia kuin järvialueiden monimuotoisemmassa ympäristössä. Myös Iijoen varressa isompien asumuspainannekeskittymien synty on jotakuinkin samanaikainen kuin Muinais-Saimaalla ja Karjalankannaksella, siis noin 3700 3500 cal BC (ks. Mökkönen 2002; 2009; Núñez 2004; Vaneeckhout 2009). 16

Muutos yhteisön tavassa asua ja toimia ympäristössä on yleensä seuraus joko uuden väestön saapumisesta alueelle tai kulttuurin sisällä tapahtuneesta suuremmasta ideologisesta muutoksesta, johon liittyy yleensä myös teknologis-taloudellinen/organisatorinen muutos 3 (Rafferty 1985; Tilley 1994; Ingold 2000, 153 ff; Thomas 2001). Itä-Ruotsissa, Norlannin eteläosien keskineoliittisessa asutuksessa on havaittu samansuuntaisia muutoksia kuin tyypillisen kampakeramiikan aikana Saimaalla ja Karjalankannaksella. Norlannissa asutuksen painopiste siirtyy aikavälillä 3400 3000 cal BC jokisuista kohti ulkosaaristoa, ja keskineoliittisen kauden lopulla n. 3000 2300 cal BC asutus on siirtynyt kokonaan ulkosaaristoon (Björk 2003). Ruotsissa tähän muutokseen liittyy asuinpaikkojen määrän ja koon pieneneminen. Koska muutos asutusmallissa tapahtuu samaan aikaan kuin ensimmäiset maanviljelyn merkit tulevat alueella näkyviin, muutosten on oletettu kytkeytyvän maanviljelyn omaksumiseen ja siinä yhteydessä tapahtuviin muutoksiin tavassa asua ja toimia ympäristössä (Björk & Larsson 2007). Muutoksia asumuksissa Pitkätalo on asumustyyppi, joka selkeästi yhdistyy Itämeren alueelle etelästä leviäviin maatalouskulttuureihin (esim. Welinder 1998a, 109). Ruotsissa pitkätalot liittyvät sekä suppilopikarikulttuuriin että nuorakeramiikkaan, Itämeren itäpuolella nuorakeramiikkaan ja niin sanottuun Baltic coastal culture -hybridikulttuuriin. Suomesta puhtaasti nuorakeraamisia asumuksia ei tunneta (ks. kuitenkin Mökkönen 2008). Suomessa (sub)neoliittisen kivikauden tyyppiasumukseksi varhaiskampakeraamista vaihetta lukuun ottamatta on vakiintunut 1990-luvun aikana asumuspainanne (ks. Pesonen 2002). Rakennustradition sisällä on kuitenkin ajallista ja alueellista variaatiota. Pohjois-Euroopassa asumusten koko kasvaa yleisesti pituussuunnassa kolmannen vuosituhannen BC aikana (Norberg 2008, 159 160). Myös Suomessa asumuspainanteiden koko ja pituus kasvavat keskineoliittisella kaudella (Mökkönen 2002; Vaneeckhout 2008a; 2009). Kaikkein suurimmat asumukset ilmaantuvat paikallisten kulttuurien ja nuorakeraamisen kulttuurin kontaktivyöhykkeiden aineistoihin eri alueilla jotakuinkin samanaikaisesti. Itse pidän pitkiä ja monihuoneisia asumuspainanteita eräänlaisina nuorakeraamisesta kulttuurista omaksuttuina kulttuurilainoina asumuspainanteiden rakennustekniikalla rakennettuina pitkätaloina (Mökkönen 2008). Asumusten koon kasvun yhteydessä myös asumusten pohjan muoto muuttuu suorakulmaisemmaksi etupäässä siten, että suurempien asumusten pituus kasvaa. Tämä muutos tapahtuu Saimaalla (Mökkönen 2002), Pohjois-Pohjanmaalla Yli- Iissä (Vaneeckhout 2008b) ja Karjalankannaksella (Mökkönen 2009) pääosin tyypillisen kampakeramiikan (4000 3400 cal BC) jälkeisenä aikana. 3 Teknologis-taloudelliset/organisatoriset muutokset voivat liittyä uusien talousmuotojen omaksumiseen (maatalous) tai innovaatioihin olemassa oleven resussien hyödyntämisessä esim. säilönnän tai muiden uusien menetelmien kautta. 17

Etnografisista lähteistä (Murdock 1967) tunnetut maahan kaivetuissa asumuksissa asuvat yhteisöt ovat 77-prosenttisesti metsästäjä-keräilijöitä. Jäljelle jäävään 23 %:iin mahtuu sekä intensiivistä että pienimuotoista maataloutta harjoittavia yhteisöjä (Gilman 1987, 545 546). Vaikka maanviljely onnistuu myös asumuspainanteissa asuvalta yhteisöltä, maataloutta harjoittavassa kulttuurissa päädytään kuitenkin yleensä ennemmin tai myöhemmin asumaan useampihuoneisissa ja/ tai maanpäällisissä, pohjakaavaltaan suorakaiteisissa asumuksissa. Tämä muutos tapahtuu usein siirryttäessä aiempaa intensiivisempään maatalouteen (Gilman 1987). Asumuspainanteita pidetään usein kosteusolosuhteiden takia sopimattomina kesäasumiseen (esim. Gilman 1987, 542 543). Silti isompien, suorakulmaisten ja/tai moniovisten/-huoneisten rakenteiden ominaisuudet tuuletuksen, lämmönsäätelyn ja siten myös kosteuden poiston suhteen ovat hyvät (Wilkins 2009). Asumusten koon kasvu ja monihuoneisten asumusten/rivitalojen synty tuo keskineoliittisella kivikaudella mukanaan muutoksia, jotka parantavat asumusten käytettävyyttä muulloinkin kuin talvella. Lämpötilan ja tuuletuksen säätely ovien ja/tai väliseinien sekä eri lattiakorkeuksilla olevien huonetilojen avulla lisää mahdollisuuksia kontrolloida asumuksen lämpötilaa ja luo paremmat olosuhteet varastoinnille myös lämpiminä vuodenaikoina. Asumusten koon kasvu ja pohjakaavan muodon muutokset heijastavat monen muun asian lisäksi myös asumuksen tilankäytölle asetettuja vaatimuksia. Esimerkiksi suorakaiteen muotoisen asumuksen on olettu olevan paremmin muunneltavissa tarpeiden mukaan, minkä takia tämä pohjamuoto liittyy usein paikallaan pysyvään asutukseen (McGuire & Schiffer 1983). Suuremmat asumukset vastaavat myös paremmin kasvaneeseen sisäaskareiden ja säilönnän tarpeeseen (esim. Gilman 1987; Ames 2003). Suomessa asumuspainanteiden koon kasvun myötä asumusten yhteydessä alkaa ilmetä myös aiempaa intensiivisempää säilöntää indikoivia asumusten ulkopuolisia varastokuoppia 4, minkä voi olettaa merkitsevän myös Suomen oloissa aiempaa paikallaan pysyvämpää asutusta. Ihminen luonnonympäristön muuttajana Maatalouteen liittyvä raivaus näkyy siitepölyissä puiden osuuden vähentymisenä, valoa suosivien kasvien määrän nousuna ja hiilihiukkasten määrän kasvuna. Esimerkiksi Ruotsissa Vätternin itärannalla tutkitun Alvastran paaluasumuksen (n. 3300 3000 cal BC) luona varhaiset viljan siitepölyt esiintyvät usein samassa yhteydessä kuolleenkourasaniaisten (Pteredium) esiintymispiikkien ja hiilihiuk- 4 Tästä ovat esimerkkeinä Puumalan Kärmelahti (Katiskoski 2002), Yli-Iin Kuuselankankaan Kierikin keramiikkaan liittyvät painanteet ja varastokuopat (mm. Vaneeckhout 2009) sekä Virolahden Meskäärtty (Mökkönen 2008) ja Takala. Yli-Iin alueen asumusten koon kehityksestä ja varastokuoppien esiintymisestä ks. Vaneeckhout (2008b). 18

kasten osoittamien metsäpalonmerkkien kanssa (Göransson 1998, 67 69). Eri puolilta Suomea tunnetaan kivikautisen raivausvaiheiden merkkejä, mutta niiden yhteydestä ei kuitenkaan ole aina löydetty viljojen siitepölyjä (esim. Alenius 2008; Vuorela 1995; Vuorela & Kankainen 1992; 1993; Vuorela et al. 1993). Varhaisen maanviljelyn havaitseminen on vaikeaa ja vanhimman viljan viljelyn rajan (C 0 -raja eli ensimmäinen viljan siitepöly) löytäminen on sattumanvaraista (mm. Saastamoinen 1996, 132; 1999, 140; Vuorela 1999b, 144 145). Ensimmäiset viljellyt viljalajit vehnä, ohra ja kaura ovat itsepölyttyviä, minkä takia ne tuottavat vain vähän ja pienelle alueelle leviävää siitepölyä. Tämä heijastuu kyseisten lajien heikkoon näkyvyyteen siitepölystössä jopa aivan viljelykohteen lähistöllä (Vuorela 1973; Alenius et al. 2008). Tämän vuoksi viljan siitepölyjen puuttumista ei voi pitää maanviljelyn harjoittamista poissulkevana kriteerinä (Poska et al. 1999, 307). Myöhemmin viljelykasviksi tuleva ruis taas on tuulipölytteinen kasvi, joka levittää laajalle alueelle runsaasti siitepölyä. Koska viljan siitepölyjen puuttumista ei voida pitää maanviljelyn olemassaoloa poissulkevana kriteerinä, koko keskustelu Suomen nuorakeramiikan suhteesta maanviljelyyn tuntuu paradoksaaliselta. Kiistely nuorakeramiikan aikaisesta maanviljelystä on sidottu pitkälti viljanjyvien, keramiikassa esiintyvien jyväpainaumien ja viljan siitepölyjen puuttumiseen (esim. Edgren 1984; 1999). Nuorakeramiikan osalta voi todeta, että maanviljelyn puolesta puhuvat aihetodisteet ovat niin vahvat, ettei kyseessä voi olla muu kuin maataloutta harjoittanut kulttuuri. Nuorakeramiikan asuinpaikkojen sijoittuminen perustuu maataloutta harjoittavan kulttuurin tapaan asettua asumaan, eikä nykytietämyksen mukaan yli viisisataa vuotta kestäneen kulttuurivaiheen aikana ole tapahtunut muutosta asuinpaikan valinnassa. Mikäli nuorakeraamisen kulttuurin konsepti olisi ollut täysin sopimaton paikallisiin olosuhteisiin, se olisi oletettavasti sopeutunut ja muuttunut huomattavasti nopeammin. Lisäksi on muistettava, että nuorakeraamisen asutuksen lähistöltä tunnetaan samanaikaisia selkeitä raivaamisen ja laiduntamisen merkkejä. Yhtenä esimerkkinä on Tammelan Kuivajärvi, jonka kerrostumista tehdyt havainnot viittaavat raivaamiseen ja laiduntamiseen 3000 2700 cal BC välillä (Vuorela & Kankainen 1992). Järven ympäristöstä tunnetaan raivauksen kanssa samaan aikahorisonttiin ajoittuvat, nuorakeramiikkaa sisältävät Mattilan ja Syrjälän asuinpaikat (Pesonen 2008, 21, Taulukko 3). Vastaava tulkinta nuorakeraamisesta laidunnuksesta on tehty myös Tampereelta Mikkolanlammin ja Iidesjärven siitepölyjen perusteella (Alhonen 1988, 46 47, kuva 25). Maata on mahdollista viljellä monella tavalla, ja erilaiset maanviljelykulttuurit jättävät erilaisia siitepölyissä havaittavia jälkiä. Esimerkiksi Pohjois-Amerikan Ontarion alueen (Suurten järvien alueella) huron-intiaanien maanviljely näkyy vain vähäisenä lehtipuiden vähenemisenä sekä valoa tarvitsevien saniaisten jaruohokasvien lisääntymisenä. Polttamalla ja harventamalla tehty raivio umpeutuu nopeasti viljelyn päätyttyä, mikä näkyy lehtipuiden määrän pienoisena kasvuna. Vastakohtana tällaiselle maataloudelle on Ontarion alueen eurooppalaistyyppistä 19

viljelyä harjoittava 1870-luvulla alueelle siirtyvä uudisasutus, jonka aiheuttamat muutokset näkyvät huomattavasti voimakkaampina kuin alkuperäisväestön vaatimattomampi maanviljely 5 (Burden et al. 1986; McAndrews & Boyko-Diakonov 1989). Varhaisessa vaiheessa viljaa ei ole aina kasvatettu pelloilla. Esimerkiksi Skotlannissa 3800 3700 cal BC ajoittuvan hirsitalon (timber hall) lähiympäristön siitepölytutkimusten mukaan vehnää, ohraa ja kauraa viljeltiin asumuksen välittömässä ympäristössä pähkinäpensaistoon tehdyillä pienillä raivioilla (Tipping et al. 2009). Vastaavaa pähkinäpensaiston harventamiseen perustuvaa viljelytapaa on esitetty käytetyksi myös Ruotsin neoliittisella kivikaudella (Göransson 1988, 70 73; Welinder 1998a, 32 33). Edellä esitetyn perusteella Suomen (sub)neoliittisten kulttuurien raivaustoiminta (esim. Vuorela & Kankainen 1993; Vuorela 1995; 1996) saattaa olla liitettävissä varhaiseen maatalouteen, joka ei ole yhtä voimaperäistä ja siitepölyissä niin selkeästi havaittavissa kuin esimerkiksi nuorakeramiikan yhteydessä havaitut raivausvaiheet. Keskustelua Siitepölyanalyysien virhelähteitä Varhaisen maatalouden tunnistamiseen siitepölyanalyysin avulla liittyy useita virhelähteitä: virheellinen siitepölyn tunnistaminen, aineiston siirtyminen vääriin kerrostumiin, ajoitukseen liittyvät virhelähteet ja siitepölyjen kaukolento (Simola 1998; Welinder 1998a, 65 67; Behre 2007). Viljan siitepölyjen esiintymisen voi katsoa olevan peräisin paikallisesta maanviljelystä, jos siitepölydiagrammissa on samassa kohdassa havaittavissa merkkejä metsän raivaamisesta sekä laidunnuksesta tai ihmisen läsnäolosta kertovia indikaattorilajeja. Metsäisen Suomen tapauksessa varhaisten itsepölyttävien ja huonosti siitepölyjä levittävien viljakasvien siitepölyjen kaukolento (ja vielä näiden pölyjen löytäminen) tuntuu epätodennäköiseltä virhelähteeltä. Aineiston kontaminaation mahdollisuus on aina olemassa, mutta sekin tuntuu epätodennäköiseltä selitykseltä, mikäli diagrammissa löytyy viljan kanssa samasta tasosta muita raivauksen merkkejä ja muut kasvillisuudessa havaittavat muutokset ovat ajallisesti oikeilla paikoillaan. Jäljelle jäävä siitepölyn virheellinen tunnistaminen on todennäköisin 5 Pohjois-Amerikan huron-intiaanien pienimuotoinen viljely ja uudisasukkaiden eurooppalaistyylinen intensiivinen maatalous on otettu tässä esimerkiksi kahdesta erilaisesta tavasta harjoittaa maanviljelyä siitä huolimatta, että huron-intiaanien pääsääntöinen viljelykasvi oli maissi. Esimerkki on kuitenkin toimiva, sillä myös maissi on harvinainen pölystössä ja siksi intiaanien maanviljelyn merkkejä täytyy etsiä muita kasveja hyödyntäen (Burden et al. 1986, 49). Luonnonolojensa puolesta Suurten järvien alue sopii hyvin esimerkiksi, sillä alueen luonnonolot vastaavat napapiirin eteläpuolista Suomea sekä kasvillisuus- että ilmastovyöhykkeiden osalta. 20

varhaisen viljan siitepölyn virhelähde (ks. Behre 2007, 213 214). 6 Tätä argumenttia käytettäessä täytyy huomioida, mikä laji on sekoitettu viljaan ja kuinka todennäköisesti kyseinen laji on esiintynyt näytteenottopaikalla. Viimeaikainen keskieurooppalainen keskustelu mesoliittisen kivikauden lopun pienimuotoisesta maanviljelystä kiertyy pitkälti siitepölyjen virheellisen määrittämisen ympärille. Pienimuotoisen varhaisen maatalouden puolustajat perustavat argumenttinsa varhaisten viljan siitepölyjen esiintymiseen yhdessä muiden maataloutta ilmaisevien kasvillisuusmuutosten kanssa (Tinner et al. 2007; 2008). Nämä havainnot ajoittuvat kaikki mesoliittisen kivikauden loppuvaiheeseen. Esineoliittisen viljelyn vastustajat pitävät havaittuja viljojen siitepölymääriä niin pieninä, ettei niiden perusteella pysty varmuudella sanomaan, onko kyseessä oikean viljakasvin siitepöly vai poikkeuksen suurikokoinen luonnonkasvista peräisin oleva siitepöly. Behren (2007; 2008) mukaan modernien siitepölyanalyysien suuret pölyjen laskumäärät lisäävät todennäköisyyttä löytää kooltaan normaalia suurempia luonnonkasvien siitepölyjä, joita sitten usein erehdytään tulkitsemaan viljakasveiksi. Hän pitää kaikkia Keski-Euroopan mesoliittiseksi ajoitettuja viljan siitepölyjä tavalla tai toisella virheellisesti tulkittuina. Viljan ja luonnonkasvien siitepölyjen sekoittumisen mahdollisuus kasvaa myös sen seikan takia, että varhaisten viljojen siitepölyt ovat kooltaan nykyisten lajitovereittensa siitepölyjä pienempiä. Esimerkiksi ohran siitepölyn halkaisijan alarajana perustuen modernien lajikkeiden siitepölyjen kokojakaumaan pidetään eri tutkimuksissa vaihtelevasti 0,036 0,04 mm (esim. Vuorela et al. 1993; Poska & Saarse 2006 viitteineen). Näin ollen esimerkiksi Laukaan Vuojärven kerrostumista löydetty, muita heinien pölyjä selvästi isompi siitepöly (halkaisijaltaan 0,03 mm) jää tulkinnaltaan avoimeksi. Kyseessä saattaa olla varhainen viljelty ohra tai sitten luonnon laji (rantavehnä tai juolavehnä) (Vuorela et al. 1993, 39 40). Tässä tapauksessa tilanne on kutkuttava. Kyseinen mahdollisen ohran löytökerros ajoittuu noin 2300 2000 cal BC ja sen kanssa samalla korkeustasolla yhdistyvät selvät raivauksen merkit: kuusen osuuden sekä hehkutushäviön lasku ja hiilihiukkasten määrän ja ruohojen pölymäärän nousu. Nyt havainnot on tulkittu mahdollisten laidunten raivaamiseksi (Vuorela et al. 1993). 6 Helpoiten viljan siitepölyihin sekoittuvia lajeja ovat: villivehnät (Aegilops ovata pukinvehnä, Agropyron caninum koiranvehnä, Agropyron intermedium engl. intermediate wheatgrass), Avena fatua hukkakaura, Bromus erectus pystykattara, Bromus inermis rehukattara, Bromus mollis mäki-/ nurmikattara, Glyceria fluitans ojasorsimo, Glyceria plicata matala ojasorsimo, Hordeum murinum hiirenohra, Lygeum spartum engl. esparto grass, Secale montanum ruis, Setaria glauca/setaria pumila sinipantaheinä, Triticum aegilopoides villi yksijyvävehnä ja Triticum dicoccoides villi kaksijyvävehnä (emmervehnä) (Behre 2007 mukaan). Suomessa mahdollisia sekoitettavia lajeja ovat lisäksi ainakin Leymus arenarium rantavehnä ja Elymus repens juolavehnä (Vuorela et al. 1993). 21

Erilaiset maanviljelyn merkit Edellä esiteltiin erilaisia maanviljelyn siitepölystöön jättämiä merkkejä. Voimaperäisempi maatalous raivattuine pelto- ja niittyalueineen näkyy siitepölyissä selvinä muutoksina, toisin kuin muun kasvillisuuden sekaan tehdyillä raivioilla harjoitettava ja vain vähäisiä merkkejä jättävä kuokkaviljely tai puutarhatalous (horticulture). Arvelen, ettei voimaperäisempään maanviljelyyn keskittynyt tutkimus ole oikealla tavalla orientoitunut löytääkseen pienimuotoisen maanviljelyn merkkejä. Siksi Pohjois-Euroopan maanviljelytutkimus ei voi seurata samoja polkuja kuin nauhakeraamisen kulttuurin (Linearbandkeramik) ympärille kietoutuva keskieurooppalainen tutkimus. Jos nauhakeramiikkaa pidetään neoliittisen maanviljelykulttuurin ainoana oikeana esikuvana (ks. Behre 2007), muunlaista maataloutta on mahdotonta löytää. Aiemmin viitattu keskustelu myöhäismesoliittisen maatalouden olemassaolosta (Behre 2007; 2008; Tinner et al. 2007; 2008) on mielenkiintoinen, sillä siinä heijastuu myös kaksi perusolettamuksiltaan hyvin erilaista tapaa lähestyä maanviljelyä: neoliittisen voimaperäisen maatalouden lähtökohdista tehtävä tutkimus ja pienimuotoisempaa, luonnonmuutoksiltaan vähäisempää pienimuotoista maataloutta etsivä tutkimus. Vaikuttaakin siltä, että lähtökohtaisesti intensiivistä maataloutta etsivä tutkimus ei voi ikinä hyväksyä siitepölyissä vain heikosti näkyviä maanviljelyn merkkejä, ellei lisätodisteeksi löydy arkeologisista konteksteista löydettyjä ja ajoitettuja viljanjyvien makrofossiileja. Pienimuotoisen maanviljelyn kannalta takertuminen pelkkiin viljakasvien siitepölyihin tuntuu oudolta. Mikäli metsän raivauksen merkit yhdistyvät muihin viljelyä ja/tai laidunnusta ilmaisevien lajien esiintymiseen, vaatiiko tämä aina seurakseen vaikeasti löydettävän viljan siitepölyn? Siitepölyanalyysien tulkinnassa tuntuu vallitsevan suuntaus, jossa jopa yksittäiseen viljan siitepölyyn suhtaudutaan suurella varauksella, vaikka kyseessä olisi vaikeasti löydettävä, heikosti siitepölyä levittävä itsepölytteinen viljalaji, jonka löytöhorisontissa on havaittavissa myös muita viljelytulkintaa tukevia kasvillisuusmuutoksia. Onko pohjoisen havumetsävyöhykkeen varhaisen maanviljelyn tutkimus keskittynyt liiaksi viljan siitepölyihin? Tutkimus ja odotukset tulisi suunnata vastaamaan pienimuotoista viljelyä: raivauksen merkit ovat ensimmäinen indikaattori ihmistoiminnasta, lajisto ja siinä havaitut muutokset antavat viitteen millaisesta raivaustoiminnasta on kysymys, ja viljan siitepölyn löytäminen on tämän päälle mukava tulkintaa varmistava lisä. Kuusen ja viljelyn suhde Kuusen leviäminen liittyy kahdella tavalla Suomen varhaisesta maanviljelystä käytävään keskusteluun. Näistä ensimmäinen liittyy nuorakeraamisen kulttuurin 22

ja kuusen suhteeseen. Toinen liittyy vanhimpiin viljan siitepölyihin ja kuusen leviämiseen, jotka esiintyvät usein samanaikaisesti siitepölydiagrammeissa. Nuorakeramiikan ja kuusen leviämisrajojen yhteneväisyyden on oletettu merkitsevän nuorakeraamisen kulttuurin tarkoituksellista pysyttelyä leviävän havumetsävyöhykkeen ulkopuolella (Edgren 1970, 48; 1984; Luoto 1986). Tämän yhdessä sen kanssa, että kuusen levinneisyyden raja vuoden 2000 ekr. tienoilla vastaa jotakuinkin tammikuun -8 C keskilämpötilan rajaa (Äyräpää 1939) on katsottu merkitsevän kulttuurin asettumista maatalouden kannalta otollisimmalle alueelle. Todellisuudessa kuusen leviämisen raja (Korhola 1990; Giesecke & Bennett 2004) ja nuorakeramiikan levinneisyyden raja (ks. kuva 1) kohtaavat vasta Suomen nuorakeraamisen kulttuurin (3200 2350 cal BC) loppuvaiheessa ja tuolloinkin vain osin kulttuurin reuna-alueella Hämeessä ja Kaakkois-Suomessa (ks. myös Siiriäinen 1981, 24; Edgren 1984, 13). Näin syy siihen, että nuorakeramiikan leviämisraja vakiintuu tiettyyn paikkaan, ei löydy kuusen leviämisestä. Nuorakeramiikan levinneisyyttä ei voi selittää myöskään alueellisella kasvukauden pituudella eikä muilla ilmastollisilla seikoilla (ks. Solantie 2005), sillä nuorakeramiikan aikaisia maanviljelyn merkkejä on nuorakeramiikan pohjoisrajaa huomattavasti pohjoisempaa, myös havumetsävyöhykkeeltä. Huomiota kannattaa kiinnittää myös siihen, että kuusi leviää Baltiaan ja Länsi-Venäjälle huomattavasti Suomea aikaisemmin (Giesecke & Bennett 2004) eikä kuusen esiintyminen näillä alueilla ole ollut esteenä nuorakeraamisten kulttuurien leviämiselle. Kuusen leviämisen, tammikuun lauhkeimpien keskilämpötilojen ja nuorakeramiikan levintäkartta on luonut jokaisen suomalaisen arkeologin päähän voimakkaan mielikuvan maanviljelyskulttuurista, joka on levittäytynyt myös Suomen suotuisimmille viljelysseuduille. Kun samansuuntainen kuva on nähtävissä myös rautakauden lopun (nykyisin tunnetun) kiinteän asutuksen levinnässä, on helppo ajatella, ettei kivikautiselle maanviljelylle olisi ollut mahdollisuuksia nuorakeraamisen kulttuurin levintäalueen pohjoispuolella. Kuitenkin lämpökauden aikana 6000 2500 cal BC lauhkean lehtimetsävyöhykkeen raja on kulkenut jotakuinkin leveyspiirillä 63 N (Heikkilä & Seppä 2003) eli likimain Vaasan korkeudella (kuva 2). Tämä näkyy hyvin myös nykyisessä lehmuksen levinnässä, sillä Suomen nykyilmastossa lehmus ei lisäänny siemenestä vaan ainoastaan uudentuu juuriversojen avulla. Kuusen leviäminen idästä Suomeen n. 4500 cal BC alkaen eteni siten, että noin 2500 cal BC Etelä-Suomen itäosa oli jo osa boreaalista havumetsävyöhykettä samalla kun länsiosa oli vielä lehtisekametsää. Kuusen katsotaan levinneen tietylle alueelle, kun sen osuus puiden siitepölyistä nousee 5 10 % tasolle. Leviämisen jälkeen kuuselta kesti 100 550 vuotta saavuttaa nykyistä vastaava osuus puustosta. Tässä kilpailussa häviäjinä olivat erityisesti lehmus- ja pähkinäpensasalueet (Seppä et al. 2009). Toinen tässä käsiteltävä asia on se, miksi viljan siitepölyjä löytyy usein samasta aikahorisontista, jossa näkyy kuusen leviäminen (Vuorela 1999a, 343). Maa- 23

Kuva 2. Nykyinen lehmuksen levinneisyys harmaalla (Lampinen & Lahti 2008) ja lämpökauden 6000 2500 cal BC aikaisen lauhkean lehtimetsävyöhykkeen suuntaa-antava pohjoisraja leveyspiirillä 63º N (Heikkilä & Seppä 2003). Ylempi katkoviiva on nykyinen napapiiri. Fig. 2. The present distribution of linden (Tilia cordata) in grey (Lampinen & Lahti 2008) and a indicative north limit of the boreo-nemoral zone during the Holocene thermal maximum ca. 6200 2500 cal BC along 63º N (Heikkilä & Seppä 2003). The upper dashed line marks the current Arctic Circle. talouden harjoittamisen kannalta tässä ei ole mitään järkeä, sillä leviävä kuusi nimenomaan valtaa varhaisen maatalouden kannalta otollisimmat lehmus-pähkinäpensasalueet huonontaen muun kasvillisuuden seassa tehtävän pienimuotoisen maanviljelyn mahdollisuuksia. Yhtenä selitysvaihtoehtona arvelin, että kuusen leviämisen alku olisi pyritty selvittämään muuta diagrammia tarkemmin, ja siksi myös vähäisten viljan siitepölyjen löytyminen tästä horisontista olisi todennäköisempää. Näin ei tilanne kuitenkaan ole muissa kuin nimenomaan metsähistoriaan keskittyvissä tutkimuksissa (Vuorela pers. comm.) Toisaalta havainnossa kuusen yleistymisestä ja ensimmäisistä siitepölyistä samassa aikahorisontissa saattaa piillä myös osittainen totuuden siemen, sillä ihmisen nimenomaan lehmus-pähkinäpensaskasvustoon kohdistuva, varhaiseen maanviljelyyn liittyvä raivaustoiminta on omalta osaltaan varmasti auttanut ja nopeuttanut kuusen leviämistä ja yleistymistä (Huttunen 1980). Syynä tälle ilmiölle voi olla myös se, että kuusen leviäminen on muuttanut olosuhteita siten, ettei viljely muun kasvillisuuden seassa ole ollut enää mahdollista. Kuusi on saattanut pakottaa aiempaa voimaperäisempään raivaamiseen, mikä on saattanut helpottaa siitepölyjen leviämistä ja sitä kautta parantanut mahdollisuuksia viljelyn havaitsemiseen siitepölyissä. Kuusen leviäminen ei ole ollut esteenä maanviljelylle. Nuorakeraaminen kulttuuri ja kuusi kohtaavat Suomessa vain pienellä alueella ja ajallisesti lyhyen ajan ko. kulttuurin loppuvaiheessa. Siksi nuorakeramiikan levinneisyys ei selity kuusen yleistymisellä. Nuorakeraamisen kulttuurin pohjoispuolelta on löydettävissä myös maataloudelle sopivat olosuhteet. Tämä on näkyvissä selkeästi myös varhaisten viljan siitepölyjen levinnässä (kuva 1). Kuusen leviämisen ja varhaisen maanviljelyn suhde on yhä arvoitus. On mahdollista, ettei kyseessä ole puhdas sattuma ja että nämä asiat todella kietoutuvat yhteen. 24

Kampakeraamisen kulttuurin kontaktiverkostot Pohjoisen metsävyöhykkeen laajat yhteysverkostot tulevat erityisen hyvin näkyviin tyypillisen kampakeramiikan aikana (4000 3400 cal BC). Itämeren eteläosan meripihka, Luoteis-Venäjän piikivi, Äänisen kupari ja Uralin alueen sembramännystä valmistetut lusikat ja jalakset leviävät tänä aikana laajalle alueelle. Samaan aikaan itse ns. kampa-kuoppakeraamiset kulttuurit ovat levinneet laajimmilleen ja ovat etelässä jo kosketuksissa maanviljelykulttuurien kanssa. 2000-luvulla tehdyn tutkimuksen mukaan Baltian alueen tyypillisen kampakeraamisen kulttuurin katsotaan harjoittavan jo maanviljelyä (Kriiska 2003). Tyypillisen kampakeramiikan aikana selvästi näkyvät kontaktit pysyvät aktiivisina myös myöhemmin kolmannella vuosituhannella eaa. (esim. Edgren 1992, 68 70; Carpelan 1999; Zvelebil & Lillie 2000; Zvelebil 2006; Zhulnikov 2008). Nämä kontaktiverkostot toimivat alueella, jolla on maataloutta harjoittavia kulttuureita. Ainakin Virossa maanviljely liitetään jo osaksi tyypillisen kampakeramiikan edustamaa kulttuurivaihetta. Vaikuttaisi oudolta, jos maanviljely ei olisi levinnyt näiden kontaktien myötä myös Suomeen jo tyypillisen kampakeramiikan aikana. Tänä aikana ainakaan luonnonolot eivät ole olleet esteenä maanviljelyn omaksumiselle. Yhteenveto Uusimpien ajoitusten perusteella Baltian pohjoisosien (ja todennäköisesti myös Suomen) maatalouden alkusysäys ei tule eteläiseltä Itämereltä vaan todennäköisemmin idempää ja kauan ennen eteläisen maatalouskulttuurin, tässä tapauksessa nuorakeraamisen kulttuurin, leviämistä alueelle. Itäistä tulosuuntaa on ehdotettu aiemminkin Sisä-Suomen maatalouden saapumissuunnaksi, mutta liittyen varhaismetallikautiseen tekstiilikeramiikkaan yhdistettyihin viljan siitepölylöytöihin (Meinander 1984; Taavitsainen et al. 1998). Nykyisessä tutkimustilanteessa Suomen vanhimmat kiistattomat maanviljelyn ajoitukset ovat tuhannesta kahteentuhanteen vuotta nuorempia kuin lähialueiden vanhimpien viljansiitepölyjen perusteella voisi olettaa. Artikkelin alussa mainitut, vuonna 1980 julkaistut (Reynaud & Hjelmroos 1980) Utajärven ja Raahen ajoitukset vaikuttavat nykytutkimuksen kautta katsottuina suorastaan juuri siltä, mitä olisi syytä odottaa. On huomionarvoista, että vanhimmat viljan siitepölyt löytyvät selkeästi nuorakeraamisen kulttuurin leviämisalueen pohjoispuolelta. Nykyisessä tutkimustilanteessa Puolangan 2900 2800 cal BC ajoittuva ensimmäinen viljan siitepöly ja heikot raivauksen merkit (Vuorela 1998; 1999a; 2002) eivät tunnu enää niin sensaatiomaisilta. Oletan maanviljelyn omaksumisen tapahtuneen tyypillisen kampakeramiikan aikana saman kontaktiverkoston välityksellä, joka levitti piitä, meripihkaa ja kuparia laajalle alueelle Pohjois-Itämerellä ja Itä-Euroopan havumetsävyöhykkeellä. 25

Varhainen viljely ei ole ollut intensiivistä, vaan oletettavasti pienimuotoista, muun kasvillisuuden sekaan tehdyillä raivioilla harjoitettua viljelyä. Oletettavasti maanviljely on ollut alkuvaiheessa hyvin pesäkkeistä, kuten etelämpänä Euroopassa. Suomen (sub)neoliittisten kulttuurien osalta tämä tarkoittaa sitä, että vain osa tietyn materiaalisen kulttuurin omaavista yhteisöistä on saattanut harjoittaa sivutoimista maanviljelyä, metsästyksen ja kalastuksen ollessa yhä pääelinkeinona. Nuorakeraamisen kulttuurin levintäalue yhdistettynä kuusen leviämiseen ja moderneihin ilmastotutkimusten tuloksiin on luonut voimakkaan kuvan tämän kulttuurin rajoittumisesta maatalouden kannalta otolliselle alueelle. Tätä kuvaa on vahvistanut vielä se, että kiinteän maanviljelyasutuksen raja rautakauden lopulla kulkee samalla alueella. Kivikauden Suomessa maatalouden harjoittaminen on ollut kuitenkin mahdollista myös nuorakeraamisen kulttuurin pohjoispuolella varsinkin kivikauden lämpökauden lauhkean lehtimetsävyöhykkeen alueella, joka näkyy vielä nykyisin lehmuksen levinnässä. Tässä artikkelissa esitettyjä muutoksia tavassa asua ei voi suoraan liittää maanviljelyyn. Kuitenkin Suomen lähialueilta saadut entistä vanhemmat maanviljelyn alun tulokset yhdessä Kainuun ja Pohjois-Pohjanmaan varhaisten viljan siitepölyjen kanssa antavat syyn olettaa, että maatalouden harjoittaminen on mahdollista jo tyypillisen kampakeramiikan aikana. Siksi on täysin mahdollista, että asutuksessa tapahtuneet muutokset liittyvät tavalla tai toisella prosesseihin, joissa viljely oli jo mukana. Ainakin kylämäisen asutuksen synty ja asumuspainanteiden sijoittamisen muutokset pitempiaikaista asutusta ja säilöntää suosiviin paikkoihin tapahtuvat samaan aikaan kun Virossa ja Äänisen länsirannalla on jo harjoitettu maanviljelyä. Uudet (sub)neoliittisten kulttuurien piiristä saadut, aiempaa huomattavasti vanhemmat maanviljelyn ajoitukset pakottavat miettimään viljelyn leviämistä uudestaan. Baltian pohjoisosien osalta on jo selvää, ettei viljely levinnyt alueelle vasta nuorakeraamisen kulttuurin myötä vaan jo huomattavasti aiemmin (Kriiska 2003). Uudet tulokset edellyttävät koko subneoliittisten kulttuurien uudelleen arviointia juuri tuottavien maatalouselinkeinojen puuttuminen on erottanut subneoliitiset kulttuurit muista neoliittisista kulttuureista. Artikkelissa oletetaan Suomen varhaisimman viljelyn tapahtuneen pienillä, muun kasvillisuuden sekaan raivatuilla alueilla, samoin kuin muilla maatalouden kannalta perifeerisillä alueilla. Tämän takia varhaisimpien viljan siitepölyjen yhteydessä havaittavat metsänraivauksen merkit ovat yleensä vähäisiä ja heikosti erottuvia, ainakin verrattaessa niitä intensiivisemmän peltoviljelyn aikaansaamiin kasvillisuusmuutoksiin. Oletan intensiivisemmän maatalouden saapuneen Suomeen, samoin kuin Viroon, nuorakeraamisen kulttuurin myötä. Mikäli olettamus nuorakeramiikkaa vanhemmasta maanviljelystä Suomessa pitää paikkansa, tällä on mielenkiintoisia seuraamuksia nuorakeramiikan aikaisen kulttuurikehityksen kannalta. Kampa-, asbesti- ja nuorakeraamiset kulttuurit eivät ehkä sittenkään olleet kaikilta osin niin erilaisia kuin aiemmassa tutkimuksessa on oletettu. 26

Täysin varmojen, ajoitukseltaan vanhempien kuin 3000 cal BC viljan siitepölyjen puuttuminen Suomen tämänhetkisestä aineistosta on tunnustettava tosiasia. Menneisyyttä tutkittaessa löytää kuitenkin usein vain sitä, mitä osaa etsiä tai mitä on suuntautunut etsimään. Kivikauden maanviljelyn merkkien puuttumista ei voi perustella jo tehtyjen siitepölyanalyysien perusteella, mikäli näiden tutkimusten painopiste on ollut rautakauden tutkimuksessa (esim. Grönlund 1995). Kivikautisen, nykytutkimuksessa esitettyä vanhemman maanviljelyn olemassaolosta näkyy jo positiivisia viitteitä, mikäli siihen haluaa uskoa. Esimerkiksi Keski-Suomessa sijaitsevan Hankasalmen Särkilammen näytesarjassa on nähtävissä tulella tehtyä metsien raivausta tähän yhdistyvät häiriöt puulajien osuuksissa, niittylajiston määrän kasvu ja muiden asutusindikaattorien määrän kaksinkertaistuminen jotakuinkin tyypillistä ja myöhäiskampakeramiikkaa vastaavana aikana n. 3800 3200 cal BC (Vuorela 1995). Varhaisten metsänraivausvaiheiden liittäminen maanviljelyyn voidaan varmistaa uusilla tutkimuksilla. Vaikka ensimmäisen viljan siitepölyn löytäminen on osittain onnenkauppaa, tutkijalla on mahdollisuus tehdä valintoja, joilla onnistumisen todennäköisyyttä voidaan parantaa. Varhainen maanviljely kaipaisi omaa tutkimusprojektiansa, jossa varhaisimpia viljan siitepölyjä pyrittäisiin löytämään esimerkiksi laskettavien siitepölyjen määrää kasvattamalla. Todettakoon vielä, että nykyistä varhaisemman maanviljelyn todistaminen tavalla, joka vakuuttaisi epäluuloisimmatkin tutkijat, vaatisi siitepölyjen rinnalle myös muita todisteita, kuten viljansiementen makrofossiileja ja/tai jyväpainaumia keramiikassa. Jälkikirjoitus Artikkelin jättämisen jälkeen tietooni on tullut yksi uusi kivikautista maanviljelyä koskeva tutkimustulos. Nurmeksen Lehmilammesta otetuissa siitepölydiagrammeissa ensimmäiset ihmistoiminnan merkit ilmaantuvat välillä 3000 2500 kalenterivuotta ennen ajanlaskun alkua (Augustsson et al., käsikirjoitus). Kohteen tulokset ovat erityisen mielenkiintoisia, sillä ajoitukset eivät perustu radiohiiliajoituksiin vaan vuosilustoihin. Näin ollen havaintojen ajoitusta ei ole mahdollista epäillä liian vanhaksi esimerkiksi radiohiiliajoituksessa käytetyn näytteen kontaminaation tai muun häiriön takia. Nurmeksen Lehmilammen siitepölyprofiilissa vanhin viljan siitepöly (Hordeum t.) esiintyy 3000 cal BC. Samaan vaiheeseen liittyy muitakin ihmistoiminnan merkkejä. Tuhkaa esiintyy runsaasti aikavälillä 3100 2800 cal BC. Välillä 3000 2500 cal BC aineistossa esiintyy säännöllisesti pieniä määriä seuraavia kulttuuri-indikaattorilajeja: aho-/niittysuolaheinä, pujo, piharatamo, savikkakasvit, hienäkasvit, niittyleinikki ja kataja (Augustsson et al., käsikirjoitus). Tutkimuksen tekijät ovat tulkinneet tulokset merkeiksi lammen ympäristössä harjoitetusta laiduntamisesta ja pienimuotoisesta maanviljelystä. Nurmeksen tulos kiilaa Suomen viiden vanhimman viljan siitepölylöydön joukkoon. Vanhem- 27

pia havaintoja viljan siitepölyistä on esitetty ainoastaan Reynaudin ja Hjelmroosin (1980) kiistanalaisessa artikkelissa. Nurmeksen Lehmilammen tulokset vahvistavat kuvaa nuorakeraamisen kulttuurin esiintymisaluetta pohjoisemman kivikautisen maanviljelyn olemassaolosta. Kiitokset Esitän kiitokseni geologi Teija Aleniukselle (FT) ja arkeologi Vesa-Pekka Hervalle (FT) artikkelin syntyä edesauttaneista keskusteluista ja käsikirjoituksen kommentoinnista. Kiitokset Irmeli Vuorelalle (FT) 4.8.2009 käydystä puhelinkeskustelusta liittyen kuusen leviämisen yhteydestä löydettyihin ensimmäisiin viljan siitepölyihin. Kiitokset Liisa Lohtanderille (FM) artikkelin oikoluvusta ja Jarmo Kankaanpäälle (Ph.D.) viimetingan kommenteista ja englanninkielisen tiivistelmän kielenhuollosta. Lopuksi vielä kiitos professori Marie-Jose Gaillardille (Kalmarin yliopisto, Ruotsi) heidän tutkimusryhmänsä tulosten (Augustsson et al. käsikirjoitus) antamisesta käyttööni. Tämä tutkimus on tehty Arkeologian valtakunnallisessa tutkijakoulussa (2006 2009). Lähteet Tutkimusraportit Holback, T., Lindholm, P. & Runeson, H. 2004. Bjästamon. Ett kustbundet boplatskomplex från slutet av neolitikum, Bottniabanan, Västernorrland. Rapport 2004:1. Riksantikvarieämbetet, UV mitt, Stockholm. Vuorela, I. 1994. Keski-Suomen asutushistoriaa Keuruun ja Jämsän järvikerrostumien arkistossa. Geologian tutkimuskeskus, Maaperäosasto, Tutkimusraportti P 34.4.111. Vuorela, I. 1995. Keski-Suomen asutushistoriaa Hankasalmen ja Saarijärven järvikerrostumien arkistoissa. Geologian tutkimuskeskus, Maaperäosasto, Tutkimusraportti P 34.4.114. Vuorela, I. 1997. Pihtiputaan Karvalampi paleoekologisen tutkimuksen kohteena. Geologian tutkimuskeskus, Maaperäosasto, Tutkimusraportti P 34.4.119. Vuorela, I. & Kankainen, T. 1991. Siitepölyanalyyttinen tutkimus asutuksen vaikutuksesta kasvillisuuteen Puolangan kunnan Kotilan kylässä. Geologian tutkimuskeskus, Maaperäosasto, Tutkimusraportti P 34.4.100. Vuorela, I. & Kankainen, T. 1992. Tammelan asutushistoria Kuivajärven kerrostumien paleoekologisen tutkimuksen valossa. Geologian tutkimuskeskus, Maaperäosasto, Tutkimusraportti P 34.4.104. Vuorela, I. & Kankainen, T. 1993. Luonnon- ja kulttuurimaiseman kehitys Taipalsaaressa. Geologian tutkimuskeskus, Maaperäosasto, Tutkimusraportti P 34.4.107. 28

Vuorela, I., Uutela, A, Saarnisto, M., Ilmasti, M. & Kankainen, T. 1993. Vuojärven ja Antinlammen kerrostumat Laukaan asutus- ja luonnonhistorian arkistona. Geologian tutkimuskeskus, Maaperäosasto, Tutkimusraportti P 34.4.108. Kirjallisuus Aalto, M., Siiriäinen, A. & Vuorela, I. 1985. Humppila Järvensuo a preinvestigation for an archaeological and palaeobotanical project in SW Finland. Iskos 5, 165 177. Alenius, T. 2008. The palaeoecological study of three mires on the island Kemiönsaari, SW Finland. Appendix 2 teoksessa H. Asplund 2008. Kimittæ. Sites, centrality and long-term settlement change in the Kemiönsaari region in SW Finland. (Turun yliopiston julkaisuja, Sarja B 312, Humaniora). Turun yliopisto, Turku, 565 584. Alenius, T., Mikkola, E. & Ojala, A.E.K. 2008. History of agriculture in Mikkeli Orijärvi, eastern Finland as reflected by palynological and archaeological data. Vegetation History and Archaeobotany, Vol. 17, No. 2, 171 183. Alenius, T. & Lavento, M. 2007. 1.2 Katajajärvi osa sisämaan asutus- ja viljelyhistoriaa. Teoksessa A. Lahelma & M. Lavento (toim.), Sama maisema, eri kulkijat Repoveden kansallispuisto kivikaudelta 1900-luvulle. (Metsähallituksen luonnonsuojelujulkaisuja. Sarja A 165). Metsähallitus, Helsinki, 14 19. Alenius, T., Lavento, M. & Saarnisto, M. 2009. Pollen-analytical results from Lake Katajajärvi Aspects of the history of settlement in the Finnish inland region. Julkaistaan Acta Borealia 26, No. 2, 136 155. Alhonen, P. 1988. Tampereen luonnonympäristön kehitysvaiheet. Teoksessa P. Alhonen, U. Salo, S. Suvanto & V. Rasila (toim.), Tampereen historia 1. Tampereen kaupunki, Tampere, 3 50. Ames, K.M. 2003. The Northwest Coast. Evolutionary Anthropology 19, 19 33. Augustsson A., Gaillard M.J., Peltola P., Mazier F., Haltia-Hovi E., Bergbäck B. & Saarinen T. käsikirjoitus. Mid- and Late Holocene sediment chemistry of a small lake in northern Karelia, Finland - the role of forest dynamics, human impact and climate change. Tullaan lähettämään julkaistavaksi julkaisusarjaan Journal of Paleolimnology. Behre, K.-E. 2007. Evidence for Mesolithic agriculture in and around central Europe? Vegetation History and Archaeobotany 16, 203 219. Behre, K.-E. 2008. Comment on: Mesolithic agriculture in Switzerland? A critical review of the evidence by W. Tinner, E-H Nielsen and A.F. Lotter. Quanternary Science Reviews 27, 1467 1468. Björk, N. 2003. The Neolithic coastal settlements cosmology and ideology in a Baltic Sea perspective. Teoksessa C. Samuelsson & N. Ytterberg (eds.), Uniting Sea. Stone Age Societies in the Baltic Sea region. (Occasional Papers in Archaeology 33). Uppsala universitet, Uppsala, 20 42. Björk, N. & Larsson, F. 2007. Det Neolitiska samhället i Östra Mellansverige. Teoksessa N. Stenbäck (red.), Stenåldern i Uppland, Uppdragsarkeologi och eftertanke. (Volum 1, Arkeologi E4 Uppland studier). Riksantivarieämbetet, Uppsala, 67 96. Bocquet-Appel, J.-P., Naji, S., Vander Linden, M. & Kozlowski, J.K. 2009. Detection of diffusion and contact zones of early farming in Europe from the space-time distribution of 14C dates. Journal of Archaeological Science 36, 807 820. 29

Borić, D. 2005. Deconstructing essentialisms: unsettling frontiers of the Mesolithic-Neolithic Balkans. Teoksessa D. Bailey, A. Whittle & V. Cummings (eds.), (un)settling the Neolithic. Oxbow books, Oxford, 16 31. Burden, E. T., McAndrews, J. H. & Norris, G. 1986. Palynology of Indian and European forest clearance and farming in lake sediment cores from Awenda Provincial Park, Ontario. Canadian Journal of Earth Sciences, vol. 23, No. 1, 43 54. Carpelan, C. 1999. Käännekohtia Suomen esihistoriassa aikavälillä 5100 1000 ekr. Teoksessa P. Fogelberg (toim.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. (Bidrag till kännedom av Finlands natur och folk 135). Suomen Tiedeseura, Helsinki, 249 280. Carpelan, C. 2002. Arkeologiset löydöt aikaportaina. Teoksessa R. Grünthal (toim.), Ennen, muinoin. Miten menneisyyttämme tutkitaan. (Tietolipas 180). Suomalaisen Kirjallisuuden Seura, Helsinki, 188 212. Davidson, K., Dolukhanov, P.M., Sarson, G.R., Shukurov, A. & Zaitseva, G.I. 2009. Multiple sources of the European Neolithic; Mathematical modelling constrained by radiocarbon dates. Quaternary International, Volume 203, Issues 1 2, 10 18. Edgren, T. 1970. Studier över den snörkeramiska kulturens keramik i Finland. (Suomen Muinaismuistoyhdistyksen aikakauskirja 72). Suomen muinaismuistoyhdistys, Helsinki. Edgren, T. 1984. On the economy and the subsistence of the Battle-Axe Culture in Finland. Iskos 4, 9 15. Edgren, T. 1992. Den förhistoriska tiden. Teoksessa M. Norrback (red.), Finlands historia 1. Schildts, Esbo, 9 270. Edgren, T. 1999. Käännekohtia Suomen kivikaudessa. Teoksessa P. Fogelberg (toim.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. (Bidrag till kännedom av Finlands natur och folk 135). Suomen Tiedeseura, Helsinki, 281 293. Giesecke, T. & Bennett, K.D. 2004. The Holosene spread of Picea abies (L.) Karts. Teoksessa Fennoscandia and adjacent areas. Journal of Biogeography, Vol. 31, 1523 1548. Gilman, P. 1987. Architecture as artifact: pit structures and pueblos in the American Southwest. American Antiquity, Vol. 52, N.o 3, 538 564. Gkiasta, M., Russell, T., Shennan, S. & Steele, J. 2003. Neolithic transition in Europe: the radiocarbon record revisited. Antiquity 77, 45 62. Grönlund, E. 1995. A palaeoecological study of land-use history in East Finland. (Joensuun yliopiston luonnontieteellisiä julkaisuja, No 31). Joensuun yliopisto, Joensuu. Göransson, H. 1998. Neolithic Man and the Forest Environment around Alvastra Pile Dwelling. (Thesis and Papers in North-European Archaeology, 20). Lund University Press, Lund. Hallgren, F. 2008. Identitet i praktik. Lokala, regionala och överregionala sociala sammanhang inom nordlig trattbägarkultur. (Coast to Coast-books 17). Uppsala Universitet, Uppsala. Halinen, P. & Mökkönen, T. 2009. Between Lake and Sea Stone Age settlement by Ancient Lake Ladoga on the Karelian Isthmus. Fennoscandia Archaeologica XXVI, 107 132. Heikkilä, M. & Seppä, H. 2003. A 11,000 yr palaeotemperature reconstruction from the southern boreal zone in Finland. Quaternary Science Reviews 22, 541 554. Huttunen, P. 1980. Early land use, especially the slash-and-burn cultivation in the commune of Lammi, southern Finland, interpreted mainly using pollen and charcoal analyses. Acta Botanica Fennica 133, 1 45. 30

Huurre, M. 1983. Pohjanmaan ja Lapin esihistoria. Pohjois-Pohjanmaan ja Lapin historia 1. Pohjois-Pohjanmaan maakuntaliitto ja Lapin maakuntaliitto, Oulu. Ingold, T. 2000. The perception of the environment. Essays in livelihood, dwelling and skill. Routledge, London & New York. Karlsson, S. 2007. Stenåldersvegetation i Uppland, Östra Svealand. Teoksessa N. Stenbäck (red.), Stenåldern i Uppland, Uppdragsarkeologi och eftertanke. (Volum 1, Arkeologi E4 Uppland studier). Riksantivarieämbetet, Uppsala, 137 153. Katiskoski, K. 2002. The semisubterranean dwelling at the Kärmelahti site in Puumala, Savo province, Eastern Finland. Teoksessa H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 171 200. Korhola, A. 1990. Suomen metsien kehitysvaiheet. Terra 102:4, 268 274. Kriiska, A. 2003. From hunter-fisher-gatherer to farmer Changes in the Neolithic economy and settlement on Estonian territory. Archeologia Lituana 2003, Vol. 4, 11 26. Kotsakis, K. 2005. Across the border: unstable dwellings and fluid landscapes in the earliest Neolithic of Greece. Teoksessa D. Bailey, A. Whittle & V. Cummings (eds.), (un) settling the Neolithic. Oxbow books, Oxford, 8 15. Königsson, L.K., Possnert, G. & Hammar, T. 1997. Economical and cultural changes in the landscape development at Novgorod, Russia. Tor 29, 353 382. Lang, V. 1999. Pre-Christian history of farming in the Eastern Baltic Region and Finland: a synthesis. Teoksessa U. Miller, T. Hackens, V. Lang, A. Raukas, & S. Hicks (eds.), Environmental and Cultural History of the Eastern Baltic Region. (PACT 57). Conseil de l Europe, Rixensart, 359 372. Lampinen, R. & Lahti, T. 2008. Kasviatlas 2007. Helsingin yliopisto, Luonnontieteellinen keskusmuseo, Kasvimuseo, Helsinki. Levinneisyyskartat osoitteessa <http://www.luomus.fi/kasviatlas>. Lempiäinen, T. 1999. Hiiltyneet viljanjyvät ja maanviljelyn alku Suomessa. Teoksessa P. Fogelberg (toim.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. (Bidrag till kännedom av Finlands natur och folk 135). Suomen Tiedeseura, Helsinki, 152 154. Luoto, J. 1986. Problem inom Finlands mellanneolitikum. Finskt Museum 1986, 9 21. McAndrews, J.H. & Boyko-Diakonov, M. 1989. Pollen analysis of varved sediment at Crawford Lake Ontario: evidence of Indian and European farming. Teoksessa R.J. Fulton (ed.), Quaternary Geology of Canada and Greenland, Geological Survey of Canada, Geology of Canada No. 1, 528 530. McGuire, R.H. & Schiffer, M.B. 1983. A Theory of architectural design. Journal of Anthropological Archaeology 2, 277 303. Meinander, C.F. 1984. Om introduktionen av sädesodling i Finland. Finskt Museum 1983, 5 20. Murdock, G.P. 1967. Ethnographic atlas. University of Pittsburgh Press, Pittsburgh. Mökkönen, T. 2002. Chronological variation in the locations of hunter-gatherer occupation sites vis-à-vis the environment. Teoksessa H. Ranta (ed.). Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 53 64. Mökkönen, T. 2008. A Review of Neolithic multi-room housepits as seen from the Meskäärtty site in Virolahti parish, extreme south-eastern Finland. Estonian Journal of Archaeology, Vol. 12, No. 2, 114 151. 31

Mökkönen, T. 2009. Neolithic housepits in the River Vuoksi Valley, Karelian Isthmus, Russia chronological changes in size and location. Fennoscandia Archaeologica XXVI, 133 161. Nunez, M. 1999. Role of food production in Stone Age Finland. Teoksessa P. Fogelberg (toim.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. (Bidrag till kännedom av Finlands natur och folk 135). Suomen Tiedeseura, Helsinki, 133 142. Núñez, M. 2004. All Quiet on the Eastern Front? Teoksessa H. Knutsson (ed.), Coast to Coast Arrival. Results and Reflections. Uppsala University, Uppsala, 345 367. Norberg, E. 2008. Boplatsvallen som bostad i Norrbottens kustland 5000 till 2000 före tideräkning. En studie av kontinuitet och förändringar. (Studia Archaeologica Universitatis Umensis 23). Umeå Universitet, Umeå. Pesonen, P. 2002. Semisubterranean houses in Finland a review. Teoksessa H. Ranta (ed.), Huts and Houses. Stone Age and Early Metal Age Buildings in Finland. Finnish National Board of Antiquities, Helsinki, 9 41. Pesonen, P. 2008. Mitä uutta Lounais-Hämeestä? Tuloksia vuosien 2001 2006 inventoinneista. Muinaistutkija 2/ 2008, 19 32. Pluciennik, M. 2008. Hunter-Gatherers to Farmers? Teoksessa A. Jones (ed.), Prehistoric Europe. Theory and practice. Blackwell Publishing, Oxford, 16 34. Poska, A. & Saarse, L. 2006. New evidence of possible crop introduction to north-eastern Europe during Stone Age. Cerealia pollen finds in connection with the Akali Neolithic settlement, East Estonia. Vegetation History and Archaeobotany 15, 169 179. Poska, A., Saarse, L., Veski, S. & Kihno, K. 1999. Farming from the Neolithic to the Pre- Roman Iron Age in Estonia, as Reflected in Pollen Diagrams. Teoksessa U. Miller, T. Hackens, V. Lang, A. Raukas, & S. Hicks (eds.), Environmental and Cultural History of the Eastern Baltic Region. (PACT 57). Conseil de l Europe, Rixensart, 303 317. Poska, A., Saarse, L. & Veski, S. 2004. Reflections of pre- and early-agrarian human impact in the pollen diagrams of Estonia. Palaeogeography, Palaeoclimatology, Palaeoecology 209 (2004), 37 50. Rafferty, J. E. 1985. The Archaeological record on sedentariness: Recognition, development, and implications. Teoksessa M. B. Schiffer (ed.), Advances in Archaeological Method and Theory, Vol. 8. Academy Press, New York, 113 156. Raska-Jasiewiczowa, M. & Latałowa, M. 1996. Poland. Teoksessa B.E. Berglund, H.J.B. Birks, M. Ralska-Jasiewiczowa and H.E. Wright (eds.), Palaeoecological events during the last 15 000 years. Regional synthesis of palaeoecological studies of lakes and mires in Europe. Wiley, Chichester, 403 472 Reynaud, C. & Hjelmroos, M. 1980. Myöhäiseltä mesoliittiselta ajalta lähtien siitepölyanalyysillä todistettu ja radiohiilimenetelmällä ajoitettu ihmisen vaikutus luontaiseen metsäkasvillisuuteen Pohjois-Pohjanmaan alueella. Faravid 4 (1980), 41 75. Rimantienė, R. 1992. The Neolithic of the Eastern Baltic. Journal of World Prehistory, Vol. 6. N.o 1, 97 143. Robb, J. & Miracle, P. 2007. Beyond migration versus acculturation: new models for the spread of agriculture. Teoksessa A. Whittle & V. Cummings (eds.), Going Over: The Mesolithic-Neolithic Transition in North-West Europe. (Proceedings of the British Academy 144). Oxford University Press, Oxford, 99 115. Rowley-Conwy, P. 2004. How the west was lost. A reconsideration of argricultural origins in Britain, Ireland, and Southern Scandinavia. Current Anthropology, Vol. 45, Supplement. S83 S113. 32

Saastamoinen, S. 1996. Prehistoric settlement in the Sätös area in Eastern Finland reflected in a pollen analysis made from the sediments of lake Saari-Oskamo. Teoksessa T. Kirkinen (ed.), Pithouses and Potmakers in Eastern Finland. Reports of the the Ancient Lake Saimaa Project. (Helsinki papers in archaeology, N:o 9). Helsingin yliopisto, Helsinki, 119 134. Saastamoinen, S. 1999. Investigating Prehistoric Economy with Pollen Analysis. The Early Metal Age dwelling site at Kitulansuo in Ristiina, East Finland. Teoksessa M. Huurre (ed.), Dig it all. Papers dedicated to Ari Siiriäinen. Jyväskylä. Finnish Antiquarian Society, Helsinki, 131 148. Salo, U. 1997. Ihmisen jäljet Satakunnan maisemassa. Kulttuurimaiseman vuosituhannet. Suomalaisen Kirjallisuuden Seura, Helsinki. Seppä, H., Alenius, T., Bradshaw, R.H.W., Giesecke, T., Heikkilä, M. & Muukkonen, P. 2009. Invasion of Norway spruce (Picea abies) and the rise of the boreal ecosystem in Fennoscandia. Journal of Ecology 2009, 97, 629 640. Siiriäinen, A. 1981. On cultural ecology of the Finnish Stone Age. Suomen Museo 1980, 5 40. Siiriäinen, A. 2003. Esihistoria. Teoksessa S. Zetterberg (toim.), Suomen historian pikkujättiläinen. WSOY, Porvoo, 14 43. Simola, H. 1998. Siitepölyanalyysin mahdollisuudet ja rajat. Teoksessa P. Fogelberg (toim.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. (Bidrag till kännedom av Finlands natur och folk 135). Suomen Tiedeseura, Helsinki, 155 158. Solantie, R. 2005. Aspects of some prehistoric cultures in relation to climate in Southwestern Finland. Fennoscandia Archaeologica XXII, 28 42. Stančikaitė, M., Daugnora, L., Hjelle, K. & Hufthammer, A.K. 2009. The environment of the Neolithic archaeological sites in Šventoji, Western Lithuania. Quaternary International (2009), doi:10.1016/j.quaint.2009.01.012, 1 13. Taavitsainen, J.-P., Simola, H. & Grönlund, E. 1998. Cultivation history beyond periphery: Early agriculture in the North European boreal forest. Journal of World Prehistory, Vol. 12, No. 2, 199 253. Thomas, J. 2001. Archaeologies of place and landscape. Teoksessa I. Hodder (ed.), Archaeological theory today. Polity Press, Cambridge, 141 164. Tilley, C. 1994. A Phenomenology of landscape. Places, paths and monuments. Berg Publishers, Oxford. Tinner, W., Nielsen, E.H. & Lotter, A.F. 2007. Mesolithic agriculture in Switzerland? A critical review of the evidence. Quanternary Science Reviews 26, 1416 1431. Tinner, W., Nielsen, E.H. & Lotter, A.F. 2008. Reply: Evidence for Late-Mesolithic agriculture? A reply to Karl-Ernst Behre. Quanternary Science Reviews 27, 1468 1470. Tipping, R., Bunting, M.J., Davies, A.L., Murray, H., Fraser, S. & McCulloch, R. 2009. Modelling land use around an early Neolithic timber hall in north east Scotland from high spatial resolution pollen analysis. Journal of Archaeological Science 36, 140 149. Tolonen, M. 1980. Postglacial pollen stratigraphy of lake Lamminjärvi, S. Finland. Annales Botanici Fennici 17, 15 25. Tolonen, M. 1987. Vegetational history in coastal SW Finland studied on a lake and a peat bog by pollen and charcoal analyses. Annales Botanici Fennici 24, 353 370. Vaneeckhout, S. 2008a. Sedentism on the Finnish Northwest Coast: shoreline reduction and reduced mobility. Fennoscandia Archaeologica XXV, 61 72. 33

Vaneeckhout, S. 2008b. Dwelling depressions at Kierikki: results of a GPS-survey. Faravid 32, 1 12. Vaneeckhout, S. 2009. 2500 years of social evolution in a northern Finnish Stone Age village. Teoksessa T. Mökkönen & S.-L. Seppälä (toim.), Arkeologipäivät 2008. Sosiaaliarkeologiaa yhteisöt arkeologisen aineiston takana & Tutkimushistorian painolasti. Suomen arkeologinen seura, Helsinki, 63 71. Vuorela, I. 1973. Relative pollen rain around cultivated fields. Acta Botanica Fennica 102, 1 27. Vuorela, I. 1990. Pollenanalytiska studier. Teoksessa K. Zilliacus (red.), Finska skären : studier i Åboländsk kulturhistoria. (Föreningen Konstsamfundets publikationsserie 7). Konstsamfundet, Helsingfors, 115 132. Vuorela, I. 1996. Palynological indication of the Stone Age dwelling site of Pörrinmökki, Rääkkylä, Eastern Finland. Teoksessa T. Kirkinen (ed.), Environmental Studies in Eastern Finland. Reports of the Ancient Lake Saimaa Project. (Helsinki Papers in Archaeology, No. 8). University of Helsinki, Helsinki, 129 142. Vuorela, I. 1998. The Transition to Farming in Southern Finland. Teoksessa M. Zvelebil, L. Domańska and R. Denenell (eds.), Harvesting the Sea, Farming the Forest. The Emergence of Neolithic Societies in the Baltic Region. (Sheffield Archaeological Monographs, 10), 175 180. Vuorela, I. 1999a. The Beginnings of Agricultural Land Use in Finland: an Assessment based on Palynological Data. Teoksessa U. Miller, T. Hackens, V. Lang, A. Raukas, & S. Hicks (eds.), Environmental and Cultural History of the Eastern Baltic Region. (PACT 57). Conseil de l Europe, Rixensart, 339 351. Vuorela, I. 1999b. Viljelytoiminnan alku Suomessa paleoekologisten tutkimusten kohteena. Teoksessa P. Fogelberg (toim.), Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. (Bidrag till kännedom ab Finlands natur och folk 135). Suomen Tiedeseura, Helsinki, 143 151. Vuorela, I. 2002. Luonnon kerrostumat säilövät menneisyyttä. Teoksessa R. Grünthal (toim.), Ennen, muinoin. Miten menneisyyttämme tutkitaan. (Tietolipas 180). Suomalaisen Kirjallisuuden Seura, Helsinki, 76 92. Vuorela, I. & Hicks, S. 1996. Human Impact on the Natural Landscape in Finland. A Review of the Pollen Evidence. Teoksessa A.-N. Rebertson et al. (eds.), Landscapes and Life. Studies in Honour of Urve Miller, (PACT 50), 245 257. Vuorela, I. & Lempiäinen, T. 1988. Archaeobotany of the site of the oldest cereal grain in Finland. Annales botanici Fennici 25, 33 45. Vuorela, I., Saarnisto, M., Lempiäinen, T. & Taavitsainen, J.-P. 2001. Stone Age to recent land-use history at Pegrema, northern Lake Onega, Russian Karelia. Vegetation History and Archaeobotany 10, 121 138. Welinder, S. 1998a. Del 1. Neoliticum Bronsålder, 3900 500 f.kr.. Teoksessa J. Myrdahl (red.), Det Svenska jordbrukets historia. Jordbrukets första femtusen år. Natur och Kultur/LTs förlag, Lagersberg, 11 236. Welinder, S. 1998b. Pre-Neolithic Farming in the Scandinavian Peninsula. Teoksessa M. Zvelebil, L. Domańska and R. Denenell (eds.), Harvesting the Sea, Farming the Forest. The Emergence of Neolithic Societies in the Baltic Region. (Sheffield Archaeological Monographs, 10), 164 174. 34

Werbart, B. 1998. Subneolithic: What it is? Subneolithic societies and the censervative economies of the Circum-Baltic Region. Teoksessa M. Zvelebil, L. Domańska & R. Denenell (eds.), Harvesting the Sea, Farming the Forest. The Emergence of Neolithic Societies in the Baltic Region. (Sheffield Archaeological Monographs, 10), 37 44. Wilkins, H. 2009. Transitional change in proto-buildings: a quantitative study of thermal behaviour and its relationship with social functionality. Journal of Archaeological Science 36, 150 156. Zernitskaya, V. & Mikhailov, N. 2008. Evidence of early farming in the Holocene pollen spectra of Belarus. Quanternary International 2008, doi.10.1016/j.quaint.2008.04.01, 1 14. Zhulnikov, A. 2008. Exchange of amber in northern Europe in the III millennium as a factor of social interactions. Estonia Journal of Archaeology, Vol. 12, Issue 1, 3 15. Zvelebil, M. 1998. Agricultural frontiers, Neolithic origins, and the transition to farming in the Baltic Basin. Teoksessa Zvelebil, M., Domańska, L. & Dennell, R. (eds.), Harvesting the sea, farming the forest. The emergence of Neolithic societies in the Baltic region. (Sheffield archaeological monographs 10). Sheffield Academic Press, Sheffield, 9 27. Zvelebil, M. 2006. Mobility, contact, and exchange in the Baltic Sea basin 6000 2000 BC. Journal of Anthropological Archaeology 25, 178 192. Zvelebil, M. & Dolukhanov, P. 1991. The Transition to Farming in Eastern and Northern Europe. World Prehistory, vol. 51, No. 3, 233 278. Zvelebil, M. & Lillie, M. 2000. Transition to agriculture in eastern Europe. Teoksessa T. Douglas Price (ed.), Europe s First Farmers. Cambridge University Press, Cambridge, 57 92. Zvelebil, M. & Rowley-Conwy, P. 1984. Transition to farming in Northern Europe: a hunter-gatherer perspective. Norwegian Archaeological Review, 17, 104 128. Äyräpää, A. 1939. Suomen kivikauden kulttuurimuodot. Suomalaisen tiedeakatemian Esitelmät ja Pöytäkirjat 1937. Suomalainen tiedeakatemia, Helsinki, 101 126. 35

SUMMARY Neolithic cereal cultivation in Finland This paper outlines the oldest cereal cultivation in Finland and in the neighbouring areas, and presents an hypothesis concerning the probable age of cereal cultivation in Finland. New pollen analytical results from the Baltic countries and northwestern Russia indicate that cereal cultivation was practiced there already during the 4 th millennium cal BC. These results are some one to two thousand years older than the oldest Finnish cereal pollen finds currently accepted as indicators of cultivation. The presence of cereal cultivation in Finland during the Typical Comb Ware Period is suggested in the paper. This view is argued on the basis of pollen analytical results as well as changes in the settlement pattern. At present, the oldest commonly accepted cereal cultivation in Finland dates to the Late Stone Age, ca. 2500 1800 cal BC (see Vuorela & Hicks 1996; Vuorela 1998; 1999a-b). However, there are a number of older cereal pollen finds from the Finnish interior and from Northern Finland (Fig.1 and Table 1) that have been generally rejected by scholars. In most cases, the rejected results have been contradictory to the prevailing state of research, or palynologists have hesitated to interpret certain early cereal pollens as positive indicators of cultivation. No new dates that would have moved the beginning of cereal cultivation in Finland further back in time have been forthcoming since the 1990 s. Over the last decade, numerous new dates on cereal pollens have been published in Estonia (see Kriiska 2003; Poska et al. 2004; Poska & Saarse 2006). These results show that cereal cultivation was practiced by (sub)neolithic cultures in Estonia at least from ca. 4000 cal BC onwards (Kriiska 2003). Nearly equally old results have been obtained from the eastern shore of the Lake Onega, Russia (Vuorela et al. 2001). It has previously been thought that agriculture was introduced into the Baltic countries by the Corded Ware Culture from 3200 cal BC onwards. The new dates now show that cereal cultivation was practiced in Estonia notably earlier (Kriiska 2003). The new results challenge the conventional models concerned with the spread of agriculture into the Baltic countries and Finland. In these models, agriculture spreads through the interaction between foragers and farmers (e.g. Zvelebil & Rowley-Conwy 1984; Zvelebil 1998; Lang 1999; Zvelebil & Lillie 2000). The models were constructed at a time when the spread of agriculture and the Corded Ware culture were thought to be fairly contemporaneous events. Now it appears that cultivation was known among (sub)neolithic cultures long before direct contacts between local and Corded Ware cultures were established. In Finnish archaeology, the beginning of cereal cultivation has been a difficult subject to deal with. Neolithic phases with agriculture are nearly absent, and even the Corded Ware Culture is not accorded agrarian status without objection. Actually, Finnish archaeologists are of two opinions concerning the practice of agriculture by the Corded Ware population. The sceptics reject the idea of agriculture because concrete evidence, such as cereal grains, grain impressions in pottery, and cereal pollens, is lacking. Others seek to ascertain the presence of agriculture through indirect evidences, such as site placement. 36 Suomen Museo 2009

With this background, it is not surprising that among Finnish archaeologists the Comb Ware cultures (ca. 5100 3200 cal BC) have been treated as pure hunter-fisher-gatherers. The Corded Ware Culture has had a deep impact on the way Finnish archaeologists think about early agriculture. For example, notions proposed by Aarne Äyräpää (1939) and later reiterated by Torsten Edgren (1970), according to which the Corded Ware Culture settled the area most suitable for agriculture and actively stayed away from the coniferous forest marked by the presence of spruce (Picea) spreading from east, have created a situation where the recognition of Stone Age cereal cultivation outside the temporal and areal distribution of the Corded Ware Culture has been complicated. There are good reasons for demolishing this myth. First, according to current data, the spread of spruce and the area of the Finnish Corded Ware Culture overlapped only towards the end of the Corded Ware Phase and in few areas. Second, cereal cultivation coeval with the Corded Ware Culture is known from the coniferous forest zone beyond the northern limit of the distribution of Corded Ware (see Fig 1 and Table 1). Third, during the Holocene thermal maximum (c. 6000 2500 cal BC) the deciduous forests suitable for early cereal cultivation reached much further north than Corded Ware (Fig. 2). Although no undisputed cereal pollen dating older than the Corded Ware Culture (3200 2350 cal BC) is known from Finland, there are other positive indicators of possible cultivation during the Comb Ware Period. There are signs of forest clearing with a concomitant increase in charcoal particles, increases in plants usually connected with the presence of meadows/fields, and an increase in plants indicating the presence of man, but no signs of cereals. This is not surprising, since early cultivation has most probably been practiced in small patches cleared in the forest, and the earliest cultivated cereals are selfpollinating crops with a low yield of pollen grains and poor pollen dispersal. Therefore, the absence of cereal pollens does not exclude the possibility of cultivation (see Poska et al. 1999). During the Finnish Typical Comb Ware Period (4000 3400 cal BC), cultivation was practiced within the same cultural entity in Estonia and by people using Rhomb-Pit Ware by Lake Onega in Russia. Archaeological data shows that at this time, vast interaction networks connecting the hunter-gatherer groups of the eastern boreal forest zone were actively spreading exotic goods such as Baltic amber, Russian flint, copper, and green slate over vast areas (e.g. Edgren 1992; Carpelan 1999; Zvelebil & Lillie 2000; Zvelebil 2006; Zhulnikov 2008). Therefore, there is good reason to suppose that a knowledge of cereal cultivation could also have spread via this network much like any other goods. During the Typical Comb Ware Period a number of changes in the settlement pattern took place: pithouses began to occur in village-like concentrations and pithouse sites shifted from sheltered loci to unsheltered ones. Such places were often islands or windy capes. This change is seen in the Lake Saimaa area in the Finnish interior (Mökkönen 2002) as well as on the Karelian Isthmus in Russia (Mökkönen 2009). According to ethnographic sources, pithouses have been used primarily as winter dwellings (Gilman 1987). However, during the Typical Comb Ware Period pithouses began to be constructed in locations that were more suitable for summer than for winter habitation, and at the same time more suitable for long term habitation and storage. I have interpreted 37

this change as a mark of growing sedentism (Mökkönen 2009). Another phenomenon, the increase in housepit size, takes place in northern Europe approximately at the same time, i.e., mainly after 3500 cal BC (Norberg 2008). The growing size of the dwellings also increases their suitability for storage and indoor activities. As an hypothesis, I would propose that cereal cultivation was already present in this development. To sum up: The earliest cereal cultivation most probably arrived in Finland from the east already before the beginning of the Corded Ware Period. Judging from (1) the age of the earliest cereal cultivation in Estonia and NW Russia, (2) the changes in housing and in the way the sites were located, and (3) the dense network that during the 4 th millennium cal BC was spreading goods and transmitting ideas between cultivators and non-cultivators, the initial stage of cereal cultivation most likely arose during the Typical Comb Ware Period. I do not consider an even earlier start as totally unthinkable. In the light of current research, the early dates for the first occurrence of cereal pollens (Reynaud & Hjelmroos 1980; Vuorela & Kankainen 1991) are precisely what one would expect. Previously, these early results have either been excluded from discussion or their value as an indicator of cereal cultivation has been doubted. In order to clarify the date of the beginning of cereal cultivaltion in Finland, new archaeological, palaeobotanical, and palynological studies are needed. 38