The Spatial Organization of Bone Crafting During the Middle and Late Mesolithic at Ringsjöholm and Strandvägen in Sweden

Journal of Field Archaeology

ISSN: 0093-4690 (Print) 2042-4582 (Online) Journal homepage: https://www.tandfonline.com/loi/yjfa20

Sara Gummesson, Fredrik Molin & Arne Sjöström

To cite this article: Sara Gummesson, Fredrik Molin & Arne Sjöström (2019) The Spatial Organization of Bone Crafting During the Middle and Late Mesolithic at Ringsjöholm and Strandvägen in Sweden, Journal of Field Archaeology, 44:3, 165-179, DOI: 10.1080/00934690.2019.1580093 To link to this article: https://doi.org/10.1080/00934690.2019.1580093

View supplementary material

Published online: 01 Mar 2019.

Submit your article to this journal

Article views: 252

View Crossmark data

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=yjfa20 JOURNAL OF FIELD ARCHAEOLOGY 2019, VOL. 44, NO. 3, 165–179 https://doi.org/10.1080/00934690.2019.1580093

The Spatial Organization of Bone Crafting During the Middle and Late Mesolithic at Ringsjöholm and Strandvägen in Sweden Sara Gummessona, Fredrik Molinb, and Arne Sjöströmc aUppsala University, Uppsala, Sweden; bNational Historical Museums, Linköping, Sweden; cLund University, Lund, Sweden

ABSTRACT KEYWORDS This paper focuses on the spatial distribution of bone tool production waste from two Mesolithic sites bone tool production; spatial in Sweden, Ringsjöholm and Strandvägen, with well-preserved faunal remains including bone and distribution; Mesolithic antler artifacts. Local production on both sites has generated a variety of identifiable waste period; taphonomy products deriving from complete chains of production, including unmodified bones, debitage and finished products. Identified categories include: blanks, removed epiphyses, bone flakes, and preforms. Identification of species shows that antler and bone from red deer were the preferred raw materials. Spatial statistical analyses confirm that different stages of bone tool production were organized within separate areas of the sites and that larger items were discarded in the water along the shorelines. Interestingly, blanks and preforms seem to have been stored under water for future use and demarcated clusters of bone flakes in association with dwellings represent “bone knapping floors” where production was more intense than in other areas.

Introduction been reported in Sweden, such as at Norje Sunnansund (Kjäll- One of the more important aspects of animal carcass utiliz- quistetal.2014), Rönneholm (Larsson and Sjöström 2011), ation in the Mesolithic was the use of skeletal elements as Tågerup (Cronberg 2001), and Lassebacken I and II (Persson raw material for bone tool production. Bone and antler 2012). These studies have enhanced our understanding of were obviously important raw material for various kinds of the spatial settlement organization and of human activities tools, hunting implements, or personal ornaments. Such on site. Often these patterns indicate the presence of specific finds appear occasionally on Scandinavian Mesolithic sites, activity areas at these sites (Stapert 1989). Our knowledge of only when the preservation of organic material is favorable. activities related to bone tool production, however, is limited. In recent years information about these finds has increased Poor conditions for preservation of osseous material have (David and Kjällquist 2018; Gummesson and Molin in often limited the spatial studies only to lithics; or, at best, press), but still our knowledge of the spatial organization of when faunal assemblages are analyzed only species distri- this craft is limited. Spatial aspects of craft production can butions are used (Olson et al. 2011; Kjällquist et al. 2014). provide further insights as they may highlight important An important exception is the bone tool fabrication floor at aspects of craft organization and specialization (Costin 1991). Kasteelberg B, South Africa (Smith and Poggenpoel 1988). Production is often modeled through technological organ- Spatial patterns in osteoarchaeology also have an obvious ization (Nelson 1991)orchaîne opératoire (e.g., Geneste link to taphonomy, where the spatial dimension of deposi- [1991]). Such perspectives may vary in their overarching tional patterns is of interest (Gummesson, Molin, et al. theoretical positions (Perlès 1993), but they both focus on 2017). Many variables, such as the level of fragmentation or and emphasize modeling of the technological process as burning, affect the possibility of identifying different traces different sequences, including steps such as raw material and marks on the bones. Also, post-depositional processes acquisition, processing, use, and discard. Furthermore, var- affect the osseous material through weathering and degra- ious alternatives that might have influenced or determined dation (Gummesson, Molin, et al. 2017). All of this affects the practices are considered. Conneller (2006) has illustrated comparisons of observed quantitative patterns. the chaîne opératoire in relation to both time and the land- This paper presents the results of spatial studies of bone scape, where places and agencies are linked in spatial net- tool production waste from two Mesolithic sites in Sweden, works that actually produce the landscape (also see Ingold both with well-preserved osteological remains. The two [1993] on taskscapes). sites, Ringsjöholm and Strandvägen, are dated to the Middle Spatial dimensions in prehistoric settings may be studied in and Late Mesolithic, with Ringsjöholm in Scania representing terms of many different aspects and scales. Grøn (2003)has an older chronological sequence than Strandvägen in Öster- shown the importance of the spatial organization of dwell- götland. Favorable preservation conditions at both sites ing-spaces among recent hunter-gatherers and how it seems have led to the recovery of bone artifacts otherwise often rooted in the social psychology of smaller groups. Spatial dis- not represented in the archaeological record. Earlier studies tribution patterns for lithics on Stone Age sites have commonly on Mesolithic bone tool production, technology and

CONTACT Sara Gummesson sara.gummesson@ofl.su.se Lilla Frescativägen 7, 114 18 Stockholm, Sweden Supplemental data for this article can be accessed at https://doi.org/10.1080/00934690.2019.1580093 © 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way. 166 S. GUMMESSON ET AL. operational chains, principally by Eva David (1999, 2007; Strandvägen David et al. 2015), have been used as a point of departure During the last decade, an important Mesolithic site com- for the present study. At both sites bone tool production plex in Motala, in Eastern Central Sweden, has seen large- has generated identifiable blanks, preforms, discarded waste, scale archaeological excavations (FIGURE 1). This work has and debitage. The finds actually constitute a large part of revealed three more or less contemporaneous sites from the two faunal assemblages; they represent chains of pro- the Late Mesolithic in close spatial relation to each other duction—from unmodified bones to debitage, finished, and and to the River Motala Ström. The river connects Lake Vät- reworked objects—from the local production of bone tools. ternwiththeBalticSeabasin,whichatthattimewassitu- The excavated areas of the two sites, although markedly ated approximately 30 km to the east. On the north side of different in size, both include dry occupation layers and sub- the river, two sites have been excavated: Kanaljorden, a merged sediments in water. Both sites exhibit settlement deposition of human and animal remains in a small lake remains with dwellings and different archaeological features. (Hallgren and Fornander 2016); and Verkstadsvägen, a The archaeological remains at Ringsjöholm provide an site on the river bank with several dwellings (Hagberg and important example of the use of a rather limited space in Westermark 2015). On the opposite river bank, on the the landscape, on a longshore bar in Lake Ringsjön. Strandvä- south side of the river, is the site Strandvägen. This settle- gen on the other hand is a large settlement with several con- ment was excavated during the years 1999–2003 and temporary occupational units on the shore of the River 2010–2013, with the site comprising extensive refuse layers Motala Ström. It was anticipated that differences in pro- along the river shore, several dwellings, knapping floors, and duction and/or settlement organization may be visible in activity areas, as well as a Mesolithic cemetery (Carlsson the spatial distribution patterns of finds that are related to 2008;Molinetal.2014; Gummesson and Molin 2016). bone craft but also that these patterns would provide Thesiteisdatedto6000–4500 CAL B.C. with the most inten- additional previously unknown information on site use and sive period of settlement at around 5500–5000 CAL B.C.An craft organization. area of 7110 m2 has been investigated, of which 3145 m2 of occupational and refuse layers have been water-sieved (using a 4 mm mesh). The excavated area thus includes Sites and Material dry occupation layers and submerged stratigraphy extend- Ringsjöholm ing into the river. Waterlogged sediments were preliminary investigated by marine archaeologists during 2003 (Carlsson Ringsjöholm is located on the western shore of Lake Ringsjön 2008), but were excavated in much larger scale from land in Scania, approximately 35 km from the coast (FIGURE 1). during the most recent excavations by building dams and The site is situated on a small longshore bar, about 250 × using electric pumps to keep water out of trenches. Six 10–20 m wide. The occupational area is estimated to cover dwellings have been documented in the settlement and approximately 5000 m2 but only 266 m2 has been excavated. each dwelling is connected to a corresponding activity area The archaeological work was carried out as research and wherelithicdebitageandfaunalremainsweredeposited, seminar excavations during the years 1994–1997 (Sjöström probably the result of accumulation over time (Molin et al. 1997). The excavations covered both dry occupational layers 2018). These accumulations have been considered to be pri- on land and previously submerged sediments (fen peat and mary deposits of refuse consisting of a diversified assem- drift gyttja). In the settlement area, one dwelling and several blage of finds. In addition, four larger craft areas have smaller archaeological features, such as two hearths, were dis- been identified on the basis of the spatial distribution pat- covered. Unfortunately, parts of the site have in recent times terns of lithics. Eight different zones with accumulations been affected by post-depositional activities, such as tren- of bone finds have been identified (Gummesson, Molin, ching, which could have disturbed the archaeological remains et al. 2017). Many of the finds associated with bone craft and affected the preservation of the organic remains. The site have been recovered in submerged sediments along the is dated to ca. 7100–6000 CAL B.C., with a possibly more inten- shoreline but a rather large amount of material, although sive (or better preserved) phase of activities 6400–6000 CAL more heavily fragmented was also recovered in the dry occu- B.C. (Sjöström 1997). The dry occupation layer shows no pational layers on the settlement (Gummesson and Molin in clear stratigraphic separation, while the submerged sediments press). The osteological analysis, together with archaeobota- exhibit several substrata, of which some layers have an older nical analyses, suggests a more or less continuous human formation date. Artifacts from the previously submerged and presence on the site throughout the year (Gummesson, stratigraphically lowest layer, L3, can be dated to ca. 7100– Karlsson, et al. 2017). 6200 CAL B.C. Dispersed finds of this period were also recovered in other excavated areas of the site. The osteological material from the 1994–1997 field seasons Methods has previously been analyzed by students at the Department of Osteology and manufacturing Historical Osteology at Lund University (see Magnell [2006: 119] and references therein). The results indicate a rather The faunal assemblages from both sites were analyzed in equal representation between different large ungulate species regard to the production of bone tools (including artifacts and a high frequency of fur animals compared with other made of antler), focusing on finished artifacts, blanks, pre- Mesolithic materials from Scania. The minimal season of forms, and debitage. Detailed taphonomic analyses were car- occupation for the complete assemblage was estimated to be ried out on all material from Ringsjöholm and on material from July–February. However, in his analysis of wild boar from Strandvägen excavated during the years 2010–2013. hunting Magnell (2006: 78, fig. 29) reports that the shortest All finished artifacts, blanks, preforms, and debitage were possible hunting period at the site was from July to October. recorded and quantified according to weight and to number JOURNAL OF FIELD ARCHAEOLOGY 167

Figure 1. Ringsjöholm and Strandvägen marked on a map of Sweden, showing present day shorelines. The River Motala Ström is marked by a blue line connecting Lake Vättern with the Baltic Sea. of identified specimens (NISP), minimum number of Distribution patterns and statistical analysis elements (MNE), and minimum number of individuals The aim of this study was to test observed spatial patterns of (MNI) (Lyman 2012). Bone fragments were identified to bone craft using GIS and statistical analysis. At Ringsjöholm species level if possible (or to higher taxa if not), to bone the exact positions of all finds were individually documented element and side, and to part of element. Fragment size and using a local coordinate grid. In this study the coordinates has level of firing, following Stiner et al. (1995), was recorded for been denoted to the nearest excavation square (1 × 1 m) to all fragments and a fracture analysis was carried out on avoid stacking and to allow summation of the total number unburned long bone fragments, according to methods pre- of finds for each square meter. At Strandvägen all finds sented by Outram (2001) and Johnson (1985). Artifacts were documented and related to stratigraphic layer and exca- were also assessed and registered to part of the object and vation unit (1 × 1 m), using the software Intrasis 3.0. to the general degree of completeness (i.e., fragmented, half The spatial analysis and layout was created using ArcGis of a complete object, or complete artifacts). The orientation 10.2.1. Statistical analyses of find distributions were per- of artifacts (i.e., working edges) follows the definitions of formed, using different methods, based on NISP values. Voruz (1984: 287, fig. 7). Spatial autocorrelation Moran’s I (Mitchell 2005; Dormann The analyses focused on the identification of production et al. 2007) (for archaeological application, see Lindholm waste such as flakes and blanks, but also debitage in the et al. [2013]) was applied in order to evaluate whether the form of cut-off metapodial epiphyses and bone fragments data expressed clustering or a dispersed or random distri- with visible surfaces modifications. The finds were assessed bution. In addition, Ripley’s K function, previously success- and categorized in regard to different stages of bone tool pro- fully applied for archaeological point data (Winter-Livneh duction. The categories identified were blanks (anatomical et al. 2010; Tacher et al. 2017) was used to test whether the and technical), removed epiphyses, preforms, flakes, and spatial distributions were significantly clustered or dispersed. finished artifacts. The registration criteria are presented in A total of 99 permutations were run and the “simulate outer detail in Supplemental Material 1. The analysis was carried boundary values” edge correction was used. Weight values out at the Osteoarchaeological Research Laboratory, Stock- (i.e., the NISP values) were both omitted and specified in holm University, Sweden. 168 S. GUMMESSON ET AL. order to evaluate clustering of feature values and clustering/ assemblage at Strandvägen (ca. 45% of the identified artifacts) dispersion of feature location. Given that NISP quantification while slotted points are more frequent at Ringsjöholm (24%, was used, feature location (i.e., omitted values) was regarded compared to 4% at Strandvägen). There is a difference in the as more accurately representing potential clustering. Signifi- frequency of plain points between the sites, but this may poss- cantly clustered distributions were finally mapped in a cluster ibly be an effect of fragmentation. Ringsjöholm exhibits a and outlier analysis, using Anselin Local Moran’s I (Anselin higher number of fragments from unspecified bone points 1995; Mitchell 2005), with applied False Discovery Rate that in fact may originate from plain points, arrowheads, or (FDR) correction. spearheads. Unidentified bone points and plain bone points at Ringsjöholm represent about 16% of the identified artifacts, while unidentified bone points, spearheads, arrowheads, and Results plain bone points represent about 20% at Strandvägen. The larger assemblage at Strandvägen exhibits a greater Products, artifacts, and production waste variety of artifact types, including slotted daggers, chisels, kni- fi fi ves, navettes (or net mending tools [David et al. 2015: fig. At Ringsjöholm, 316 nds of nished bone and antler arti- fl facts have been identified and 977 fragments were registered 12.2]), and antler objects such as sleeves, pressure akers fi (Pelegrin 2012; David and Sørensen 2016), and clubs, as production waste. At Strandvägen, nished artifacts ff amount to 1468 finds and production waste to 2185 frag- which are to be expected given the di erence in numbers. ments. The identified finished artifacts at Ringsjöholm rep- There are, however, some objects that only occur at Ringsjö- resent 2% (1.627 kg) and unfinished products or debitage holm, or are more common there: for example, a complete bone adze made of the metapodium of an aurochs (Bos primi- 9% (7.199 kg) of the total weight of the osteological assem- — blage (estimated to a total of 80 kg [Sjöström 1997: 12]). At genius). Neither the artifact type which has been considered fi a type artifact of the Maglemose Culture (Larsson 1978; Strandvägen, the nished artifacts represent 18% (9.9 kg) of — fi the total weight (56.3 kg) and the production waste represents David 2003) nor aurochs bone have been securely identi ed 17% (9.3 kg) of the faunal remains (corresponding to 4% and at Strandvägen. Personal ornaments like beads, pendants, and 6%, respectively, of the total number of fragments). Ringsjö- pearls are also much more common at Ringsjöholm, repre- holm exhibits a higher frequency of finished artifacts per senting around 18% of the assemblage, while at Strandvägen 2 only about 1% can be identified as ornaments of this type. square meter, at 1.19/m , than Strandvägen, where the fre- ff fi quency was 0.47/m2. There is also a slight di erence in the number of sh hooks The two assemblages show some obvious differences in the and gorges from the sites, but due to fragmentation and the fragility of the artifact type, this is hard to evaluate. At Ring- inventories of artifacts (TABLE 1). Barbed points dominate the sjöholm, one fish hook (made of enamel from a boar tusk) was identified. At Strandvägen, on the other hand, three Table 1. Number of bone and antler artifacts from Ringsjöholm and gorges and only one possible broken shaft of a fish hook Strandvägen. In parentheses: total number of finds identified as representing fi more than 50% of the complete artifact. Classification according to: Montelius have been recovered, although the latter identi cation is not (1994 [1917]), Althin (1954), Larsson (1978, 2005), Stewart (1996), Vankina secure. (1999), Karsten and Knarrström (2003), Browall (2011), Pelegrin (2012), The amount of production waste is rather similar on the Galiński (2013), and David et al. (2015). two sites (TABLE 2). Flakes represent 46% of the production Bone and Antler Artifacts Ringsjöholm Strandvägen waste at Ringsjöholm and 39% at Strandvägen. At Ringsjö- Barbed point 6 446 (121) fl Harpoon head 1 14 (1) holm akes have also, to a higher degree than at Strandvägen, Slotted bone point 34 (1) 43 (5) been identified to a specific type. Plain bone point 5 180 (27) Arrowhead 8 (5) Spearhead 16 (8) Species and element distribution Bone point, unspec. 17 (1) Needle/Pin 15 (1) 35 (2) The species distribution of skeletal elements utilized in tool Slotted dagger 8 (4) ff Ulna dagger 1 production in the two assemblages are somewhat di erent Awl 7 (6) 71 (47) (TABLE 3). Red deer (Cervus elaphus) is the dominating Chisel 53 (43) species in both assemblages—quantified either by NISP, Knife 5 (5) — fi Wedge 1 (1) MNE, or MNI regarding nished artifacts and production Navette 7 (5) waste. Roe deer (Capreolus capreolus) and moose (Alces Fish hook 1 (1) 1 alces) are also commonly represented. Auroxen (Bos primi- Gorge 3 (2) Pressure flaker/Puncher 8 (7) 33 (18) genius), wild cats (Felis silvestris), and pine martens (Martes Antler axe/Adze/Mattock 3 (2) 10 (8) martes) have been identified at Ringsjöholm, but not at Bone adze (metapodium) 1 Strandvägen, while beavers (Castor fiber) were only identified Antler club 6 (3) Club (bone) 1 (?) at Strandvägen. The latter species is, however, represented by Antler shaft 4 (4) rather few finds. Species representation is more varied among Antler sleeve 3 (3) the finished products at both sites, exemplified by a worked Decorated antler object 6 (2) Stopper/Cork 3 (3) beaver incisor, bird beads, and carnivore pendants. Knife/scraper, boar tusk 2 (1) 7 (3) The most frequently utilized raw material at both sites is Enamel tool 1 13 (7) antler (SUPPLEMENTAL MATERIAL 2). However, a higher frag- Beads/Pearls/Pendant 26 (19) 13 (7) Unident. antler object 14 31 (9) mentation level for antler than for bones and some level of Unident. 174 (4) 447 (7) bias in identification need to be considered, as these may lead to inflation in NISP quantification. This has been studied Total 316 (42) 1468 (334) at Clos de Poujol, France, in regards to the production of JOURNAL OF FIELD ARCHAEOLOGY 169

Table 2. Categories of production waste (NISP). In brackets; total numbers of Strandvägen it is the metatarsal. Wild boar (Sus scrofa)is fi fi nds identi ed as representing more than 50% of the complete product. mostly represented by canines, whereas small carnivores are Production Waste Ringsjöholm Strandvägen represented by several teeth, and birds (Aves) by long bones. Anatomical blank 201 (8) 255 (1) Technical blank 82 (26) 172 (13) Preform 20 (4) 101 (12) Flake, unident. 151 (150) 594 (464) Level of firing and fragmentation Flake, divergent 148 (146) 168 (159) Flake, squared 147 (147) 85 (80) There is a difference in the level of firing of finished artifacts Calibration flake 1 (1) 3 (2) Epiphyses (removed) 66 (50) 64 (34) and production waste in the two assemblages (TABLE 4). Debitage, unident. 161 (6) 744 (9) Burning of bone fragments and fragmentation might render Total 977 (538) 2185 (489) objects less likely to be identified, and, thus, affect comparisons between different assemblages. At Ringsjöholm, 56% fi antler tools and the exploitation of red deer (Bridault et al. of the nished artifact fragments exhibit some degree of fi – 2009). Most common at both sites are antler of red deer, fol- ring (score 1 6). Most common are artifacts with a high fi lowed by roe deer, whereas moose antler was only identified degree of ring (score 5 and 6), comprising 78% of the frag- ff fi at Strandvägen. Tines are the most common part of the red ments a ected by re. At Strandvägen, only 14% of the arti- deer antler at both Ringsjöholm and Strandvägen, rep- fact fragments are burned and most that are burned have a fi resented by 12 and 45 finds respectively. The burr is some- low degree of ring. However, if we restrict our focus to one of the dwelling floors at Strandvägen (dwelling 6), what more common at Strandvägen, represented by 13 2 pieces at that site and only by three finds at Ringsjöholm. measuring about 26 m , the number of burned artifact frag- At Ringsjöholm, two of the three burrs are from shed antlers, ments actually increases to roughly 36%, a frequency more ff whereas eight of the 13 finds at Strandvägen were still comparable to that of Ringsjöholm. The di erences in degree fi attached to crania. Antlers of roe deer are more equally rep- of ring are not as pronounced among the production waste: resented regarding burr and tines. At Ringsjöholm, seven 12% are burned at Ringsjöholm, and 7% at Strandvägen. It is, fi complete or almost complete antlers with burr, beam, and however, worth noting that the highest degrees of ring tines were recovered; at Strandvägen, eight antlers with burr (scores 4, 5, and 6) are more common also among the pro- were recovered, as well as the same number of antler tines. duction waste at Ringsjöholm (58%) than at Strandvägen The beam was represented by a lower number of finds. All (24%). roe deer antlers at Ringsjöholm were still attached to the cra- The level of fragmentation in the studied assemblages is ff nium whereas five of eight were shed at Strandvägen. similar (TABLE 5). There is a marked di erence in the esti- fi The most commonly utilized bone elements of large ungu- mated level of completeness between nished artifacts and lates are those of the metapodium. Interestingly, at Ringsjö- production waste at Ringsjöholm (TABLES 1 and 2). Only fi holm the metacarpal is the more common element but at 13% of the nished artifact fragments represent half or more of a complete product. In contrast, as much as 55% of the production waste (excluding complete distal epiphysis Table 3. Species distribution of bone and antler artifacts and production waste and broken blanks) can be estimated to represent more at Ringsjöholm and Strandvägen. than half of an originally whole product. At Strandvägen Ringsjöholm Strandvägen such a difference cannot be identified, instead about 20% of NISP MNE MNI NISP MNE MNI the finished artifacts and the waste products can be estimated Artifacts to represent more than half of a complete product. Alces alces 22 1 3591 Bos primigenius 11 1 Small unburned fragments are more common than large Capreolus capreolus 12 4 1 30 11 2 unburned fragments on both sites. Finds larger than 10 cm Cervus elaphus 26 10 2 281 28 6 only account for roughly 7% of all finds. The most common Felis silvestris 11 1 – Martes martes 33 1 fragment size is 1 4 cm, representing some 70% of both Sus scrofa 11 3 2 24 6 1 assemblages. Fragmentation, however, differs for different Ursus arctos 22 1 111 Castor fiber 11 1 Cervidae 9 13 Table 4. Level of firing (Stiner et al. 1995) among artifacts and production waste Small carnivore 4 at Ringsjöholm and Strandvägen (NISP). Carnivora 2 – Large Ungulate 6 381 Ringsjöholm Strandvägen (2010 2013) Ungulate 3 Artifacts Aves 11 1 1 0 138 1017 Unident. 234 691 11584 (incl. small, medium and large 2419 mammals) 328 Total 316 26 9 1468 57 13 42517 53011 Production waste 6 102 20 Alces alces 19 7 2 17 8 2 Total 316 1176 Capreolus capreolus 72 23 5 57 14 1 Production waste Cervus elaphus 379 20 5 1083 39 5 0 861 1675 Sus scrofa 74 3 115112760 Castor fiber 21 122024 Cervidae 30 97 317 Large Ungulate 26 82 4496 Unident. (incl. small, medium and 444 836 581 large mammals) 61122 Total 977 54 15 2185 67 7 Total 977 1795 170 S. GUMMESSON ET AL.

Table 5. Fragmentation level of unburned artifacts. Fragments exhibiting low levels of ﬁring (0, 1, and 2), i.e., fragments that were only charred, are also included. Ringsjöholm (NISP) Strandvägen (NISP) Fragment size Dry stratigraphy Submerged stratigraphy Total Dry stratigraphy Submerged stratigraphy Total 0–1 cm 14 1 15 39 6 45 1–2 cm 42 4 46 258 34 292 2–4 cm 39 10 49 139 112 251 4–6 cm 5 4 9 16 121 137 6–10 cm 5 7 12 12 235 247 > 10 cm 6 11 17 3 149 152 Total 111 37 148 467 657 1124

categories of finds and at different areas of the sites (TABLE 5 more common at the proximal end (FIGURE 4). At Ringsjö- and 6). Generally, fragmentation is higher in dry occupation holm the fracture distribution is more even. layers on land than in the submerged sediments. It may be noted that the estimated wear of artifacts is similar In the dry occupation layers at Ringsjöholm, 18% of the at both sites. Wear was only estimated macroscopically and no finished artifacts are smaller than 2 cm and 65% of the use-wear analysis by microscope has yet been carried out. At waste products are (in maximum dimension) smaller than Ringsjöholm, 94% of the artifacts exhibit some to moderate 2 cm. At Strandvägen, about 62% of both finished artifacts wear. At Strandvägen, 85% exhibit the same. Major wear was and the waste products in the dry layers are smaller than registered on only 5% and 9% of the artifacts. At Ringsjöholm 2 cm in maximum dimension. At Ringsjöholm, 9% of the artifacts without wear represent 1% of the collection, corre- artifact fragments from dry sediments are larger than 6 cm, sponding to 3% in the Strandvägen collection. Of the complete and even fewer finds of the production waste (5%). At Strand- items with no or dry breaks at Ringsjöholm, 16 finds exhibit vägen only 3% of the artifact fragments and 2% of the pro- only some wear and could be regarded as still useable. These duction waste in the dry occupation layers are larger than were mainly tooth pendants (n = 9) or pressure flaker/punch- 6cm. ers (n = 4). At Strandvägen, 51 unbroken artifacts exhibited In submerged sediments, 14% of the artifact fragments and only some wear; the majority of these were awls (n = 14), chi- 21% of the production waste are smaller than 2 cm at Ring- sels (n = 9), and bone points (n = 7). sjöholm. At Strandvägen, 6% of the artifact fragments and 37% of the production waste recovered in submerged sediments are smaller than 2 cm. 48% of the artifact fragments Production techniques and 25% of the unfinished products or debitage recovered It is evident that the two assemblages differ in general artifact in submerged sediments at Ringsjöholm are larger than inventory and fragmentation level. This calls for a comment 6 cm, while the same number at Strandvägen corresponds on some identified differences in the mode of tool production to 58% and 17%, respectively. Smaller fragments of finished and the techniques utilized (for further details see Gummes- artifacts are thus more common in both dry and submerged son [2018]). These differences are difficult to quantify and sediments at Ringsjöholm compared to Strandvägen. Larger evaluate as some of the techniques leave more recognizable fragments are generally more common at Strandvägen and debitage than others: percussion, for example, leaves both more common in submerged than dry sediments. negative flake scars on the bone and, detached flakes, as In order to further evaluate the effects of post-depositional opposed to grinding which usually only leaves diagonal stria- processes, an analysis of fracture patterns supplements the tions on the worked bone. study of fragment size. The fracture patterns of production Percussion is evidenced at both sites from impact marks waste are different from those of finished artifacts, in that and negative flake scars on worked bones and from the large fractures produced in fresh bone are more common even amounts of bone flakes. Calibration of the proximal end is ver- on small fragments, especially at Ringsjöholm. Among the ified by a few diagnostic calibration flakes and corresponding finished artifacts, dry breaks (i.e., fractures produced in dry negatives at the proximal end of the metapodia (David 2003, bone) are more common than fresh breaks (i.e., fractures pro- 2007; David and Johansen 1996). Sawing as a technique of duced in fresh bone), especially for the smaller fragments (1– detachment was only identified on two finds from Ringsjö- 4cm)(FIGURES 2, 3). On finished artifacts from Strandvägen, holm: on a removed distal epiphysis from a metapodial of fresh breaks are more common among the larger fragment red deer, and on the palmar portion of a red deer antler. The sizes. In addition, there are different fracture patterns between epiphysis exhibits a sawn groove around the circumference the proximal and distal end of the artifacts, and slightly vary- above the distal end of the bone. Additionally, there are traces ing between different types of objects. Generally, fractures are of axial sawing on the lateral and medial trochleae. On the

Table 6. Fragmentation level of unburned production waste. Fragments exhibiting low levels of ﬁring (0, 1, and 2), i.e., fragments that were only charred, are also included. Ringsjöholm (NISP) Strandvägen, 2010–2013 (NISP) Fragment size Dry stratigraphy Submerged stratigraphy Total Dry stratigraphy Submerged stratigraphy Total 0–1 cm 81 12 93 74 168 242 1–2 cm 244 72 316 376 207 583 2–4 cm 98 154 252 217 331 548 4–6 cm 26 57 83 40 134 174 6–10 cm 13 42 55 12 117 129 > 10 cm 11 58 69 2 56 58 Total 473 395 868 721 1013 1734 JOURNAL OF FIELD ARCHAEOLOGY 171

Figure 2. Relative frequencies of fresh and dry bone break on waste products from A) Ringsjöholm and B) Strandvägen (%NISP). Fresh denotes breaks made in fresh bone, dry denotes breaks made in dry bone, and intermediate denotes fracture that exhibits both fresh and dry bone features (see Outram [2001]).

Figure 3. Relative frequencies of fresh and dry bone breaks on finished artifacts from A) Ringsjöholm and B) Strandvägen (%NISP). Fresh denotes breaks made in fresh bone, dry denotes breaks made in dry bone, and intermediate denotes fracture that exhibits both fresh and dry bone features (see Outram [2001]). antler, one tine had been removed by sawing and there are also the authors do not consider this groove to have been made traces of sawing on the beam of the antler, but detachment had for a prepared break. There are also several smaller fragments not been completed. At Strandvägen, no traces of sawing for at Strandvägen exhibiting traces of axial grooving but in detachment have been identified; instead, nicking (i.e., a per- neither case could the technique be identified as sawing as a cussion technique using a sharp edged tool) was used to divide mean for detachment. antlers and detach epiphyses. Nicking was also identified at Incisions (which may have been created by sawing move- Ringsjöholm. Axial sawing has previously been identified on ments) are more common at Strandvägen, which is associated a metatarsal on Strandvägen (David et al. 2015), however, with a large number of decorated artifacts from the site

Figure 4. Relative frequencies of fracture location on ﬁnished artifacts from Ringsjöholm and Strandvägen (%NISP). Fresh denotes breaks made in fresh bone, dry denote breaks made in dry bone, and intermediate denotes fracture that exhibits both fresh and dry bone features (see Outram [2001]). 172 S. GUMMESSON ET AL.

Figure 5. Spatial distribution of bone and antler artifacts at A) Ringsjöholm and B) Strandvägen. The trench of 1999–2003 excavation at Strandvägen is marked with a gray and black broken line. Finds that represent more than 50% of a complete product are marked with stars.

(Larsson and Molin 2017); in total, 5.5% of the finished arti- Strandvägen, the majority of the finished artifacts were recov- facts are decorated. In contrast only 1.6% is decorated at ered in the submerged sediments along the shore of the river. Ringsjöholm. Traces of grinding also seem to be more fre- The finds show a dense, even distribution in the water, while quent at Strandvägen, where 5.3% of the finished artifacts in dry sediments, they are aggregated in four more or less and the waste products have identifiable traces, whereas spatially limited areas, at dwellings and activity areas. Frag- only 1.2% at Ringsjöholm exhibits such traces. At Strandvä- ments that represent more than 50% of a complete product gen, grinding was commonly observed at the basal end of leis- were predominately recovered in the water. Interestingly, ter points, a tool type represented only by one find at even though bone and antler artifacts were recovered along Ringsjöholm. the complete shore of Strandvägen they are significantly clustered only in the southern part of zone 8 in the water and on land, in zones 1, 2, and 3. Spatial organization of production Technical blanks were recovered from both dry and sub- At both Ringsjöholm and Strandvägen, spatial concentrations merged sediments, but a majority of the finds were recovered of finished bone and antler artifacts and different waste pro- in the water or near the shore at both sites (FIGURE 6). At ducts occur in different areas of the sites, often adjacent to Ringsjöholm, many (n = 26 of 82) technical blanks represent dwellings but also in the water along the shorelines. All distri- more than 50% of the complete product and these are all bution patterns and statistical tests are presented in Sup- found adjacent to the shore. A similar spatial pattern is plemental Material 3A–L. At Ringsjöholm, a much smaller seen at Strandvägen, but the number of finds that represent area was excavated than at Strandvägen, complicating more than 50% is lower (n = 13 of 172). At both sites, Moran’s interpretations of the site organization, but when explored I does not identify the patterns significantly different than in relation to each other, the spatial patterns at the two random, but Ripley’s K analyses (omitted weight value) indi- sites are valuable. cate that the feature location is clustered, and that at Strand- The spatial distributions of finished artifacts are signifi- vägen the feature value also touches on clustering. A mapping cantly clustered at both Ringsjöholm and Strandvägen (SUP- of the distribution (Anselin Local Moran’s I) at Ringsjöholm PLEMENTAL MATERIAL 3AAND3G). At Ringsjöholm, finished does not identify clustering (even if a high low outlier is ident- artifacts (including artifacts representing > 50% of a complete ified). At Strandvägen, on the other hand, a cluster appears in product) are more commonly found on land, in dry occu- a small area in zone 6, in the water (SUPPLEMENTAL MATERIAL pation layers, adjacent to the dwelling (FIGURE 5). At 3B AND 3H). JOURNAL OF FIELD ARCHAEOLOGY 173

Figure 6. Spatial distribution of technical blanks at A) Ringsjöholm and B) Strandvägen. The trench of 1999–2003 excavation at Strandvägen is marked with a gray and black broken line. Finds estimated to represent more than 50% of a complete product are marked with stars.

Preforms are represented by only a few finds at both sites Strandvägen. Single flakes were recovered in the submerged and have more restricted spatial distributions (FIGURE 7). The stratigraphy at both sites. At Ringsjöholm, flakes are signifi- low number of finds at Ringsjöholm hinders the use of spatial cant clustered in only one area, in the western part of the statistics for evaluation of the distribution patterns. In fact, transect. At Strandvägen, significant spatial clustering is one single deposit, already noticed during the excavations, identified in zones 2 and 3. The spatial distributions of differ- constitutes the majority of the finds (FIGURE 8). It consists ent flake types do not show large differences. In areas with of nine split long bones of large ungulates and one finished many flakes, the variation of flake types is higher, which is artifact fragment (TABLE 7), and was recovered in a transect not surprising. One exception is dwelling 4 at Strandvägen, in the former water off the shoreline. A few finds of more where only squared flakes have been identified. The few complete (< 50%) preforms were also recovered in the finds of calibration flakes were recovered at Ringsjöholm water and some fragments were recovered adjacent to the adjacent to the dwelling and at Strandvägen in the southern dwelling and along the shore. At Strandvägen, the distri- part of zone 2 and in zone 8. bution pattern is not recognized to be significantly different Anatomical blanks were often recovered in spatially than random, using Moran’s I, but a clustering of feature confined concentrations at both sites (SUPPLEMENTAL location is suggested (SUPPLEMENTAL MATERIAL 3I). Preforms MATERIAL 3EAND3K). They represent skeletal elements that were recovered along the entire shore except in the northern are commonly used in tool production but do not exhibit part; on land they were found in zone 1 and zone 2. Two finds traces of human modification, mostly in form of antler frag- were also recovered adjacent to the eastern border of the ments or red deer metacarpals and metatarsals (Strandvägen) south-western excavation area. Complete finds (> 50%) of and roe deer metacarpals and metatarsals (Ringsjöholm and preforms were recovered in the water, in zone 8, bordering Strandvägen). At Ringsjöholm the number of finds is too on zone 7, and in dry occupation layers near dwellings in small to allow spatial statistics. At Strandvägen the distri- zone 2. bution is not significantly different than random (Moran’s Flakes exhibit distinctly different spatial distribution pat- I), however feature location may be considered clustered at terns than the other debitage (FIGURE 9). Flakes are mainly smaller distances (Ripley’s K). distributed on land, in dry sediments, and significantly clus- Finally, removed epiphyses have a wide spatial distribution tered at both sites (SUPPLEMENTAL MATERIAL 3D AND 3J), but on Ringsjöholm (SUPPLEMENTAL MATERIAL 3FAND3L). At the patterns are not as obvious at Ringsjöholm as at both sites, the patterns are not significantly different than 174 S. GUMMESSON ET AL.

Figure 7. Spatial distribution of preforms at A) Ringsjöholm and B) Strandvägen. The trench of 1999–2003 excavation at Strandvägen is marked with a gray and black broken line. The position of the identified deposit at Ringsjöholm is marked with a black circle. Finds estimated to represent more than 50% of a complete product are marked with stars. random (Moran’s I) but feature location, again, may be seen traditions and site use (SUPPLEMENTAL MATERIAL 4A–B). as clustered (Ripley’s K). The epiphyses are generally found The utilization patterns differ somewhat between the two close to the shore, at Ringsjöholm they also occur adjacent sites, perhaps related to chronological differences, site func- to the dwelling. At Strandvägen, however, only nine finds tion, or geographical location, or due to the excavated con- are found outside of the shore area. texts. In general, the utilization of bone may be related to subsistence patterns and availability of game animals or technological traditions (Noe-Nygaard 1977). This also entails Discussion consequences for the availability of suitable raw material for The bone tool inventory and spatial patterns at two Meso- bone and antler craft. The species distributions of the utilized lithic sites have provided insights into Mesolithic craft bones differ between Ringsjöholm and Strandvägen, but the

Figure 8. A) Photograph of the deposit at Ringsjöholm during excavation in one square meter unit, and B) a close up. In both pictures north is to the right and the photographer is facing westward. Photo: Arne Sjöström. JOURNAL OF FIELD ARCHAEOLOGY 175

Table 7. The content of the deposit recovered at Ringsjöholm. At Ringsjöholm, most of the finished bone and antler arti- Species Element Side Category NISP Comments facts were recovered on land; at Strandvägen, on the other Alces alces Radius Sin Preform 1 hand, most of the artifacts were recovered in the water Cervus elaphus M.c. Sin Preform 1 along the shore. Interestingly, the distribution patterns of M.t. – Preform 1 Tibia Dxt Preform 1 finished artifacts observed (by the naked eye) at Strandvägen Large Ungulate Radius Dxt Preform 1 Probable differ from the statistically significantly clustered finds of Alces alces finished artifacts. The artifacts cluster in dry sediments, adja- Tibia – Preform 1 M.t. Sin Preform 1 Cervus/Alces cent to dwellings and activity areas, a pattern more similar to M.t./ – Debitage, 1 Technical that observed at Ringsjöholm. This clustering could be inter- Tibia unident. blank? preted in terms of organized spaces at the settlement area as Unident. (incl. small, Ossa Artifact 1 With medium and large plana fragment incisions the results of repeated everyday activities, probably connected mammals) to several smaller household units (Molin et al. 2018). The Total 9 activities in the water, on the other hand, did not result in significant spatial clusters. The distribution patterns clearly need to be considered in relation to the artifact inventory. At differences are represented by only a small number of finds. Strandvägen, larger areas in the water with submerged sedi- fi The raw materials were mainly from ungulates and especially ments have been excavated, constituting shing stations red deer, and the most common utilized elements were antler where barbed points would break and be deposited during fi and metapodial bone. Personal ornaments were made of active shing. Slotted points, on the other hand, which were bones from more uncommon species, but may have been most common at Ringsjöholm, may not mainly be associated introduced to the sites as finished products. These items are with active use in the water, but rather with hunting and other not necessarily representative of the local bone tool industry. activities on land. At Strandvägen, the distribution of slotted Importantly, species distribution seen in the artifact inven- points is also more concentrated on land; 70% of the slotted tory of the two sites is related to the specific utilization of points were recovered from dry sediments corresponding to different elements for different products, such as tooth only 26% of the barbed points (Gummesson and Molin in pendants. press). The low number of personal ornaments at

Figure 9. Spatial distribution of diﬀerent categories of ﬂakes at A) Ringsjöholm and B) Strandvägen. The trench of 1999–2003 excavation at Strandvägen is marked with a gray and black broken line. 176 S. GUMMESSON ET AL.

Strandvägen is noticeable as such large areas of the settle- Osipowicz 2007). This practice is also known from ethno- ment, including burials, have been excavated. graphic records (Semenov 1976 [1964]; Stewart 1996: 84). The spatial distributions need to be considered in relation Soaking is supposed to make bone and antler easier to form to many factors, such as the estimated level of completeness and could also prolong a fresh bone response (Karr and Out- and fragmentation. The fragmentation level of artifacts is ram 2015) as deposition in wet environments affects the plas- higher at Ringsjöholm, but there is also a difference in frag- tic characteristics of bones and thus their responses to mentation between finished artifacts and waste products. mechanical forces. This is known, for example, to have com- The higher number of fresh fractures on the production plicated interpretations of bone fractures on finds from wet waste, mainly among small fragment sizes, must be under- depositions (Kjellström and Hamilton 2014). Technical stood in regard to the identified bone flakes (preserved with blanks at Strandvägen were found spatially clustered in a fresh breaks), but also concerning the fragmentation of small restricted area in the waters off the shore. Similarly, finished artifacts during use. The level of fragmentation also the deposition of preforms at Ringsjöholm occurred just off differs in different locations of the sites, as do the levels of the former shoreline. A similar find is, in fact, known from firing. Generally, fragmented artifacts are more frequently the Mesolithic settlement site of Ageröd V, in Scania, dated burned and exposed to higher temperatures at Ringsjöholm to 5700–5500 CAL B.C., where 14 split long bones were than at Strandvägen; however, in restricted contexts such as found lying horizontally within a 33 × 15 cm area and recov- concentrations associated with dwellings, the patterns at ered with two strips of bark wrapping crossing the length of Strandvägen become more similar to those at Ringsjöholm. the bones (Larsson 1983: 79). Grønnow (1987: 144) has pre- Thus, specific contexts clearly have impacts on taphonomic viously also suggested that parts of the faunal assemblage at patterns. Molin et al. (2014) have suggested that complete the Paleolithic site of Stellmoor in Northern Germany— decorated objects were ritually deposited along the shore of more specifically reindeer antlers recovered from shallow Strandvägen. The observation of larger artifact fragments in waters near the shore—are repeated depositions of blanks. the submerged stratigraphy at Strandvägen could support It is therefore likely that the blanks and preforms in the this interpretation and further explain some of the differences waters of Ringsjöholm and Strandvägen do not represent in the observed patterns. The artifacts from the two sites debitage from attempts gone wrong or bones thrown into apparently have different taphonomic histories and were the water as ordinary refuse, but rather also represent a simi- probably handled differently at the sites during the settlement lar kind of storage and/or preparation of raw material for activities, but were also differently affected by post-deposi- future use. This also attests to storage and planning ahead tional processes. The spatial distribution of finished artifacts (Cunningham 2011; Boethius 2018; Molin et al. 2018) and at Ringsjöholm also differs from the distribution of the a social organization were raw material was brought to majority of the waste products, but not so at Strandvägen. specific places with the anticipation of return and a continu- Probably there were differences in the utilization and pro- ous presence in the landscape. duction of tools at the two sites, as various types of tools Another spatial clustering pattern may be identified on and waste products exhibit different spatial patterns. land, in relation to dwellings and different archaeological fea- It must be stressed that though the perspectives of an oper- tures. Here, secondary reduction debitage, mostly represented ational chain provide a general frame of reference to study by bone flakes, were found to cluster. Flakes are produced technological processes, the specific sequences of reduction when a bone is struck by impact, such as in association techniques (i.e., methods) may vary between different sites with marrow exploitation, in tool production, or perhaps and also within a specific assemblage. Production must be even by accidental impact. According to the description understood as a process related to preceding events, or tech- given by Binford (1978: 54, 153–155) (see also Enloe nological action and choices, but also to expectations on [1993]) of the Nunamiut marrow fracturing, fracturing is pre- future form and/or function (Conneller 2006). It is therefore formed near the articular end of long bones. A single blow possible to find any piece of debitage or even a finished pro- creates a spiral fracture and the articular end is then twisted duct at any stage of production. This is perhaps best exem- off. The impact of the blow creates smaller bone chips but plified by tools that were reshaped due to breakage, the numbers of flakes are relatively low, as is the number of excessive use, or cases where the manufacturing process impact points. This is also consistent with patterns described changed during the course of work, possibly due to handling by Pickering and Egeland (2006), in relation to experimental errors or accident. Along the shores, a combination of differ- patterns from hammer stone percussion. ent depositions were identified, likely of both primary and Different types of flakes occur at Ringsjöholm and Strand- secondary deposits and from different stages of the oper- vägen and they could be subdivided into several categories. ational chain. Mostly primary reduction debitage, such as Divergent flakes and calibration flakes, due to their respective removed distal epiphyses, blanks, and preforms, were recov- morphology and/or location on the skeletal element, are more ered from the waters. The distribution patterns of anatomical easily connected to intentional breakage and shaping of blanks and distal epiphyses do not, however, exhibit any bones. Most squared flakes, however, can only be attributed spatial clustering, and it seems that the distributions of pro- to a general process of fresh bone fracturing. There is good ducts from the initial reduction cannot be statistically separ- reason to believe that most of the larger and elongated ated from a random distribution at the sites. Technical blanks squared flakes derive from controlled and deliberate actions and preforms were most often found in the water on both aiming to shape the bones, rather than from other activities sites, probably in a secondary location in relation to the such as marrow extraction, as many of the flakes exhibit nega- area of production. tive flake scars from previous blows. There are also flakes with It is likely that relatively fresh bones were used for making traces of grooves created to aid the controlled splitting of the tools and it has also been suggested that bone and antler were bones (Clark and Thompson 1953; David 2007). As the deliberately soaked prior to manufacturing (Newcomer 1976; spatial distribution of the different type of flakes is similar JOURNAL OF FIELD ARCHAEOLOGY 177 it seems likely that the delimited areas where flakes cluster household archaeology and intra-site studies, as well as raw materials represent craft areas, or bone knapping floors. However, procurement and technology. though squared flakes mostly coincide with the distribution Arne Sjöström (B.A. 1995, Lund University) is a Ph.D. student at the patterns of other type of flakes, they were also recovered out- Department of Archeology and Ancient History at Lund University. fi fl side these knapping floors, such as at Strandvägen in the floor His interests are mainly in nds of Mesolithic int tools that are found by the lakes in Scania. Through the years he has found hundreds area of dwelling 4. It should therefore be considered that these of Stone Age sites that have been more or less preserved. These sites are flakes may not only be debitage of bone tool production, but not only special regarding the good preservation conditions, but also could also derive from any other activity involving cracking because the abandoned artifacts are found exactly where they were open long bones, such as marrow extraction. At Ringsjöholm, left. This makes the sites very suitable for intra-site spatial analysis. the limited excavated area has not allowed the identification of knapping floors, but there are depositions of flint blades adjacent to the dwelling that could indicate lithic production References in this area. At Strandvägen, the bone knapping floors actu- Althin, C.-A. 1954. The Chronology of the Stone Age Settlement of Scania, ally coincide with identified knapping floors for lithics Sweden. I. The Mesolithic Settlement. Acta Archeologica Lundensia, (Molin et al. 2014). Series 4° 1. Lund: C. W. K. Gleerups förlag. Anselin, L. 1995. “Local Indicators of Spatial Association – LISA.” Geographical Analysis 27: 93–115. Binford, L. R. 1978. Nunamiut Ethnoarchaeology. New York: Academic Conclusions Press. “ The results of the analysis increase the understanding of Boethius, A. 2018. Fishing for Ways to Thrive. Integrating Zooarchaeology to Understand Subsistence Strategies and their Mesolithic site organization and the association with bone Implications Among Early and Middle Mesolithic Southern tool production. Different stages in tool production are Scandinavian Foragers.” Ph.D. diss., Lund University. associated with different areas of the spatially organized Bridault, A., E. David and M. Boboeuf. 2009. “Matter and Material: Red sites. Further, the results show that the bone and antler arti- Deer antler exploitation during the Mesolithic at Clos de Poujol ” fact inventory is affected by the context of recovery. Areas for (Aveyron, France). In In Search of Total Animal Exploitation. Case Studies from the Upper Palaeolithic and Mesolithic Proceedings of deposition of both discarded and still usable artifacts, and the XVth UISPP Congress, Session C61, vol. 42, Lisbon, 4-9 waste products seem to work as integrated parts of the occu- September 2006, edited by L. Fontana, F.-X. Chauvière and A. pational process and everyday life. We can identify knapping Bridault, 135–154. BAR International Series 2040. Oxford: BAR floors for bone crafting in association with knapping floors Publishing. for lithics, whereas larger debitage was thrown away, into Browall, H. 2011. Alvastra Pålbyggnad. 1909-1930 års Utgrävningar. Stockholm: Kungliga Vitterhets Historie och Antikvitets Akademien. the water. In wet environments, preforms and blanks were Carlsson, T. 2008. Where the River Bends: Under the Boughs of Trees. also deposited, probably kept in the shallow water for storage Strandvägen – A Late Mesolithic Settlement in Eastern Middle and preservation for later use. At Strandvägen these areas Sweden. Linköping, Sweden: Riksantikvarieämbetet. along the shore also seem to coincide with fishing stations Clark, J. G. D., and M. W. Thompson. 1953. “The Groove and Splinter and utilization zones for barbed points and simultaneously Technique of Working Antler in Upper Palaeolithic and Mesolithic Europe, with Special Reference to the Material from Star Carr.” functioned for more ceremonial depositions of decorated Proceedings of the Prehistoric Society 19: 148–160. artifacts. Conneller, C. 2006. “The Space and Time of the Chaine Operatoire: More than providing a spatial understanding of the sites, Technological Approaches to Past Landscapes.” Archaeological the implications of the amounts of finds associated with Review from Cambridge 21: 38–39. “ fi bone tool production in the faunal record is important. Costin, C. L. 1991. Craft Specialization: Issues in De ning, Documenting, and Explaining the Organization of Production.” Osseous assemblages need to be analyzed also in relation to Archaeological Method and Theory 3: 1–56. taphonomic parameters and to contextual data in order to Cronberg, C. 2001. “Husesyn.” In Tågerup Specialstudier, edited by P. answer questions about prehistoric utilization. Karsten and B. Knarrström, 82–155. Skånska spår – arkeologi längs Västkustbanan. Lund: Riksantikvarieämbetet. Cunningham, P. 2011. “Caching your Savings: The Use of Small-Scale Acknowledgments Storage in European Prehistory.” Journal of Anthropological Archaeology 30: 135–144. We would like to thank the anonymous reviewers for valuable comments David, E. 1999. “L’industrie en Matières Dures Animals du Mésolithique that improved the manuscript. Ancien et Moyen en Europe du Nord. Contribution de l’analyse Technologique à la Definition du Maglemosien.” Ph.D. diss., Université Nanterre-Paris X, Paris. Notes on Contributors David, E. 2003. “The Contribution of a Technological Study of Bone and fi ” Sara Gummesson Antler Industry for the De nition of the Early Maglemose Culture. (Ph.D. 2018, Stockholm University) is a guest In Mesolithic on the Move. Papers Presented at the 6th International researcher at the osteoarchaeological research laboratory, Stockholm Conference in the Mesolithic in Europe, Stockholm, 4-8 September University and researcher at the Department of Organismal Biology, 2000, edited by L. Larsson, H. Kindgren, D. Loeffler and A. Human Evolution, at Uppsala University, currently working with Åkerlund, 649–657. Exeter: Oxbow Books. depositional patterns of human remains at Stone Age sites in Sweden. David, E. 2007. “Technology on Bone and Antler Industries: A Relevant Her previous work and Ph.D. project concern Mesolithic bone and antler Methodology for Characterizing Early Post-Glacial Societies (9th–8th craft and patterning in osteological assemblages from the Stone Age. Millennium BC).” In Bones As Tools: Current Methods and Central to this research is the study of taphonomy and spatial Interpretations in Worked Bone Studies, edited by C. Gates St- distribution. Pierre, and R. B. Walker, 35–50. BAR International Series 1622. Fredrik Molin (B.A. 1993, Umeå University]) is a Ph.D. student at Umeå Oxford: BAR Publishing. University. He has extensive experience in the field of contract archaeol- David, E., and L. Johansen. 1996. “Maglemosian Barbed Points Made of ogy, and currently works for the Archaeologists at the National Histori- Metapodial: Reconstructing the Chaîne Opératoire by Experiments. cal Museums in Sweden. During the last two decades, his research has Rapport D’étude Expérimentale Haf 26/96.” Unpublished report on mainly concerned remains from the Stone Age, with a focus on hun- file at Lejre Experimental Centre (Denmark). https://halshs. ter-gatherers of the Mesolithic Scandinavia. Of special interest are archives-ouvertes.fr/halshs-00124982. 178 S. GUMMESSON ET AL.

David, E., and M. Kjällquist. 2018. “Transmission of Knowledge, Sweden 7000 BP.” In Mesolithic Burials – Rites, Symbols and Social Crafting and Cultural Traditions, Interregional Contact and Organisation of Early Postglacial Communities, edited by J. Interaction, 7300 Cal BC.” In The Technology of Early Settlement in Grünberg, B. Gramsch, L. Larsson, J. Orschiedt and H. Meller, Northern Europe Transmission of Knowledge and Culture the Early 161–174. Halle (Saale): Tagungen des Landesmuseums für Settlement of Northern Europe, Volume 2, edited by H. Glørstad, K. Vorgeschichte. Knutsson, H. Knutsson and J. Apel, 231–276. Sheffield: Equinox Ingold, T. 1993. “The Temporality of the Landscape.” World Publishing. Archaeology 25 (2): 152–174. David, E., and M. Sørensen. 2016. “First Insights into the Identification Johnson, E. 1985. “Current Developments in Bone Technology.” of Bone and Antler Tools used in the Indirect Percussion and Advances in Archaeological Method and Theory 8: 157–235. Pressure Techniques during the Early Postglacial.” Quaternary Karr, L. P., and A. K. Outram. 2015. “Bone Degradation and International 423: 123–142. Environment: Understanding, Assessing and Conducting David, É., S. McCartan, F. Molin, and P. Woodman. 2015. “‘Fishing’ Archaeological Experiments Using Modern Animal Bones.” Mesolithic Settlement-Subsistence Systems using Notions of International Journal of Osteoarchaeology 25: 201–212. Economy of Debitage and Raw Material toward Production of Karsten, P., and B. Knarrström, eds. 2003. The Tågerup Excavations. Bone Points.” In Muge 150th: The 150th Anniversary of the (Skånska Spår – Arkeologi längs Västkustbanan). Lund: The Discovery of Mesolithic Shellmiddens – Volume 2, edited by N. National Heritage Board. Bicho, D. Price and E. Cunha, 173–188. Newcastle: Cambridge Kjellström, A., and M. Hamilton. 2014. “The Taphonomy of Maritime Scholars Publishing. Warfare: A Forensic Reinterpretation of Sharp Force Trauma from Dormann, C. F., J. M. McPherson, M. B. Arajújo, R. Bivand, J. Bolliger, the 1676 Wreck of the Royal Swedish Warship Kronan.” In G. C. Richard, G. Davies, A. Hirzel, W. Jetz, W. D. Kissling, I. Kühn, Bioarchaeological and Forensic Perspectives on Violence: How R. Ohlemüller, P. R. Peres-Neto, B. Reineking, B. Schröder, F. M. Violent Death is Interpreted from Skeletal Remains, edited by D. L. Schurr, and R. Wilson. 2007. “Method to Account for Spatial Martin and C. P. Anderson, 34–50. Cambridge: Cambridge Autocorrelation in the Analysis of Species Distributional Data: A University Press. Review.” Ecography 30: 609–628. Kjällquist, M., A. Emilsson, and A. Boethius. 2014. Norje Sunnansund. Enloe, J. G. 1993. “Ethnoarchaeology of Marrow Cracking: Implications Boplatslämningar från Tidigmesolitikum och Järnålder. Särskild for Recognition of Prehistoric Subsistence Organization.” In From Arkeologisk Undersökning 2011 och Arkeologisk Förundersökning Bones to Behavior. Ethnoarchaeological and Experimental 2011 och 2012, Ysane Socken, Sölvesborgs Kommun i Blekinge län. Contributions to the Interpretations of Faunal Remains, edited by J. Blekinge Museum Rapport 2014:10. Karlskrona, Sweden: Blekinge Hudson, 83–97. Carbondale: Southern Illinois University. Museum. Galiński, T. 2013. “Typological, Chronological and Cultural Verification Larsson, L. 1978. “The Mesolithic Bone and Antler Artefacts From of Pleistocene and Early Holocene Bone and Antler Harpoons and Central Scania.” Meddelande från Lunds universitets historiska Points from the Southern Baltic Zone.” Przegląd Archeologiczny 61: museum 1976–1977: 28–67. 93–144. Larsson, L. 1983. Ageröd V: An Atlantic Bog Site in Central Scania. Geneste, J.-M. 1991. “Systèmes Techniques de Production Lithique: Acta Archaeologica Lundensia, Series 8° 12. Lund: University of Variations Techno-économiques Dans les Processus de Réealization Lund. des Outillages Paléolithiques.” Techniques et culture 17–18: 1–35. Larsson, L. 2005. “Regional or Interregional Representation? A Slotted Grøn, O. 2003. “Mesolithic Dwelling Places in South Scandinavia: Their Bone Dagger from Offerdal, Jämtland.” In En Lång Historia: Definition and Social Interpretation.” Antiquity 77: 685–707. Festskrift Till Evert Baudou på 80-årsdagen, edited by R. Grønnow, B. 1987. “Meiendorf and Stellmore Revisited: An Analysis of Engelmark, T. Larsson and L. Rathje, 261–273. Umeå: Institutionen Late Paleolithic Reindeer Exploitation.” Acta Archaeologica 56: 131– för arkeologi och samiska studier. 166. Larsson, L., and F. Molin. 2017. “Decoration of Bone and Antler Gummesson, S. 2018. “Points in Production. Taphonomic Research on Artefacts as an Indication of Mesolithic Networks. Finds from a Mesolithic Osseous Assemblages in Sweden.” Ph.D. diss., Central Swedish Late Mesolithic Site.” Самарский научный Stockholm: Stockholm University. вестник (Samara Journal of Science) 6 (3 [20]): 85–91. Gummesson, S., J. Karlsson, H. Kivikero, O. Magnell, and J. Storå. 2017. Larsson, L., and A. Sjöström. 2011. “Hut and House Structures in the “Osteologisk Analys av Djurben Från Strandvägen.” Rapport OFL Mesolithic of Southern Scandinavia.” In Site-Internal Spatial 2017:4. Unpublished report on file at Osteoarchaeological Research Organization of Hunter-Gatherer Societies: Case Studies from the Laboratory, Stockholm University. European Palaeolithic and Mesolithic. Papers Submitted at the Gummesson, S., and F. Molin. 2016. “The Mesolithic Cemetery at Session (C58) “Come In … and Find out: Opening a New Door Into Strandvägen, Motala, in Eastern Central Sweden.” In Mesolithic the Analysis of Hunter-Gatherer Social Organization and Behavior”, Burials – Rites, Symbols and Social Organisation of Early Postglacial Held at the 15th U.I.S.P.P. Conference in Lisbon, September 2006, edi- Communities. International Conference Halle (Saale), Germany, ted by S. Gaudzinski-Windheuser, O. Jöris, M. Sensburg, M. Street 18th-21st September 2013, edited by J. Grünberg, B. Gramsch, L. and E. Turner, 233–247. Mainz: RGZM – Tagungen. Larsson, J. Orschiedt and H. Meller, 145–159. Halle: Landesamt für Lindholm, K.-J., E. Sandström, and A.-K. Ekman. 2013. “The Denkmalpflege und Archäologie Sachsen-Anhalt, Landesmuseums Archaeology of Commons.” Journal of Archaeology and Ancient für Vorgeschichte Halle (Saale). History 10: 1–49. Gummesson, S., F. Molin, and J. Storå. 2017. “Taphonomy, Bone Surface Lyman, R. L. 2012. Quantitative Paleozoology. Cambridge: Cambridge Characteristics and Assemblage History: Finding Mesolithic Bone University Press. Depositions at Strandvägen, Motala.” Journal of Archaeological Magnell, O. 2006. “Tracking Wild Boar and Hunters: Osteology of Wild Science: Reports 13: 11–25. Boar in Mesolithic Scandinavia.” Ph.D. diss., Lund University. Gummesson, S., and F. Molin. In press. “Points of Bone and Antler from Mitchell, A. 2005. The ESRI Guide to GIS Analysis. Volume 2: Spatial the Late Mesolithic Settlement in Motala, Eastern Central Sweden.” In Measurements and Statistics. Redlands, CA: ESRI Press. Working at the Sharp End: From Bone and Antler to Early Mesolithic Molin, F., G. Gruber, and L. Hagberg. 2014. “Motala – A North Life in Northern Europe, edited by D. Groß, H. Lübke, J. Meadows and European Focal Point?” In Lateglacial and Postglacial Pioneers in D. Jantzen. Untersuchungen und Materialen in Schleswig-Holstein Northern Europe, edited by F. Riede, and M. Tallavaara, 91–102. und im Ostseeraum 10. Kiel/Hamburg: Wachholtz / Murmann BAR International Series 2599. Oxford: Archaeopress. Publishers. Molin, F., H. Hagberg, and A. Westermark. 2018. “Living by the Shore: Hagberg, L., and A. Westermark. 2015. “Material Culture and Mesolithic Mesolithic Dwellings and Household in Motala, Eastern Central Actions in Space: Results of the Spatial Analysis of the Sweden, 5600–5000 cal BC.” Journal of Archaeological Science 18: Verkstadsvägen Site in Motala.” In Muge 150th: The 150th 913–924. Anniversary of the Discovery of Mesolithic Shellmiddens – Volume 2, Montelius, O. 1994 (1917). Minnen Från vår Forntid. I Stenåldern och edited by N. Bicho, D. Price and E. Cunha, 219–234. Newcastle: Bronsåldern. Gamleby, Sweden: ARKEO-Förlaget. Cambridge Scholars Publishing. Nelson, M. C. 1991. “The Study of Technological Organization.” In Hallgren, F., and E. Fornander. 2016. “Skulls on Stakes and Skulls in Archaeological Method and Theory, Volume 3, edited by M. Water. Mesolithic Mortuary Rituals at Kanaljorden, Motala, Schiffer, 57–100. Tucson, AZ: University of Arizona Press. JOURNAL OF FIELD ARCHAEOLOGY 179

Newcomer, M. 1976. “Experiments in Upper Palaeolithic Bone Work.” Semenov, S. A. 1976 (1964). Prehistoric Technology. An Experimental In Methodologie Appliquee a L’industire de l’os Prehistorique. edited Study of the oldest Tools and Artefacts from Traces of Manufacture by H. Camps-Fabrer, 293–301 Paris: C.N.R.S. and Wear. 4th ed. Wiltshire: Moonraker Press. Noe-Nygaard, N. 1977. “Butchering and Marrow Fracturing as a Sjöström, A. 1997. “Ringsjöholm: A Boreal-Early Atlantic Settlement in Taphonomic Factor in Archaeological Deposits.” Paleobiology 3 (2): Central Scania, Sweden.” Lund Archaeological Review 3: 5–20. 218–237. Smith, A. B., and C. Poggenpoel. 1988. “The Technology of Bone Tool Olson, C., N. Björck, and J. Storå. 2011. “Huts and Deposition of Refuse Fabrication on the South-Western Cape, South Africa.” World at Fräkenrönningen, a Neolithic Coastal Dwelling Site in Eastern Archaeology 20 (1): 103–115. Middle Sweden.” International Journal of Osteoarchaeology 21: Stapert, D. 1989. “The Ring and Sector Method: Intrasite Spatial Analysis 173–186. of Stone Age Sites, with Special Reference to Pincevent.” Osipowicz, G. 2007. “Bone and Antler: Softening Techniques in Palaeohistoria 31: 1–57. Prehistory of the North Eastern Part of the Polish Lowlands in the Stewart, H. 1996. Stone, Bone, Antler and Shell: Artifacts of the Northwest Light of Experimental Archaeology and Micro Trace Analysis.” Coast. Vancouver: Indiana University. euroREA 4: 1–22. Stiner, M. C., S. L. Kuhn, S. Weiner, and O. Bar-Yosef. 1995. “Diﬀerential Outram, A. 2001. “A New Approach to Identifying Bone Marrow and Burning, Recrystallization and Fragmentation of Archaeological Grease Exploitation: Why the Indeterminate Fragments should not Bone.” Journal of Archaeological Science 22: 223–237. be Ignored.” Journal of Archaeological Science 28: 401–410. Tacher, D., S. B. Milne, and P. Park. 2017. “Applying GIS and Statistical Pelegrin, J. 2012. “New Experimental Observations for the Analysis to Assess the Correlation of Human Behavior and Characterization of Pressure Blade Production Techniques.” In The Ephemeral Architectural Features among Palaeo-Eskimo Sites on Emergence of Pressure Blade Making: From Origin to Modern Southern Baﬃn Island, Nunavut.” Journal of Archaeological Science Experimentation, edited by P. Desrosiers, 465–500. London: Springer. 14: 21–30. Perlès, C. 1993. “Ecological Determinism, Group Strategies, and Vankina, L. 1999. The Collection of Stone Age Bone and Antler Artefacts Individual Decisions in the Conception of Prehistoric Stone from Lake Lubãna Riga, Latvia: History Museum of Latvia. Assemblages.” In The Use of Tools by Human and Non-Human Voruz, J. 1984. Outillage Osseux et Dynamisme Industriel Dans le Primates, edited by A. Berthelet and J. Chavaillon, 266–277. Néolithique Jurassien. Cahier d’archéologie Romande 29 [Bone tools Oxford: Clarendon Press. and Industrial Dynamism in the Jura Neolithic. French Archaeology Persson, C. 2012. “Den Hemliga Sjön – en Resa till det Småländska Notebook 29]. Lausanne, Switzerland: Bibliothèque Historique Inlandet för 9000 år Sedan.” Ph.D. diss., University of Göteborg. Vaudoise. Pickering, T. R., and C. P. Egeland. 2006. “Experimental Patterns of Winter-Livneh, R., T. Svoray, and I. Gilead. 2010. “Settlement Patterns, Hammerstone Percussion Damage on Bones: Implications for Social Complexity and Agricultural Strategies during the Chalcolithic Inferences of Carcass Processing by Humans.” Journal of Period in the Northern Negev, Israel.” Journal of Archaeological Archaeological Science 33: 459–469. Science 37: 284–294.