Microblade Technology in Northern Sweden Chronological and Cultural Irnplications

This article deals with Mesolithic microblade technology in northern Sweden. The artifacts in question are keeled scrapers, microblade cores, i.e. handle cores (also called wedge-shaped cores) and conical/cylindrical microblade cores, and microblades from Norrland and the provinces of Dalarna and Värmland. It is proposed that microblade production from handle cores was introduced perhaps as early as 7700/7500 BP in northern Sweden, but at least some time during the period 8000—7000 BP. It is possible that this type of core survives right up to ca. 5500 BP. The north Swedish handle core tradition is compared with its neighboring cultures. It is argued that microblade production from oblong handle cores was an innovation that spread from southern Scandinavia or southeastem Norway/ western Sweden to northern Sweden during the Early Atlantic period. The Scandinavian handle core tradition as a whole is further compared with its counterparts in northeastern Asia and North America.


INTRODUCTION
This article is based on parts of a more thorough study on microblade technology in northern Sweden originally published in 1995 (Olofsson 1995), with an addition of the latest research on the subject.
The main focus is on artifacts, namely microblade cores, microblades and so-called keeled scrapers found in northernmost Sweden, that is Norrland, and the provinces of Dalarna and Värmland in upper middle Sweden (Fig. 1). These artifact types have been recovered from practically the whole of Scandinavia down to Denmark and northern Germany. They are frequently found together and form the so-called microblade complex, which in Scandinavia is usually dated to the Mesolithic (the Boreal and Atlantic chronozones).
The original study is a statistical analysis of 676 artifacts from Swedish museums and other collections with a detailed description of technology, raw material, geographical distribution, etc. , much of which, however, for practical reasons must be omitted here. Current Swedish Arehaeology, Vol. 10, 2002 without being retouched into microliths; this applies especially to their use as inserts.
The microblades were produced, most likely by pressure technique, from handle cores (Sw. handtagskärnor, see Fig.   4:1)and conicaVcylindrical microblade cores (see Fig. 4:2 which is a conical core, the cylindrical core has two opposite striking platforms). Closely associated with the handle core is the so-called keeled scraper (see Fig. 4:5), which by some scholars is judged to be a preform for a handle core, and by others as an actual scraper. The main focus is on the handle core since this type of core has been involved in much speculation about the diffusion of ideas and the cultural affiliation among the Stone Age cultures of the northern Boreal/Arctic zone. The microblade complex has a circumpolar distribution. The handle cores, also called wedge-shaped cores, campus-type cores, Gobi cores and so on, are found not only in most of   Cnrrent Sivedish Archaeotogy, Vot. lO, 2(i(i2 Scandinavia but also in a wide area stretching from Mongolia, northeastern Russia, across Japan and into North America. The north Swedish handle-core tradition is compared with its counterparts in the rest of Scandinavia, but also with equivalents in the more remote areas mentioned above. Another interesting aspect of the microblade technology in northern Sweden is that it has been associated with the first more substantial immigration into the area aAer the deglaciation of the Weichselian ice sheet. Also that aspect is touched upon. CHRONOLOGICAL REFLECTIONS Well into the 1960s many archaeologists were doubtful whether Norrland was populated during the Mesolithic, i.e. before 5000 BP (e.g. Janson 1966:22). However, in 1964 the C 14-dates from Döudden were published, and they were clearly Late Mesolithic -6105+100 BP and 6115+100BP (Forsberg 1985:4).The year 1969 marked the start of the excavation that probably has been most influential on the re-evaluation of the prehistory ofNorrland, namely the excavation at Garaselet, Jörn parish, Västerbotten.
Among the finds worth mentioning are handle cores, keeled scrapers, and microblades, artifacts that had been found earlier all over Norrland but were then assumed to be anachronistic products. Now the artifacts could be dated with three C 14-dates to ca. 8000 BP (Sundqvist 1978).After that nobody could doubt that Norrland was populated also in the Mesolithic.
During the last 25 years the north Swedish handle core tradition has been dated to ca. 8000 BP. This date is, however, based on only these three dates (8160+110BP, 8040+100 BP, and 7885+300 BP) from the Garaselet site. Moreover, in recent years it has been questioned whether the artifacts concerned here, the handle cores/keeled scrapers and the microblades, can be tied to these dates from the oldest layer of the site, or if they should be linked to a younger phase dating from ca. 6000 BP (Knutsson 1993). One of the primary aims of the investigation has been to bring more data into the discussion and, ifpossible, get a firmer grasp ofthe chronology regarding these artifacts. Therefore 17 sites (Tab. 1) have been analyzed from a chronological point of view. From 15 of them there are artifacts included in the statistical part of this investigation.
These 17 sites have been analyzed and discussed with respect to C 14-dates, find context, and similar factors of importance to the matter. Unfortunately, this is not the right place to give a detailed account of that, but the chronological position ofthe sites is summarized in Table 1. (Methodological questions related to dating of Norrlandic Stone Age sites are discussed in Olofsson 1995. ) However, I will give a short presentation of the relevant material.
Andersson offers two different alternatives for interpretation. The first is that the differences between area 1 and area 2/3 should be interpreted in chronological or seasonal terms; the other is that the differences mirror a social division of a small group (Andersson 1999:95).If we take into consideration what has been said above concerning the different nature of the areas, what seems most likely to me is that the differences reflect chronological disparity. Thus, area 1 probably dates from the Neolithic and area 2/3 from the Mesolithic (i.e. Andersson's first hypothesis, p. 77).
The handle core fragment and three certain microblades (ofwhich one is retouched) were found in the same square as the pit A 2 (244/582), and the other rejuvenation flake and the rest of the microblades were found in this and adj acent squares (244/580 and 244/581). The microblades in 244/580-581 were found in layers 0-0.20 and 0-0.10 m and the unspecified rejuvenation flake in 244/580 in layer 0-0.20 m, but there is no information about layer regarding the microblades and the handle core fragment from 244/582 (or any other flakes from that square). Melander mentions that "some flakes were also found in the pit" (Melander 1981a, my translation), but in the excavation report he does not separate the handle core fragment and the microblades in 244/582 from ordinary flakes in that square, and thus we are left in ignorance as to whether "our" artifacts were found in the pit itself or only in the same square meter but outside the pit. The cooking pit was covered with sand, and Melander thinks that this is due to a deliberate covering of the pit for hygienic reasons. Andersson does not disagree with that, but she adds a further possibility: that the artifacts, while the sandy ground surface was without vegetation due to heavy trampling, had been eroded from higher lying parts of the site down to the hollow where the pit was situated (Andersson 1999:77). Finally, it can be concluded that in spite of some uncertaintywe cannot be sure of a 100% connection between the artifacts and the datethis is as close to a sealed find that one can get in these circumstances. For now, the most plausible hypothesis must be that there is a connection between the artifacts and the date of 7715+115BP.
In the excavation of a site at Lake Limsjön, Leksand parish, Dalarna six keeled scrapers, two handle cores, four microblade cores, and as many as 263 microblades were found. Five of the keeled scrapers are of porphyry, while the sixth is made of reddish sandstone. The two handle cores are made of porphyry and tuff respectively. The artifacts are not contained in the statistical part of this investigation. Two of the C 14-dates from the site indicate the Stone Age: 7640+85 BP from hearth A 17 and 5000+100 BP from post-hole A 15. The remaining six C 14-dates from the site belong mainly to the Late Iron Age. The archaeological record unambiguously indicates the Mesolithic, while the hearths with Iron Age dates probably should be connected with clearance activities. To sum up, Larsson dates the site to ca. 8000/7500 -7000 BP (Larsson 1994).
A site at Nyluspen I:10,RAÄ no. 553 in Vilhelmina parish, Lapland has produced similar artifacts (Spång 1983). The site, which was excavated in 1977, has remains &om several periods and the dates span from 8180 to 450 BP. Two handle cores were found within the squares 120/326 and 328. A charcoal sample, dated to 5570+65 BP, was taken at 120.9/327. 8 -328. 1 in a cooking pit (?) with bones and fire-cracked stones, at a depth of 0.20 m (Spång 1983: charcoal sample list), and thus in proximity of the handle cores. However, the two handle cores, ofwhite quartzite/brecciated quartz ( Fig.  4:1)and flint or fine-grained igneous/volcanic rock (Sw. hälleflinta) respectively, were found at a depth of 0.05 m. Also a keeled scraper of quartzite was found at the site. One can, of course, also assume that the handle cores, instead of being tied to the date of 5570 BP, could be connected with a cooking pit, A 8, situated 20 m to the east. A 8 is dated to 7580+115 BP, and that would perhaps be in better accordance with the known age of these artifacts. But then we have to abandon the demand for absolute spatial connection between the find and the date. These early dates are both few and more or less uncertain. The date from Högland (7715 BP) is afflicted with a rather high standard deviation of +115 years. The date (7640 BP) from Limsjön is difficult to evaluate on the basis ofthe available information, even if it seems plausible for typological reasons (although I am quite aware ofthe risk of arguing in a circle here). Assigning the handle cores from Nyluspen to the 7580 BP date is admittedly speculative, since they were, in fact, found closer to the date of 5570 BP than the date of 7580 BP. But since it is not explicitly stated in the excavation report that both the handle cores and the dated charcoal came from a sealed context (and also the vertical position differed, see above), the most sensible conclusion must be that both dates are possibleespecially ifwe consider the late dates ofthese artifacts from e.g. Vuollerim in Jokkmokk parish (see below).
There are dates from the later half of the period 8000 -7000 BP as well. At a site  (Jennbert 1984).
An excavation of a site (Luspen,RAÄ no. 190)at the southern part of Lake Vojmsjön in Vilhelmina parish, Lapland yielded among other things a handle core and a conical microblade core, both of grey quartzite. A cooking pit at the site gave a date of 7070+75 BP. The conical microblade core was found only ca. 1 m from this cooking pit and thus seems to have a spatial connection to the date. The handle core was found at another part of the site. However, it is not improbable that also the handle core is linked to the date of 7070 BP (Rydström 1980). The duration ofthe handle core tradition is unclear. There are dates from the period 7000 -6000 BP. At Lake Kårtjejaure, site 1371 in Gällivare parish, Lapland a certain handle core and three "keeled scrapers" (the handle core and two of the keeled scrapers are included in this investigation) were found during excavation which possibly can be linked to a C 14-date of 6330+80 BP (Forsberg 1985:55 -58, 243 -248). One ofvery few Norrlandic sites with a stratigraphic sequence, Döudden in Arjeplog parish, Lapland, has yielded finds oftwo handle cores/keeled scrapers of dark quartzite and fine-grained, silica-rich igneous/volcanic rock (Sw. hälleflinta) respectively in the oldest layer, which is C 14-dated to ca. 5200 -3800 cal BC (Bergman 1995:95, 195 There are signs indicating that this kind of artifacts could have existed until ca. 5500/5000 BP, e.g. a handle core and several microblades from the house remains at Vuollerim, Jokkmokk parish, Lapland (Loeffler & Westfal 1985:433 -434;Loeffler 1998:71).In this context we can also remind ourselves of the specimens from RAÄ no. 553, Vilhelmina parish, mentioned aboveif we choose the late alternative, 5570 BP. A true outlier is a specimen from Kungaudden, Sorsele parish, Lapland since this site is typologically dated to the Bronze Age (Christiansson 1969). This handle core must be either a mysterious rarity in the Bronze Age or else the site also contains older material.
The so-called keeled scrapers seem to be more or less synchronous with the handle cores. However, it seems that keel-shaped scrapers (but with a less marked keel and perhaps also lower, shorter, and wider than the Mesolithic ones) were present in the Neolithic as well. Many of them are made of quartz, and it is possible that some of them are, in fact, flake cores instead of scrapers (cf. Broadbent 1979:69 -71;Halén 1994:101, Fig. 106 and pp. 126 -127 regarding platform cores of quartz). In spite of this, many of them are true scrapers. It remains, however, uncertain whether there is any closer cultural or functional relationship between these scrapers and the "classical" keeled scrapers.
Little is known about the age of the conical microblade cores in northern Sweden.
There is only one fairly certain date for this artifact category in the investigated material the above-mentioned 7070 BP from Luspen, RAÄ no. 190 in Vilhelmina parish, Lapland (Rydström 1980). It has often been claimed that the conical microblade core was introduced somewhat earlier than the handle core (Welinder 1971:118;Bille Henriksen 1976:121;Larsson 1978:149).In Denmark it is dated by Brinch Petersen (1973:126)to 8300 -7600 BP, and in that case it has a relatively short time span. The change from conical microblade cores to handle cores seems to have occurred around 7000 BP in southeastern Norway, judging from excavated sites in the OslofJord area (Ballin 1998:123).
It appears as if blades and microblades are a more timeless kind of artifact that was manufactured from different types of cores throughout time, and it is possible that the production of small blades occurred also during the Neolithic/Bronze Age (Bastuselforsen, Arvidsjaur parish; Skalka, Jokkmokk parish; and Åseleforsen, Åsele parish; see Tab. I). However, it is likely that most ofthe microblades from these late sites were fortuitously produced, since "microblades" can result from other kinds of stone knapping, e.g. pressure retouching of bifacially worked points, manufacturing ofburins, retouching of scrapers, etc. (Cook 1968:123,125;Nygaard 1974:26;Callahan 1987:41-42), without being connected with the so-called microblade tradition, neither functionally nor culturally. Despite that, microblades are so strongly associated with the Mesolithic microblade tradition that for excellent reasons it can be assumed that the majority of these artifacts belong to this specific cultural context. Early dates of microblade finds but without microblade cores have been obtained from Foskvattnet, At the Lidsbron site, Sunnemo parish, Värmland, a pit (rubbish pit?) with charred hazelnut shells was excavated in 1998.The hazelnuts were dated to 6965+80 BP. This feature also included lithics, among other things microblades. In addition, several other microblades and a few microblade cores, at least one of which can be classified as a handle core, were found at the site, however without any clear association with a dated feature. There are also dates ranging from 7920+70 BP to 7765+80 BP &om the site, and it is at present unclear whether the microblades and the microblade cores should be attributed to these dates or to the date of6965 BP mentioned above. (Olsson 2000:17-18).Another site in Värmland with microblades and microblade cores is Kvarnåsen in Norra Råda parish, which with typological and geological methods can be dated to roughly 7000 BP (Olofsson k Olsson 1999).

COMPARATI VE STUDIES
In the following a comparison is made between northern Sweden and surrounding cultures regarding the relevant arti fact types; the main focus is, however, on the handle core. The areas concerned are southern Scandinavia, western Sweden, southern and middle Norway, northern Norway, eastern middle Sweden, Finland, and the southern coast of the Baltic. Furthermore, a brief overview is given of more remotely situated microblade cultures in northeastern Asia and North America.
The handle core is considered to appear ca. 8000 BP in southern Scandinavia (Denmark and Scania). Together with the other arti fact categories in this investigation, it constitutes a characteristic feature of the younger Maglemose culture, 8300/8000-7500 BP, and the Kongemose culture, ca 7500 -6500 BP (Becker 1952:143 -144;Bille Henriksen 1976;Vang Petersen 1984).The Norrlandic specimens show strong similarities with the south Scandinavian ones. However, there seem to be some differences. For readers not familiar with Scandinavian archaeology, it can be noted that one such difference is that most specimens (as well as many other artifact categories) are made of flint in the southern parts of Scandinavia, while in the northern parts the raw material is more mixedhere a variety of other fine grained materials as volcanic/igneous rocks, quartzite, jasper, and to some extent quartz were used instead (although specimens of flint do exist in northern Sweden). This is due to the fact that, in Scandinavia, flint suitable for tool making is found almost entirely in the southernmost areas (Denmark and Scania), in chalk deposits in the bedrock. Flint occurs also along the Swedish west coast and the Norwegian coast, transported there by drifting ice as well as in moraine deposits (see e. g. Berg-Hansen 1999:255 -266). Judging from the literature the south Scandinavian handle cores seem to be somewhat larger than their Norrlandic counterparts (comparisons made with data in Andersen 1985:21;Callahan 1985:34).Furthermore, it is possible that a platform angle over 90 degrees is more common in southern Scandinavia than in Norrland (comparisons made with data in Callahan 1985:25;see also Callahan 1984and illustrations in Andersen 1985:28 and Brendsted 1966.Another interesting fact is that microliths, which are so intimately associated with the microblade tradition in southern Scandinavia, are extremely rare in northern Sweden. Cnrrent Swedish Arehaeology, Vol. l0, 2002 As far as triangular (scalene) microliths are concerned, they are to my knowledge non-existent in Norrland.
Handle cores occur also further north in southern Sweden, for example in Småland (Lidén 1936:155 -213) and in Västergötland (Fredsjö 1953:156;Kindgren 1991).At the site of Högby in Östergötland (near Lake Vättern) finds of handle cores are correlated to an occupation phase that is C14-dated to ca. 7600 -6000 BP (Carlsson et al. 1999:56 -57 where an uncertain handle core was found in a cultural layer dated to 8180+170 BP (Mikkelsen 1975b:29;Ballin 1999:208).Torben Bjarke Ballin has re-examined the assemblage from Terkop, and he states that there are no handle cores from the site and that the supposed handle core might instead be a rough-out for a conical core (Ballin 1999:208).It has long been an established archaeological "truth" that the handle core was introduced on the Swedish west coast ca. 7000 BP, or after the maximum postglacial transgression. However, in an excavation of a site called Dammen in Bro parish, Bohuslän, with dates ranging from roughly 8000 to 7500 BP, a handle core was recovered (Kindgren 8c Schaller Åhrberg 1999:225,231).Furthermore, microblades "which most likely have been struck from a handle core" have been found at a site dated to 8200 BP at Munkeröd in southern Bohuslän (Kindgren & Schaller Åhrberg 1999:231).Taken together, this indicates that the appearance of the handle core on the Swedish west coast may have taken place somewhat earlier than previously thought. The west Swedish (e.g. Alin 1955:66,Fig. 20;111,Fig. 37;112, Fig. 38), and perhaps even more so, the east Norwegian material (Mikkelsen 1975a;Mikkelsen 8c Nybruget 1975; Gustafson 1987) show strong similarity to the north Swedish finds.
In the west Norwegian Fosna culture handle cores seem to be comparatively rare. In this area conical and cylindrical cores were most frequently used. A cylindrical microblade core from Linsell parish, Härjedalen bears a strong resemblance to a core from Åmot in Hedmark fylke, eastern Norway (Mikkelsen k Nybruget 1975:101, Fig.   9). It is possible that we have a connection here, since this type of core (at least in this particular context) is believed to be a west Norwegian innovation in the Late Mesolithic that diffused south and eastwards (lndrelid 1973:79;Nygaard 1974:25 -29).
In the Komsa culture in northern Norway there is, to all appearances, an early (ca. 9000 -7500/7000 BP) microblade production from conical and cylindrical cores (Woodman 1993:72 -75; Olsen 1994:31 -35). It is, however, uncertain whether typical handle cores of south Scandinavian and north Swedish type exist here. Keeled scrapers do, however, seem to occur (see Olofsson 1995:116 -118 for a more comprehensive discussion concerning northern Norway; see also Siiriäinen 1982:8).
Also regarding Finland it is uncertain whether any true microblade production from handle cores, in the sense we are implying here, has taken place. The only convincing case that I know of is a specimen from Lake lnari, Finnish Lapland, made of finegrained, probably non-local, rock (Siiriäinen 1982). It should be noted that there are finds of quartz from the Askola/Suomusjärvi culture (9200 -6200 BP) that have been classified as handle cores (Luho 1956(Luho , 1967Schulz 1990), but these have been questioned by several researchers (cf. Knutsson 1993:11 -12;Siiriäinen 1982:8;Edgren 1993:28).It deserves mention that Siiriäinen, regarding the specimen from Inari, believes that the core or the technotradition behind it could have come from northern Norway or Sweden into Finland. He does not consider it likely that it could have reached the area from the Finnish or East Carelian region, "as the technique in question does not seem to have had a uniform occurence in these areas" (Siiriäinen 1982:8).Siiriäinen mentions a few east European finds, but he thinks that they " are not typical because they are not related definitely to blade striking" (Siiriäinen 1982:8).
In central and eastern middle Sweden handle cores occur (Welinder 1977;Larsson 1994:247),and regular handle cores of quartz (not anvil technique) have been found in middle Sweden (Knutsson 1993:12).In the latest part of the Mesolithic there seem to have been influences from eastern middle Sweden northwards into the eastern parts of Norrland. This can be seen from the typical round-butted axes (Sw. trindyxor) that are so frequent in eastern middle Sweden and Dalarna and also occur in the eastern parts of southern and middle Norrland. Our artifacts, however, cluster mainly in the western parts. It is possible that here there are cultural and/or chronological differences (see also Baudou 1992:57;Lindholm k, Runeson 1990:12 -14, Fig. 3 and Fig. 42). Nevertheless, in a recent article by Knutsson et al. (1999:99,96, Fig. 4) there is mention of a site with early dates of handle cores also in the southeastern part of Norrland-Gardsjösund in Hälsingland, C 14-dated to 8000 -7500 BP. Also south of the Baltic handle cores are found. The cultural groups with handle cores in northern Germany and Poland (primarily the Oldesloe culture 7400 -5200 BP, the roughly synchronous Junsdorf group, and the Pomeranian group, see Gramsch 1973; Schwantes 1939;Schwabedissen 1944) are considered to belong to the Maglemosian technocomplex, and most probably these artifacts can be associated with Maglemose/Kongemose in southern Scandinavia.
In the east Baltic cultures of interest in this context, namely the Kunda culture (primarily Estonia and Latvia, ca. 9500 -5000 BP, see Kozlowski 1988)  conical and cylindrical blade and microblade cores (see e. g. Gurina 1987:37and Sumkin 1990 Fig. 2.32.). In addition to Siberia, these artifacts are found in Mongolia (Nelson 1935:356;1937:267 -272), northern China (Hayashi 1968:166 -169;Chen & Olsen 1990;Chang 1986:60), and Japan (Hayashi 1968;Kobayashi 1970). This Northeast Asian-Northwest American Microblade tradition, NANAMT (Smith 1974), seems to have been established in the Angara and Aldan area some time after 18,000 BP (this area, together with Mongolia and northern China, is usually seen as the core area of this particular technotradition). From there it spread through far-eastern Siberia, Mongolia, and northern China to Japan where it existed at least 15,000 BP (Smith 1974:351 -352).

THEORIES OF MIGRATION AND CULTURAL CONTACTS
During the last decades there have been mainly two opposing views on the earliest colonization of Norrland. Since the handle cores, at least since the above-mentioned Garselet specimens were recovered in the early 1970s, have been so closely associated with this earliest phase, these two schools of thought have been applied to the handle core tradition as well.
The one school of thought claims a migration/diffusion of ideas from the south, southwest, or west into Norrland. It could here be fruitful to take a look at some of the research connected with it. Evert Baudou has treated this issue in his "The Prehistory ofNorrland" (Not"rlands forntid) published in 1992.Baudou emphasizes the similarities between artifacts ofthe Fosna culture in middle Norway (primarily Trendelag) and the Norrlandic ones (Baudou 1992:54 -56).
Kjel Knutsson (1993)separates the earliest phase from the handle core phase. He illustrates that with the Garaselet site, which in his opinion has one early phase ca. 8000 BP and one later phase with handle cores, dated to around 6000 BP (although he is not willing to apply the date of ca. 6000 BP as the date ofintroduction of the handle core in Norrland). Knutsson states that there is a consensus among archaeologists that the handle core tradition originated and developed in southern Scandinavia during the Late Boreal -Atlantic chronozones. Furthermore, Knutsson hints at the possibility that the supposed earliest phase consists of hunting parties from the Norwegian Fosna culture visiting the area in the Late Boreal.
Also Lars Forsberg (1996)sees an early period before the handle core. By means of statistical analysis of 33 Norrlandic sites dating from the Mesolithic, he obtains three main groups: 1) Sites with exotic raw materials, borers, grey color, and a dating from the early period (8600 -7400 BP).2) Microblade technology (microblades, handle cores/ keeled scrapers) quartzite dominance, heterogeneous petrography, and a predominantly dark color of the raw material (7400 -5800 BP). 3) Quartz technology, white color, beaver and elk bones, and a dating from the late period (5800 -5400 BP). Forsberg points to the fact that the oldest sites occur in a zone in central Norrland (northern Jämtland, southern Lapland, and Västerbotten). This is not far from the Trendelag area in Norway, with old dates like Ålbusetra (8800 BP). In this area there seems to be an expansion of the Fosna groups from the coast inland. There seems to be a gradual transition from the typical grey flint of the coastal sites to more local raw materials such as quartzite. This tallies with the earliest material from the Norrlandic sites, which contains local raw materials but also a few instances of exotic raw materials, amongst others grey flint. Thus, Forsberg suggests (as does Baudou, above) that the earliest migration into Norrland may have come from the Trendelag area. He also suggests a movement from the southwest 6 -700 years later, which brought the handle core technology into the area. Forsberg also hints at the possibility of an early migration from northern Finland and the northern part of Russian Carelia into northeastern Norrland (Forsberg 1996:248 -249).
Mats P. Malmer (1992) equates the microblade inventory from northern Sweden with that from southern Scandinavia, and brings up the notion of an all-embracing Maglemose culture in the whole of Scandinavia.
The other school of thought emphasizes the easterly connections, and opens the door for a diffusion from the east and northeast. Lars Larsson (1978:176)calls attention to the fact that the handle core has a circumpolar distribution. He also brings up the notion that, since the specimens &om Norrland (e.g. Garaselet) seem to be of the same age as the south Scandinavian ones and the specimens from Siberia and Alaska even older, we could here be dealing with a distribution from north to south. Also other researchers have touched upon the idea of a connection to the northeast Asiannorthwest American wedge-shaped core technology (e.g. Broadbent 1982:27 28;Andersen 1985:45;Schulz 1990:22). Already in 1968 Helge Larsen, in a paper presented at the eighth International Congress of Anthropological and Ethnological Sciences, pointed to the common components ofthe circumpolar microblade tradition: the bone and antler shafts (points) with microblade inserts, burins to make the groove for the inserts in the points, and the specialized microblade core. Larsen also noticed the vast distribution of the slotted bone point, from Trans Baikal in the east to Belgium in the west, roughly in a belt between the parallels 50 and 60 N, and that wedge-shaped cores occur "in Mongolia, at Verkholenskaya Gora and in the Maglemose culture" (Larsen 1968:338 -340).
E. James Dixon, building upon Larsen 1968, pays attention to the fact that these points are actual composite tools, and that they represent a quite different conceptual approach to making a projectile compared with the method of producing a point by means of reducing a blank to the desired shape (Dixon 1985:59). Both Loeffler & Westfal (1985:432 -434) and Halén (1994:105 -122), in their presentation of the Vuollerim and Lillberget sites respectively (both sites situated in Norrbotten, the latter a comb ceramic site and thus Neolithic), assume that the flint from the localities could be of Russian origin. Halén also thinks that the flint from the Mesolithic site of Alträsket, Norrbotten, closely resembles the Russian flint (Halén 1994'. 43). Lennart Falk (1997b) focuses on early sites on the Bothnian coast from the Stockholm area in middle Sweden and northwards along the coast of Norrland. If the sites were related to the ancient beach, they should be dated to ca. 6500 -5000 BC (some possibly even earlier). These sites are characterized by quartz inventory and bipolar reduction technique. Falk interprets this as indicating eastern influences (from the Finnish Suomusjärvi culture) that are just as early as the western ones, bringing unifacial platform technology with handle cores/keeled scrapers of fine cryptocrystalline materials (see also Falk 1997a). CONCLUSION Before I sum up, it is important to look at the possibility of an indigenous development of a handle core industry in northern Sweden. It is theoretically possible that this could have happened during the period between the deglaciation and the appearance of the handle core. However, this lacuna does not seem to have been a very long one. The available data suggest that most of Norrland was ice-free around 8100 BP (Forsberg 1996 BP). No handle cores can safely be connected with these dates, and it is possible that these finds belong to a period before the handle core tradition (cf. Sundlin 1986;Knutsson 1993:37;Forsberg 1993;Forsberg 1996).
We must also consider the microblade production from conical cores, which, at least in the northernmost areas, might be connected with the Komsa culture. For the time being we can only speculate about that, however.
What can be said with some degree of certainty is that in the eastern parts of southern and central Norway, on the Swedish west coast, in southern Scandinavia including eastem Denmark, and in northern Germany-Poland, microblade technology with handle cores and keeled scrapers becomes a characteristic tool-kit during the period 8000-7000 BP. As regards northern Norway and Finland, there are doubts about whether unambiguous handle cores exist (except for a few isolated finds). From middle Sweden there does not seem to have been any substantial influence until the Late Mesolithic. The conclusion must be that the handle core tradition during the Early Atlantic period was spread, by the diftusion of ideas and/or actual migration, from southern Scandinavia and/or the Swedish west coast-eastern Norway northwards through western Norrland.
As we have seen, there is a microblade tradition including wedge-shaped cores in northeastern Asia and in North America. It should, however, be mentioned that in spite of a striking resemblance there are some differences in the way the cores were shaped. In Scandinavia it seems as if the reduction in most cases started with the making of a striking platform. Then the rest of the core was shaped (Knutsson 1980:86, Fig. 1;Andersen 1985:23 -24, Fig. 5, see also Olofsson 1995). In northeastern Asia and in North America it was customary to reduce the blank into a bifacially worked, "almond shaped" preform, and after that, as the second step, the platform was produced, but even in these areas different variants of the type where the platform blow precedes the core shaping occur (Kobayashi 1970;Smith 1974).It is, however, possible that there is a closer connection between the Asian-American microblade tradition and its Scandinavian counterpart. In such case, southern Scandinavia could be looked upon as a "melting pot" where easterly strains like an elaborated microblade technology on handle cores, in which the microblades are used as they are without being retouched into microliths, meet the European microlith tradition with narrow triangular microliths (Sauvetterian) and broad microliths (Castelnovian/Tardenoisian).
There is one additional phenomenon that could be of interest in the matter. As I have hinted at earlier, artifacts resembling handle cores/keeled scrapers occur in some Upper Palaeolithic cultures in western and central Europe, e.g. the Aurignacian complex, ca. 36 -21,000 BP (Westerby 1927:56 -62;Muller-Karpe 1966:307and Tafel 205, 207;Champion et al. 1984:48 -54;Gamble 1986:245).It seems to be unclear whether these artifacts should be regarded as core-scrapers or as cores for bladelets (lamelles Dufour), or ifboth functions are represented in the assemblages. Furthermore, there is a steadily increasing microlithization including a bladelet component among the European Upper Palaeolithic cultures, especially from ca. 20,000 BP onwards (Gamble 1986:136,183, 199, 220). The relationship between the Mesolithic handle core tradition in Scandinavia Current Swedish A rchaeology, Vol. l 0, 2002 and both the Asian-American wedge-shaped core tradition and the small-blades of the European Upper Palaeolithic needs further investigation before we can definitely declare anything on this point.
Finally we have to consider the possibility of an autonomous development in southern Scandinavia during the Late Boreal.
POSTSCRIPT It involves considerable risks to compare like this "in all directions" through both time and space. Formal and even technological similarities do not necessarily indicate cultural affinity. Is it possible that the idea of an oblong microblade core with invariable working face radius could have arisen at different occasions and at different places independent of each other? And even if these similar, although technically somewhat different (varying core-shaping methods), handle core/wedge-shaped core traditions should in fact be the results of a diffusion of ideas, we could not state that there was an allembracing circumpolar "culture" in a narrower sense. It would be premature to start talking about a "wedge-shaped core/handle core culture"; of course we cannot, by picking out a single material trait, draw any exhaustive conclusions about cultural affinity. All of this may seem obvious, but I do think it is important to keep in mind.
It is rarely an easy task to explain the occurrence of such a widespread everyday technological feature as a blade technique, even though in this particular case it is comparatively specialized and part of a technological kit including wedge-shaped cores/ handle cores, bone points with inserts, and frequently keeled scrapers/boat shaped tools (see Morian 1967) and burins. The wide distribution of these artifacts tells us only that microblade production from oblong, wedge-shaped cores must have been a successful technology that was utilized by many cultural groups in many parts ofEurasia, North America and Greenland during an immense time span (from the Upper Paleolithic to Christian times, although not all the time in all these areas). However, it is tempting to see this technology, common to a vast area throughout the circumpolar area, as an expression of a common way of thinking, but exactly what implications and connotations it had are still largely unknown to us.
Whatever the "reasons" for this widespread wedge-shaped core/handle core technique, whether it be migration/diffusion of ideas or some adaptation to similar living conditions leading to independent core areas, or something else, I strongly suspect that the final word in this discussion can not be expected for quite a while.
English revised by Laura IIrang.