23

Baltic Ware From The Fribrødre Å Site 2010

Embed Size (px)

Citation preview

Part IV

BALTIC WARE FROM THE FRIBRØDRE Å SITE

Lars Krants Larsen

469

BALTIC WARE FROM THE FRIBRØDRE Å SITE

In connection with the many excavations along the now less than impressive Fribrødre River, a mixed ceramic assemblage was recovered, embracing prac-tically all prehistoric periods; in the present paper, only Baltic Ware is discussed. The Baltic material is in all its diversity fascinatingly confusing, and even assemblages from closely situated locations tend to differ more than one would expect from other peri-ods of Danish prehistory. The apocryphal text, Book of Wisdom 15:7, reads: For a potter kneading his clay laboriously moulds every vessel for our use, but out of the self-same clay he fashions without distinction the pots that are to serve for honourable uses and the opposite; and what the purpose of each one is to be, the moulder of the clay decides – and, one might add, he also decides how each vessel is to be shaped, how the rim is moulded, and how the body is decorated. This, and much more, is the topic of the present paper.

What is Baltic Ware?To allow any serious discussion of Baltic Ware, it is necessary to define the concept as it is used in this text. Over time, the term Baltic Ware has had several names, most of which have been less than satis factory; the most common names are Wendish Ware, Baltic Ware and Slav Ware and, frequently, the concepts appear to be used somewhat at random. Below, the term Baltic Ware is used to describe a specific ceramic group, which is known from a large geographic area. It covers the region between the Jutland Ridge and Bornholm, including Funen, Zealand, Lolland and Falster. The colours of the group vary from brown through grey, to sometimes almost black, with a flamed pattern, and the pots are hand-made but wheel-finished. Most vessels have flat bases, and they may be decorated with furrows, wavy lines and other forms of decoration. The vessel shapes of this ceramic style are characterized by a

high degree of local variation, and the style covers a considerable period from the end of the Viking Age and well into the 1200s. Indeed, this broad definition covers a ceramic material so diverse that the concept may appear almost useless. However, this definition of Baltic Ware is the generally accepted one, and it is universally understood by specialists (Roslund 2001, 54). The expression Slav Ware exclusively refers to pottery manufactured in the Slav areas, i.e. mainly in the coastal region of present-day Germany and Poland. When encountered in modern Denmark, this material represents imports. The definition of Baltic Ware clearly has its problems, such as severe difficul-ties in terms of distinguishing between Danish prod-ucts and products manufactured south of the Baltic Sea. Frequently, it is only through scientific analysis that this distinction can be made.

Overview of the available literature

It is not the author’s intention to present a final and all-embracing overview of the published literature on the rich and highly varied Baltic Ware. Instead, this section constitutes a brief chronological presentation of the most significant books and papers relating to this specialist field, with particular emphasis on the publications which are relevant in a Danish context, especially those which are relevant to the analysis of the finds from the Fribrødre Å site. In case a more comprehensive overview of Danish and Swedish Baltic Ware is needed, the readers are referred to Mats Roslund (2001, 47-55). His monograph also includes a comprehensive summary of the Slav, Baltic and Rus-sian Wares (Roslund 2001, 87-136).

One of the most significant publications within this field of research is Dagmar Selling’s (1955) mono graph on the Viking Age and early medieval

470

pottery in Sweden. The two most important ele-ments of Selling’s typology are ceramic technology – that is, firing temperatures and the atmosphere during the burning – and vessel form. The typo-logical subdivision of the vessel forms was made possible by the fact that a large part of her material constitutes vessels from graves, which tend to be more intact than most domestic pottery. The various forms of the vessels are subdivided according to rim shape, with most rims being either inverted, everted or vertical, although other variants are also known (they are not dealt with in this paper); in her pub-lication, Baltic Ware is referred to as AII. Selling’s typology has its weaknesses, which are excellently explained by Roslund (2001, 49-50). This, however, has not prevented practically every specialist work-ing on Danish Baltic Ware from using major parts of her typology, in particular the subdivision of vessel form according to rim shape. This principle was also followed in the present paper.

Ewald Schuldt’s (1956) analysis of the pottery from Mecklenburg-Vorpommern is another foreign pub-lication, which has been used extensively in Danish research. It was published approximately at the same time as Selling’s work. On the basis of an unknown amount of pottery from a large number of excava-tions and assemblages, he subdivides the pottery into ten typological groups, which are named according to where they occur most commonly. These exotic names (Woldegker, Menkendorf, Teterow, Vipperow, etc.) soon found their way into the Danish archaeo-logical literature, and for the first time it is possible to distinguish between some of the different imported Middle and Late Slav types, which constantly occur in our excavations. Like Selling’s work, this publica-tion has been subjected to constant corrections, and particularly typological and chronological issues have been refined and stated more exactly (for exam-ple, Kempke 1981; 1984; 2001; Brather 1996; Schmid-Hecklau 2002). Despite these shortcomings, the pub-lication is still regarded as one of the most important works of reference within its field.

A number of publications have influenced the present paper more directly. The more important of these is Niels-Knud Liebgott’s (1979; 1982) two papers on the pottery from the excavations at Peders-borg near Sorø, and from Jernløsegård near Holbæk. The finds from Pedersborg include approximately 700 rim sherds from c. 440 vessels, dated quite nar-rowly to the period c. AD 1150-1200. This material is characterized according to Selling’s system, where the vessels are subdivided typologically in relation to whether they have inward bent, everted or ver-tical rims, followed by further subdivision of these

main types. This paper represents the first more extensive analysis of Baltic Ware, and all later works within the field refer to it. The paper on the finds from Jernløsegård represents a continuation of the Pedersborg paper, where the material was charac-terized along the same general principles.

Wilhelm Geber’s (1980) analysis of Baltic Ware from the Danish islands is another important paper. The finds were recovered from six selected locations, namely Lilleborg on Bornholm, Jernløsegård, St Ib’s Church in Roskilde, Vejleby and Pallerup in Vejleby parish on Lolland, as well as Odense. The main aims of the paper were to investigate the Slav influence on Baltic Ware and to define a chronology for this research field, as well as, subsequently, to compare the Danish and Slav finds. As the paper constitutes a preliminary presentation of Geber’s findings, it is not possible to deal with the topic in any greater detail. The final paper on this research has not yet been published, which is regretable as it represents the only attempt at analysing a large volume of finds from many different locations.

It is equally important to mention the investi-gation of pottery from Stevns, which is also situ-ated on Zealand. The sherds were recovered from a number of small, randomly selected (ie, sampled) grids, excavated in a number of selected villages (Hedeager et al. 1982). In total, 1,320 sherds were recovered, 160 of which are rims. This assemblage is dated to the period from c. AD 1000 to some time after c. AD 1200. All known types are present through this period, and in general it was not pos-sible to distinguish chronologically between those types. Only the assemblage from Store Heddinge allowed the definition of an earlier horizon, dating approximately to the 11th century AD.

The most thorough investigation of Baltic Ware is the analysis of the finds from a number of soli-tary farmsteads on Falster, the so-called Falster ‘Old Towns’. This investigation is particularly interesting as the farmsteads are located close to the Fribrødre Å site (Pedersen 1980; 1989). In these papers, Baltic Ware and Slav Ware are jointly referred to as Slav Ware, and the material is divided into two types, namely vessels with inverted rims and vessels with everted rims. These vessels are then subdivided into four different ornament classes, according to their most common form of decoration.

It is also necessary to mention a number of minor notes on the topic. The earliest paper is the one pre-senting the investigations in and around St Ib’s Church in Roskilde (Olsen 1962). This assemblage includes approximately 1,000 sherds from at least 90 vessels, predominantly dating to the 11th century AD.

471

Another paper describes the excavation at Vejleby on Lolland (Boyhus & Snedker 1971). This assemblage includes 5,800 sherds of Baltic Ware, dating to 1050-AD 1200. The Vejleby paper is a preliminary presenta-tion, and no final paper exists. This is unfortunate, as the excavation and its results appear to be exception-ally interesting. An assemblage of, inter alia, 451 rims is presented in a paper on the finds from an excava-tion of the Brovold defensive structure on Als (Koch 1978), but this assemblage differs notably from the excavated assemblages mentioned above. In addition, a small assemblage of at least 25 vessels was recov-ered from Borrebjerg on Langeland (Skaarup 1982).

Characterization of the pottery from Fribrødre ÅDuring the excavations along the Fribrødre River, 1,153 pot sherds were recovered, many of which were conjoinable. However, the finds are fragmented and chronologically mixed, and they include sherds from most of prehistory, as well as from medieval times. Although most periods of prehistory are represented, many finds date to the Neolithic and Iron Age periods. This section only deals with those sherds assumed to be contemporary with ship-related activities, i.e. Baltic Ware. The vast majority of the finds were recovered from the river, and because of this it was not possible to establish a stratigraphical sequence either for the prehistoric pottery or for the Baltic Ware. It was also not possible to define differences between the hori-zontal distribution patterns of the various ceramic types. Consequently, it is not possible to separate the different ceramic types either with reference to the stratigraphy or to the horizontal distribution of sherds, and the analysis had to be carried out entirely by visual inspection of the material.

As in practically all papers on Baltic Ware, the finds are divided into three types, with reference to whether the rims are either inverted (Type T-I which in this paper is sub-divided into six sub-groups), everted (Type T-II sub-divided into five sub-groups) or vertical (Type T-III sub-divided into three sub-groups).

Rim typesA total of 545 sherds were defined as belonging to the Baltic ceramic tradition; 122 of those are rim sherds and 50 are base sherds, representing at least 66 vessels. Several rims are thought to derive from the same vessel, but as those pieces did not join, it was not possible to prove the assumption. If all rim

sherds, thought to be from the same vessel as other sherds, actually represent individual vessels, the minimum total number of vessels rises to 75. No base sherds are included in this calculation, and it is therefore assumed that all base sherds are asso-ciated with at least one rim sherd. Below, only the lowest minimum number of vessels is referred to.

In terms of rim sherd typology, 60 sherds (49%) are from vessels with inverted rims, 59 (48%) are from vessels with everted rims, and three (2%) are from vessels with vertical rims. If this quantifica-tion is instead based on whole vessels, the above proportions are transformed into a slight over-rep-resentation of vessels with everted rims, as 31 ves-sels (47%) have inverted rims, 32 (48%) have everted rims, and three (5%) have vertical rims.

It is interesting to compare the relative propor-tions of rim types from other excavations (Table 1). In general, two-thirds of the rim sherds seem to derive from vessels with inverted rims, appar-ently unaffected by chronological or geographical factors. However, a small number of notable excep-tions are known: the assemblage from Pallerup includes approximately equal numbers of vessels with inverted, everted, and vertical rims, whereas the assemblage from Brovold is composed almost exclusively of everted rim sherds. There seems to be a weak trend towards higher numbers of ves-sels with inverted rims in the earliest assemblages, such as those from Store Heddinge on Stevns and St Ib’s Church in Roskilde, just as the oldest horizon in the Falster farmsteads is composed exclusively of inverted rims. This, however, is not an unequivo-cal trend. The assemblage with a vessel type com-position most comparable to that of the Fribrødre Å material is that of Borrebjerg on Langeland. The most likely conclusion is that the relative propor-tions of the rim types is not a diagnostic feature.

Vessel formsThe fragmentation of the vessels makes it diffi-cult to determine the original vessel shape, but an attempt is made below. It appears that the vast majority of the 31 vessels with inverted rims are vessels with gently curved bodies, where the broad-est part of the vessels is located approximately one-third below the rim. Selling refers to this type as AII:3a1 or AII:3a2. A small number of vessels with a relatively sharp angle (carination) at the point of greatest breadth may belong to Selling’s Type AII:2a (Fig. 2:g). In the Borrebjerg assemblage, this type represents an exception (Skaarup 2005). A soli-tary vessel is shaped almost like an urn, whereas

472

Table 1. Relative distribution (in percent) of vessels with inverted, everted and vertical rims. Left columns: the relative distribution calculated on the basis of rim numbers. Right columns: the relative distribution calculated on the basis of whole vessels.

Note 1: This assemblage was dated on the basis of Olsen (1962). Note that Gebers (1980) calculates the proportions on the basis of the entire assemblage, including lids, lamps, etc. For the sake of comparison, the present table offers re-calculated figures, based on rims only. In the table, the first figures are Geber’s original sums, whereas the figures in bold (after the slash) are the re-calculated sums.

Note 2: Numbering only 25 vessels, this assemblage is numerically small; only 19 of those vessels are typologically definable.

Note 3: Due to their small numbers of sherds, the two remaining ‘Old Towns’, Ravnsø and Egelev, have been excluded.

Table 2. Rim diameters and base diameters are shown in relation to the total numbers of rims and bases.

Site Source DateBased on total no of rims Based on total no of vessels

Inverted rims

Everted rims

Vertical rims

Inverted rims

Everted rims

Vertical rims

Fribrødre Å AD 1050-1100 49 48 2 47 48 5St. Heddinge Hedeager, Poulsen &

Tornbjerg 1982Horizon 1 (AD 1000-1100)

85 13 3

Stevns, total Hedeager, Poulsen & Tornbjerg 1982

AD 1000-1200 71 15 14

St Ib’s Church, Roskilde 1 Gebers 1980; Olsen 1962 AD 1000-1100 61/79 11/14 5/6Borrebjerg 2 Skaarup 2005 AD 1100-1150 42 47 11Vejleby Gebers 1980;

Boyhus & Snedker 1971AD 1000-1200 65/65 13/13 23/23

Odense Gebers 1980 AD 1000-1200 80/87 5/5 7/8Pallerup Gebers 1980 AD 1000-1200 31/31 25/25 43/43Lilleborg Gebers 1980 AD 1000-1200 42/65 14/22 9/14Jernløsegård Gebers 1980 AD 1000-1200 45/60 23/31 7/9Jernløsegård Liebgott 1982 AD 1150-1250 68 23 9Pedersborg Liebgott 1979 AD 1150-1200 75 20 5 77 18 5Brovold Kock 1978 AD 1150-1200 0 92 8Kippinge ‘Old Town’ 3 Pedersen 1989 AD 1150-1300 60 40 0Nørre Vedby ‘Old Town’ Pedersen 1989 AD 1150-1300 70 30 0

Diam. (cm) 6 7 8 9 10 11 12 14 15 16 17 18 19 20 22 Total %T-I:1 T-I:1a 4 4 3,8

T-I:1b 6 1 1 1 9 8,7T-I:1c 2 2 4 3,8

T-I:2 T-I:2a 1 1 8 3 13 12,5T-I:2b 0,0T-I:2c 0,0

T-I:3 T-I:3a 2 4 6 12 11,5T-I:3b 3 1 4 3,8

T-I:4 1 1 1,0T-I:5 1 1 1,0T-I:6 1 1 1,0T-II:1 T-II:1a 6 1 3 10 9,6

T-II:1b 4 7 3 14 13,5T-II:2 T-II:2a 7 3 10 9,6

T-II:2b 1 2 2 5 4,8T-II:2c 5 5 4,8

T-II:3 T-II:3a 1 1 2 1,9T-II:3b 1 2 1 4 3,8

T-II:4 0,0T-II:5 2 2 1,9T-III:1 1 1 2 1,9T-III:2 0,0B1 2 7 7 3 1 2 22 21,2B2 6 2 2 1 11 10,6B3 2 2 1,9Total rims 8,0 1,0 10,0 17,0 4,0 31,0 7,0 15,0 3,0 4,0 3,0 103Total bases 2 8 7 7 5 3 3 35% rims 7,8 1,0 9,7 16,5 3,9 30,1 6,8 14,6 2,9 3,9 2,9 100% bases 5,7 22,9 20,0 20,0 14,3 8,6 8,6 100

the other more characteristic vessels of this type all have gently curved bodies (Fig. 4:b).

Most of the 32 vessels with everted rims also belong to the form with gently curved bodies, Selling’s Type

AII:3c. A small number of the vessels differ from this morphology by having almost vertical bodies, and they generally seem more coarse; they are referred to Selling’s Type AIV:3b (Figs. 6:e and 8:a).

473

The final small group of vessels, represented by only three examples, all have vertical rims. It is diffi-cult to determine the orientation of the group’s rims, as the sherds are few in number and they are highly fragmented. It is quite possible that, if greater pro-portions of the vessels had survived, the rims would have been characterized as either inverted or everted. To the extent that the group is a true formal group at all – which two of the vessels clearly suggest (Fig. 10:c) – it represents Selling’s Type AII:3c.

All the analysed vessels are hand-made, and most are wheel-finished. In connection with this process, the vessels’ exterior was burnished, the rims were wheel-finished, and the vessels were decorated with furrows, waves and spiralling furrows. In terms of these attributes, this assemblage corresponds well to other assemblages of Baltic Ware.

The colour of the vessels varies considerably, even on individual vessels. The colours embrace a spectrum from light caramel brown, through grey to dark brown and, in this sense, they do not differ from pottery recovered elsewhere. During the analysis, it was noticed that the sherds are tempered by unusually coarse inclusions. The total assemblage includes approximately 52% of sherds with fine to medium-coarse temper, whereas 48% of the sherds have medium-coarse to coarse temper. If the character of the temper is subdivided by vessel form, it is noticed that 79%

of the vessels with inward bent rims have fine to medium-coarse temper, whereas only 48% of the vessels with everted rims have (Table 4).

Vessel size and rim diameterGenerally, it is difficult to determine the size of the vessels, as only one vessel has survived intact suffi-ciently for its height to be estimated. This vessel has a height of 16 cm (Fig. 2:c and Fig. 1). In all other cases, it was only possible to measure the rim diam-eter. Only 85% of the rim sherds were large enough to allow the rim diameter to be estimated. It is generally accepted that measurements can only be taken if at least 4-5% of the rim diameter has sur-vived (Tables 2 and 3). The assemblage includes ves-sels with a rim diameter between 10 cm and 22 cm, but the vast majority of the vessels – more than 70% – have rim diameters between 12 cm and 18 cm. The most common vessels are those with a rim dia meter of 16 cm (31%). The picture changes somewhat if

Table 3. Rim diameters are shown in relation to the total numbers of whole vessels.

Diam. (cm) 6 7 8 9 10 11 12 14 15 16 17 18 19 20 22 Total %T-I:1 T-I:1a 2 2 1,9

T-I:1b 1 1 1 1 4 3,8T-I:1c 1 1 2 1,9

T-I:2 T-I:2a 1 1 1 1 4 3,8T-I:2b 0,0T-I:2c 0,0

T-I:3 T-I:3a 1 2 1 4 3,8T-I:3b 1 1 2 1,9

T-I:4 1 1 1,0T-I:5 1 1 1,0T-I:6 1 1 1,0T-II:1 T-II:1a 3 1 1 5 4,8

T-II:1b 1 2 1 4 3,8T-II:2 T-II:2a 2 1 3 2,9

T-II:2b 1 2 1 4 3,8T-II:2c 2 2 1,9

T-II:3 T-II:3a 1 1 2 1,9T-II:3b 1 2 1 4 3,8

T-II:4 0,0T-II:5 1 1 1,0T-III:1 1 1 2 1,9T-III:2 0,0Total rims 2 1 7 11 1 13 2 5 1 2 3 48% rims 4,2 2,1 14,6 22,9 2,1 27,1 4,2 10,4 2,1 4,2 6,3 100T-I 2 1 4 3 6 3 2 21% 9,5 4,8 19,0 14,3 28,6 14,3 9,5 100T-II 3 7 1 7 2 2 1 1 1 25% 12,0 28,0 4,0 28,0 8,0 8,0 4,0 4,0 4,0 100T-III 1 1 2% 50 50 100

Table 4. Forms of temper shown in relation to the num-bers of vessels with inverted, everted and vertical rims.

Type of temperVessel type

T-I T-II T-III

Fine to medium-coarse temper 79% 48% 60%

Medium-coarse to coarse temper 21% 52% 40%

474

the size distribution of whole vessels is considered in relation to the different vessel forms within the total assemblage. Amongst the most common forms, vessels with inverted rims include 19% with a rim diameter of 12 cm, and 29% with a diameter of 16 cm. Vessels with everted rims are generally char-acterized by greater size variation, with the most common vessels being those with a rim diameter of 14 cm and 16 cm (both 28%), followed by vessels with a diameter of 12 cm (12%). Amongst the vessels with vertical rims, only two were of sufficent size to allow measurements to be taken, one with a rim diameter of 14 cm, and the other with a diameter of 20 cm. Only 3% of the entire assemblage has rims with diameters exceeding 20 cm.

Not all analyses of Baltic Ware include measure-ments of the rim diameter, but the published stud-ies suggest that in general the vessels from the Fri-brødre Å site are smaller than at least those from Zealand. At Pedersborg, 29% of the vessels with inverted rims, and 27% of the vessels with everted rims, are characterized by rim diameters exceed-ing 20 cm. More than two-thirds of the vessels with inverted rims from the Jernløsegård assemblage have rim diameters exceeding 20 cm, in some cases even 30-32 cm. The assemblage from St Ib’s Church in Roskilde is characterized by the same trend, with the main publication describing most of the sherds as representing vessels with diameters of 25 cm. In addition, the assemblages from Pedersborg and Jern-løsegård include a number of very large bowls with rim diameters approaching 40 cm. The latter vessel type is absent from the Fribrødre Å assemblage. In terms of the distribution of vessel forms, as well as rim diameters, the assemblage from Borrebjerg is (as mentioned above) the one most closely resembling that of the Fribrødre Å site, and together these assem-blages differ from other assemblages. Unfortunately, rim diameters are only known in a number of cases, but examination of the illustrations suggests that the vessels may have had an average rim diameter of approximately 15 cm. In comparison, the vessels from the Fribrødre Å site have an average diameter of 15.5 cm. However, the assemblage from Borrebjerg also includes vessels with a diameter of almost 27 cm. As mentioned above, such vessels are absent in the assemblage from the Fribrødre Å site.

The various vessel sizes may, to some extent, repre-sent functional differences. The exceptionally large bowls from Jernløsegård must have had a different function to that of the many smaller vessels; large vessels may have been storage vessels, and some are evidently cooking vessels. Also, it cannot be ruled out that some of the small vessels may represent

table wares. Slightly less than 5% of the sherds from the Fribrødre Å site are characterized by interior surfaces with burnt food remains, indicating that those vessels were used for cooking. This pheno-menon is known from other assemblages, such as those from Pedersborg and Jernløsegård. The pot-tery from Pedersborg include ‘vessels with shiny black areas of soot on their exterior surfaces, and burnt food remains on their interior surfaces’ (Lieb-gott 1979, 143), and ‘a proportion of the vessels from Jernløsegård have a thick organic coating on their interior surfaces, which may be partially carbon-ized food remains’ (Liebgott 1982, 144). As would be expected, the vast majority of the sherds with burnt food remains are body sherds. Three sherds were large enough to allow determination of their vessel form, and in all three cases it was possible to define the original vessels as having everted rims. However, no wide-reaching conclusions should be sought on the basis of this small group of sherds.1 These vessels have rim diameters of 16-18 cm, and therefore belong to the site’s larger vessels.

The collected ceramic series from Pedersborg, Jern-løsegård, St Ib’s Church in Roskilde, and the farm-steads on Falster may well be domestic pottery in the wider sense of this concept, as they represent differ-ent forms of functional and household pottery: cook-ing vessels, storage vessels, bowls, and table wares. It should also be borne in mind that medieval excava-tions frequently produce rarer ceramic forms, such as lids, lugs (handles), lamps, etc., which are absent from the Fribrødre Å assemblage. Only a small number of wooden lid fragments were recovered, and it cannot be dismissed that they may have been used in con-nection with the ceramic vessels.2 The question is, whether this defines the Fribrødre Å pottery as a spe-cial case? Is it possible that the pottery from the Fri-brødre Å site predominantly represent table wares, in which the weary ships’ carpenters had their food served, and from which they then ate? The character of the evidence from comparable studies does not allow this question to be answered, but if the assumption is correct, this might explain the assemblage’s compara-tively small ceramic vessels, as well as the absence of rarer ceramic forms, such as lids, etc. However, it should be noted that lids are also rare in the assem-blages from the farmsteads of Falster, suggesting that the absence of lids may be typical of this island in general. Unfortunately, the paper on the farmsteads of Falster does not include information on the rim diameters of vessels. However, this author measured the rims on the publication’s illustrations and due to the way this information was obtained, caution in its use is recommended. The illustrations, where meas-

475

urements of the rim diameter were made, indicate that 30% of the vessels have diameters between 10 cm and 14 cm (against 44% at the Fribrødre Å site); 50% have diameters between 15 cm and 18 cm (against 44% at the Fribrødre Å site); and 20% have diameters exceeding 18 cm (against 13% at the Fribrødre Å site). It appears that the ceramic vessels from the Falster farmsteads are generally smaller than those from Zealand but, nonetheless, larger than those from the Fribrødre Å site.

BasesA total of 50 intact or partial bases were retrieved from the site; apart from one of these, it is not pos-sible to refer any of the bases to the three main vessel types. Only the above mentioned vessel is so intact that its base as well as its rim survive. It has an inverted rim and a flat base (Fig. 2:c and Fig. 1), the most common type of base (Type B1). No less than 70% of the base sherds are flat (Fig. 11:a-b). The second most common base type (26%) is one with a small foot-ring (Type B2). This phenomenon has been explained as the product of using too small a wheel, when the vessel was wheel-finished, causing the development of a drooping rim along the edge of the base (Liebgott 1982, 142-43)(Fig. 11:c). Two examples (4%) have lens-shaped bases (Type B3), and only three examples have a base impression in the shape of a cross in high relief (Fig. 11:d). Base impressions have been discussed frequently in the

archaeological literature, and they are a common feature of Baltic Ware. A vast number of varieties are known, but their meaning is still unknown (Roesdahl 1973-74). The diameter of the bases vary between 6 cm and 14 cm, but most of the base sherds (77%) are from vessels with diameters of 7-10 cm, most commonly 7 cm (23%).

DecorationSlightly more than half of the sherds, 284 pieces (52%), are decorated. There is a tendency for the vessels to have less decoration on the lower parts of their bodies; the most extensively decorated area is the broadest part of the vessel, although many vessels display wheel-turned furrows or runnels from rim to base (Fig. 11:b). Gebers has studied the parts of the vessel where the decoration is located (Gebers 1980, 143). He divides the vessels into five zones, which, to the present author, appears to be a somewhat risky approach when one is dealing with a fragmented assemblage. Basically, it is too difficult to define the original position of a given body sherd, and it was decided not to apply Gebers’ method in connection with the analysis of the finds from the Fribrødre Å site.

The rims, and the minimum number of vessels suggested by them, indicate that 24 of 31 vessels (77%) with inverted rims are decorated, 24 of 32 ves-sels (75%) with everted rims are decorated, and all three vessels with vertical rims (100%) are decorated.

Fig. 1. The two best preserved vessels: from left 85x78 (Fig. 2-c) and 85x77 (Fig. 11-b).

476

In total, 77% of the vessels from the site are decorated. However, it should be borne in mind that a proportion of the sherds used for calculating the site’s minimum number of vessels are relatively small. Approximately 15% of the decorated rim sherds are characterized as small, whereas as much as 50% of the undecorated rim sherds are characterized as small. This suggests a correlation between rim sherd size and the frequency of decoration: the smaller the sherd, the higher the likelihood of the piece being undecorated. Or, in other words, it is quite likely that small undecorated rim sherds originally formed parts of larger sherds with decoration further down the vessel wall, but that this decoration has not survived. This indicates that the original number of decorated vessels may have been somewhat higher than 77%. Most likely, almost all the original vessels were decorated, and a notable proportion of the 48% undecorated sherds are from the lower parts of decorated vessels.

If these results are compared with other studies, differences between the various assemblages are noted. The assemblages from the farmsteads of Fal-ster include more than 95% sherds decorated with at least one furrow, and the undecorated sherds are so small that they could all derive from decorated ves-sels (Pedersen 1989). No other assemblages are char-acterized by similarly high numbers of decorated sherds. The papers on the Pedersborg and Farvergade

assem blages only mention the decorated rim sherds: the Pedersborg assemblage includes 82% decorated rim sherds, whereas the finds from Farvergade only include approximately 59% decorated rim sherds, with a trend towards higher numbers of decorated vessels in later contexts (Petersen 1988). Gebers has also attempted to estimate the proportions of deco-rated sherds in his material. Unfortunately, he does not explain clearly how these figures were produced, i.e. whether he only used the rims, or whether all sherds were included. His analyses suggest that the six locations have between 54% (Jernløsegård) and 77% (Lilleborg) decorated sherds, whereas all other sites have between 65% and 72% decorated sherds (Gebers 1980). The high fragmentation ratio makes it difficult to define how the decorative elements were combined on the individual vessels.

On the whole, the most common form of decora-tion is one or more horizontal furrows; this form of decoration occurs on 93% of the decorated sherds, and no less than 73% of the sherds are decorated by furrows only (Table 5). A small proportion (c. 3%) of these were characterized as precisely spaced fur-rows, with all these sherds deriving from vessels with everted rims (Fig. 8:e and f as well as Fig. 9:a, b and d), whereas the remainder of the furrows were defined as loosely spaced furrows, forming concen-tric or spiral patterns. It is thought that most of the

Table 5. The relative distribution of decorative elements across all decorated sherds are shown, as well as in relation to the numbers of vessels with inverted (T-I), everted (T-II) and vertical (T-III) rims.

Decorated sherds

Percentage of all decorated sherds

T-I Percentage of whole vessels

T-II Percentage of whole vessels

T-III Percentage of whole vessels

Precisely spaced furrows 8 2.8 0 0.0 4 12.5 0 0.0

Furrows 199 70.1 4 13.8 10 31.3 0 0.0

Precisely spaced furrows + furrows 207 72.9 4 13.8 14 43.8 0 0.0

One or more wavy lines 7 2.5 0 0.0 1 3.1 1 33.3

Horizontal band of oblique incisions 8 2.8 2 6.9 3 9.4 1 33.3

Cordon with oblique incisions 1 0.4 0 0.0 0 0.0 0 0.0

Finger impressions 1 0.4 0 0.0 0 0.0 0 0.0

Twig-stab decoration 1 0.4 1 3.4 0 0.0 0 0.0

Precisely spaced furrows + wavy lines

1 0.4 0 0.0 0 0.0 0 0.0

Furrows + wavy lines 11 3.9 2 6.9 1 3.1 0 0.0

Precisely spaced furrows + furrows + wavy lines

12 4.2 2 6.9 1 3.1 0 0.0

Furrows + horizontal band of oblique incisions

15 5.3 1 3.4 3 9.4 1 20.0

Furrows + cordon with oblique incisions

20 7.0 9 31.0 0 0.0 0 0.0

Furrows + finger impressions 2 0.7 2 6.9 0 0.0 0 0.0

Furrows + twig-stab decoration 6 2.1 2 6.9 1 3.1 0 0.0

Wavy lines + cordon with oblique incisions

1 0.4 1 3.4 0 0.0 0 0.0

Furrows + wavy line + twig-stab decoration

3 1.1 0 0.0 1 3.1 0 0.0

Two or three decorative elements 59 20.8 17 54.8 6 18.8 1 33.3

Decorated whole vessels 24 77.4 24 75.0 3 100.0

Fine furrows + other forms of decoration

265 93.3 20 64.5 19 59.4 1 33.3

477

vessels have more than one horizontal furrow, but in many cases the sherds are so small that only a single furrow survives. When the present assemblage is compared to other assemblages of Baltic Ware, many appear similar, but much more varied assemblages are also known. The oldest horizon at Farvergade in Næstved, for example, includes 96% sherds with furrow decoration, whereas the finds from Peders-borg only include 29% (Petersen 1988, 198-201).3

Other forms of decoration rarely occur in isolation; the most common of these decorative motifs are hor-izontal bands of oblique incisions and wavy lines, which occur on 3% of the decorated sherds. These motifs are only experienced separately in one case each, and only one of these sherds is large enough to ascertain that it actually is a single isolated motif (Fig. 2:a). In most cases, the rarer forms of decoration are combined with furrows. Most commonly, fur-rows are combined either with an obliquely incised cordon (7%), with a band of oblique incisions (5%), or with wavy lines (4%). Overall, nearly 21% of all sherds have decoration in the form of furrows com-bined with other motifs. In this respect, the assem-blage compares well with the later contexts from Farvergade (30%), and Borre bjerg on Langeland (17%), whereas Pedersborg, with 48% of its sherds carrying one or more forms of decoration, deviates from this trend (Petersen 1988).

If the decoration is considered in relation to whole vessels, certain trends become clearer. Amongst the vessels with only one form of decoration, 14% of the vessels with inverted rims are decorated with fur-rows only, whereas 44% of the vessels with everted rims are decorated this way. Amongst the vessels with inverted rims, 55% of those have more than one form of decoration, most commonly in the form of an obliquely incised cordon and furrows (31%), whereas this combination of decorative elements is completely absent on vessels with everted rims. Only 19% of the vessels with everted rims carry more than one form of decoration and, in general, the decoration of ves-sels with inverted rims is considerably more varied than that of vessels with everted rims.

Characterization of the assemblageAs other analysts have realized, Baltic Ware is a difficult research topic, characterized by fruitless analyses of vast, apparently contemporary assem-blages, which in the most infuriating manner refuse to cooperate at all. In terms of form and decoration, Baltic Ware is highly varied, and it is possible to

subdivide the material into a large number of sub-types. These are restricted to relatively narrow geo-graphical areas, without regional or inter-regional types, unless the analysis is restricted to the most general features of the vessels. Consequently, this paper does not attempt to compare the typology and dating of the Fribrødre Å vessels with vessels from other assemblages, with the exception of a small number of particularly relevant cases. The numbers of the various vessel types are presented in Table 6.

Type I: Vessels with inverted rimsThis group of vessels includes 60 rim sherds, dis-tributed across at least 31 vessels. The rims are sub-divided into six sub-groups, some of which are then subdivided into variants.

The first group, T-I:1, is the largest group of vessels with inverted rims, and it includes 20 rims, each with an either flat, rounded or pointed lip. These sherds derive from at least 11 vessels. The group is subdivided into three variants, namely T-I:1a, T-I:1b and T-I:1c.

Table 6. Rim types are shown in relation to rim numbers and minimum numbers of whole vessels.

Vessel type Sub-type Rims Whole vessels

T-I:1 20 11

T-I:1a 5 3

T-I:1b 11 6

T-I:1c 4 2

T-I:2 20 11

T-I:2a 16 7

T-I:2b 1 1

T-I:2c 3 3

T-I:3 17 6

T-I:3a 13 4

T-I:3b 4 2

T-I:4 1 1

T-I:5 1 1

T-I:6 1 1

T-II:1 27 13

T-II:1a 10 5

T-II:1b 17 8

T-II:2 21 9

T-II:2a 10 3

T-II:2b 6 4

T-II:2c 5 2

T-II:3 8 8

T-II:3a 3 3

T-II:3b 5 5

T-II:4 1 1

T-II:5 2 1

T-III:1 2 2

T-III:2 1 1

Total 122 66

478

Type T-I:1a is characterized by rims with a flat lip, and most vessels are decorated with furrows; one solitary vessel is decorated by a horizontal band of round stab marks, possibly formed by a twig. This category embraces five rims from at least three vessels. They are all characterized by medium-coarse temper, and they have a rim diameter of 16 cm (Fig. 2:a-b).

Type T-I:1b consists of vessels with an undeco-rated, mostly clearly demarcated, rim zone, with the rims having either flat or pointed profiles. The group includes 11 sherds from at least six vessels. The vessels are more highly decorated than T-I:1a, and their decoration usually includes more than one decorative element; these elements may include a horizontal band of oblique incisions, horizon-tal twig-stab decoration, or an obliquely incised cordon, but these elements are always combined with furrows. All vessels have fine to medium-coarse temper, and they differ somewhat in size. It may be possible to subdivide this type into smaller pots with a diameter of 10-12 cm, and larger vessels with a diameter of 16, 18 and 22 cm (Fig. 2:c-f).

Type T-I:1c includes vessels with a sharp angle (carination) immediately below the rounded lip. Four

rim sherds are known, and they derive from at least two vessels. They are characterized by the same type of decoration, namely a horizontal band of oblique incisions along the broadest part of the vessel, but their sizes differ considerably; one vessel has a rim diameter of 12 cm, whereas the other is slightly larger and has a diameter of 18 cm (Fig. 2:g).

The second group, T-I:2, includes the same number of rim sherds as T-I:1, from the same number of ves-sels; these vessels all have a form of indentation on the rim’s lip. It is possible to subdivide the type into three variants, namely T-I:2a, T-I:2b and T-I:2c.

Type T-I:2a has a marked indentation on the upper face of the rim which, on occasion, appears to have functioned as lid-support. The type consists of 16 sherds from at least seven vessels, and they all have fine to medium-coarse temper. The decoration frequently includes more than one decorative ele-ment, either a wavy line or a band of finger impres-sions. The vast majority of the rim sherds are from vessels with a rim diameter of 16 cm, but in reality the individual vessels vary greatly, including rim diameters of 11 cm, 14 cm, 16 cm and 18 cm, respec-tively (Fig. 3:a-c).

Fig. 2. T-I:1a – a-b. (a: 85x552; b: 85x1400-A + 85x1420-E + 85x1420-F). T-I:1b – c-f. (c: 85x78; d: 85x795; e: 85x1598; f: 85x1414-I). T-I:1c – g. (g: 85x106 + 85x1021-B). M. 2:5.

Plate 1

a

b

c

d

e

f

g

479

Type T-I:2b represents finely tempered vessels with an indentation on the inner face of the rim, as well as on the rim’s outer face. The type should be considered rare, as it only includes one sherd. It was not possible to estimate the diameter of the rim (Fig. 3:d).

Type T-I:2c only includes three sherds from three different vessels; due to the small sizes of the sherds, it was not possible to estimate their rim diameters (Fig. 3:e-g). All vessels belonging to this group have lightly inverted rims with a horizontal indentation on top of the lip of the rim. One vessel is charac-terized by a clearly demarcated rim zone and may represent an independent type. However, to avoid confusing matters by defining a large number of sub-types, the author suggests including the vessel in Type T-I:2c (Fig. 3:g).

The very distinctive group, T-I:3, includes 17 rim sherds from at least six vessels, and it is subdivided into the two variants T-I:3a and T-I:3b.

With 13 rim sherds from at least four vessels, Type T-I:3a is the most common of the group’s types. The mouth of the vessel is heavily inverted, but the rim itself is then everted, and it has a horizontal indenta-tion, or, more likely, a lid-support, although no lids were found on the site (Fig. 4:a-c). The majority of the vessels have more or less curved bodies, although one has almost vertical walls and resembles an urn

(Fig. 4:b). In spite of these differences, the vessels form a homogeneous category and, in all cases, the decoration includes spiralling furrows. The upper part of the vessel, near the sharp bend (carination) of the vessel shoulder, is characterized by an incised cordon. All the sherds have fine temper, and they seem to derive from well-executed, delicate ves-sels. Despite their similarities in terms of temper and structure, differently sized vessels are known, with rim diameters of 10 cm, 12 cm, as well as 18 cm. The Slav material offers no direct parallel to this type, but the morphology of the vessels from, in particular, the Bobzin, Teterow and Weisdin groups follow the same trends, and the Fribrødre Å vessels are clearly influenced by this material. The above groups are all defined as belonging to a later Slav ceramic tradition, and they are dated approximately to the 11th and 12th centuries AD.4

Type T-I:3b is very similar to Type T-I:3a but, instead of a lid-support, it is characterized by a rounded lip. The type includes four rim sherds from at least two vessels, and they are all well-fired and finely tempered, and they have rim diameters of 14 cm and 16 cm. The decoration is closely related to that of Type T-I:3a, but one vessel has a wavy line instead of an incised cordon (Fig. 4:d-e).

The three remaining types must be characterized as rarer forms, as each is represented by a single

Fig. 3. T-I:2a – a-c. (a: 85x314; b: 85x1331; c: D 827/82-C). T-I:2b – d. (d: 85x1400-C). T-I:2c – e-f. (e: 85x1625; f: D 849/82-C; g: D 301/83). M. 2:5.

Plate 2

a b

c

d e f g

Plate 2

a b

c

d e f g

Plate 2

a b

c

d e f g

480

Fig. 4. T-I:3a – a-c. (a: 85x877 + 85x1164-A + 85x1164-B + 85x1215-A + 85x1374-D + 85x1374-E; b: 85x1367-A + 85x1367-B; c: 85x836-B + 85x1164-C). T-I:3b – d-e. (d: 85x1260-D + 85x1414-Y + 85x1414-Z; e: 85x1393). M. 2:5.

Fig. 5. T-I:4 – a. (a: 85x225). T-I:5 – b. (b: 85x760). T-I:6 – c. (c: 85x1219). M. 2:5.

Plate 3

a

b

c

d

e

Plate 4

a

b

c

sherd. The vessel T-I:4 is a finely tempered speci-men, with a decoration of fine furrows below an undecorated neck, and with a thickened everted lip. It has a rim diameter of 14 cm (Fig. 5:a). The vessel T-I:5 has fine to medium-coarse temper, and it has a cordon with finger impressions below an

undecorated rim and neck, and below are hori-zontal furrows. The rim is somewhat thickened and pointed, and the vessel has a rim diameter of 16 cm (Fig. 5:b). Although the sherd is unique in the present assemblage, sherds of this type seem to be common in the Vejleby assemblage (Boyhus

481

& Snedker 1971). The final Group I type, T-I:6, is represented by a single sherd from a rather large vessel with a rounded, thickened, inverted lip, and a rim diameter of at least 22 cm. The vessel may have been undecorated but, as only a small propor-tion of the rim survives, it is quite possible that it had one or more decorative elements on lower parts of its body (Fig. 5:c).

Type II: Vessels with everted rimsThis group of vessels includes 60 rim sherds, dis-tributed across at least 33 vessels. The rims are sub-divided into five sub-groups, some of which are then subdivided into variants.

Overall, Type T-II:1 is the largest group of the assem-blage, embracing 27 rim sherds from at least 13 ves-sels. These vessels are characterized by slightly everted rims, each with either a flat or pointed lip. The vessels are subdivided into two variants, namely T-II:1a and T-II:1b.

Type T-II:1a includes 10 rim sherds from at least five vessels. The vessels mainly have medium-

coarse to coarse temper, and they generally appear less well-executed than other vessels discussed in this section. Their rim diameters vary from 14 cm, through 18 cm to 20 cm, and the rims are generally less decorated than those of other vessel types. Some specimens are characterized by fur-rows combined with twig-stab decoration but, in most cases, the decoration consists of broad, shal-low, barely visible furrows. The vessels 85x1665 (Fig. 6:b) and D 260/83 (Fig. 6:d) are also char-acterized by broad furrows, although this is not shown clearly by the illustrations. Compared to other vessels, Type T-II:1a has a somewhat archaic appearance, and vessels belonging to this cat-egory may be a fraction older than other vessels (Fig. 6:a-e).

Type T-II:1b consists of 17 sherds from at least eight vessels with slightly pointed rims. The vessels form a homogeneous category, with rim diameters measuring 15 cm, 16 cm and 18 cm, whereas the decoration varies and includes furrows as well as wavy lines (Fig. 7:a-f). Type T-II:1b also includes the assemblage’s only vessel characterized by the com-bination of three decorative elements, namely fur-

Fig. 6. T-II:1a – a-e. (a: 85x1452; b: 85x1665; c: D 827/82-A; d: D 260/83; e: 85x1632-A). M. 2:5.

Plate 5

a

b

c

d

e

482

Fig. 7. T-II:1b – a-f. (a: D274/83; b: D827/82-D; c: 85 1344-A + 85x1344-B + 85x1398-E + 85x1401; d: 85x1374-C + 85x1398-B + 85x1398-C; e: 85x262-A + 85x262-B + 85x262-C; f: D845/82-A). M. 2:5.

Fig. 8. T-II:2a – a. (a: D474/83-A + D 498/83 + D 504/83-I). T-II:2b – b-d. (b: 85x1660; c: 85x1253; d: 85x1323-A + 85x1323-B). T-II:2c – e-f. (e: 85x1414-R; f: 85x1256 + 85x1378-A + 85x1378-B + 85x1378-C). M. 2:5.

Plate 6

a

b

c

d

e

f

Plate 7

a

b

c

d

e

f

483

rows, wavy lines and twig-stab decoration (Fig. 7:b). Most of the vessels have medium-coarse temper, with some being finely tempered.

Type T-II:2 is the second largest category of ves-sels with everted rims, numbering 21 rim sherds from at least nine vessels. They are characterized by a furrow on top of the rim’s lip. The type is sub-divided into three variants, namely T-II:2a, T-II:2b and T-II:2c.

Type T-II:2a consists of 10 rim sherds from at least three vessels. The rim has a horizontal inden-tation and the vessels, which represent relatively poor workmanship have medium-coarse temper. The three vessels are almost identical, with a hori-zontal band of oblique incisions, combined with furrows on the lower part of the body. They are roughly equal in size, with two rim diameters measuring 17 cm and one 19 cm (Fig. 8:a). Apart from the shape of the lip, these vessels correspond morphologically to the so-called Klemensker vessel, which is dated by coins to approximately AD 1050 (Liebgott 1978, 32).

Type T-II:2b includes six rim sherds from at least four vessels. In contrast to the above category, the indentation on top of the Type T-II:2b rims is not horizontal, but orientated obliquely away from the vessel (Fig. 8:b-d). Two of the vessels are character-ized by decoration in the shape of a horizontal band

of almost vertical incisions (Fig. 8:b). The decora-tion of one vessel is in the form of a few somewhat badly executed furrows (Fig. 8:d), whereas the last vessel to be mentioned belongs to a small group of apparently undecorated examples (Fig. 8:c). The rim diameter of these vessels varies between 12 cm, 14 cm and 16 cm.

The rim shape of the final type of this group, T-II:2c, is closely related to that of the previous type, but it has a very distinctive protruding lip below the indentations. This category includes five rim sherds from two quite well-fired vessels, both of which are characterized by narrowly spaced fine furrows. The rim diameter of both vessels is 16 cm (Fig. 8:e-f).

Type T-II:3 includes eight rim sherds from eight ves-sels. They are characterized by a short everted lip, and they may be subdivided into two variants, namely T-II:3a (the lip of which is characterized by a short, rounded angle) and T-II:3b (with a lid-support).

Type T-II:3a consists of three rims from an equal number of vessels. Only two of these allowed the rim diameter to be estimated; one has a diameter of 12 cm and one 14 cm. The decoration includes furrows, and in one case narrowly spaced furrows (Fig. 9:a).

Type T-II:3b consists of five rims from five ves-sels, which all have a well-defined lid-support. Most of the vessels are decorated with furrows,

Fig. 9. T-II:3a – a. (a: 85x73). T-II:3b – b-d. (b: D 404/83-A; c: 85x1400-B; d: D827/82-B). M. 2:5.

Plate 8

a

b

c

d

484

several of which are narrowly spaced, and one has a horizontal band of oblique incisions. In terms of size, this category is fairly heterogeneous, with rim diameters varying from 12 cm, over 16 cm, to 22 cm (Fig. 9:b-d).

The following types should be perceived as rarer forms, as each is represented by one vessel only.

Type T-II:4 has a typical double-sided lip but, unfortunately, the sherd is too small to allow the parent vessel’s rim diameter to be measured (Fig. 10:a). Type T-II:5 includes two sherds from the same vessel. This is a somewhat clumsy vessel with a heavily protruding everted rim. Despite the ves-sel’s coarse appearance, it has fine temper and a rim diameter of 14 cm (Fig. 10:b).

Plate 9

a

b

cd

Fig. 10. T-II:4 – a. (a: 85x1420-B). T-II:5 – b. (b: D555/83-A + D 555/83-B). T-III:1 – c. (c: 85x1555). T-III:2 – d. (d: 85x1097). M. 2:5.

Fig. 11. Vessels with flat bases – a-b. (a: D 504/83-C + D 504/83-E + D 504/83-F + D 504/83-G + D 504/83-H; b: 85x77). A vessel with a foot-ring – c. (c: D 633/83). Base impression – d. (d: 85x1676-A). M. 2:5.

Plate 10

a

b

c

d

485

Type III: Vessels with vertical rims

This group of vessels includes three rim sherds only, distributed across an equal number of vessels. The rims are subdivided into two sub-groups, namely T-III:1 and T-III:2.

Type T-III:1 includes two vessels with a short, verti-cal, undecorated rim with a rounded lip. The ves-sels vary considerably in size, with one having a rim diameter of 14 cm, the other 20 cm. They are both decorated, the larger of the two with wavy line decoration, whereas the smaller one has a horizontal band of oblique incisions above furrows (Fig. 10:c).

Type T-III:2 is represented by a single rim sherd with a flat termination and a discrete outer furrow immediately below the rim. The decoration of the sherd appears to be a double wavy line, although the sharp bends of the lines indicate that it may be a zigzag band. If this is the case, the piece is the only one in the present assemblage with this form of decoration. Similar forms of decoration are known from other archaeological sites in the area, such as a vessel from Hampegård on Falster, although that specimen represents a deviating rim type (Jensen 2004, 133). Due to the small size of the sherd, its rim diameter cannot be measured (Fig. 10:d).

Provenance and datingIt has frequently been said that the vessels on either side of the Baltic Sea are so similar that dif-ferentiation between Danish-made pottery and imported pottery is not possible. This, however, has not prevented the very same analysts from characterizing specific vessels as definitely repre-senting imports (Pedersen 1989, 41). In contrast, a large group of researchers – to which the present author belongs – claim that practically all Baltic Ware was produced in Denmark, from Danish clay and by local artisans, particularly from the second half of the 11th century AD onwards (Madsen 1991; Liebgott 1982). This view is supported by scientific analyses. In the western part of Zealand, for exam-ple, a ceramic assemblage was recovered which is strikingly similar to Slav material, but which is proven by clay analyses to be local. Even in Viborg, where Baltic Ware represents a rare exotic element, it has been shown to be based on local clay (Nielsen 2002; Rasmussen & Meier 2002; Ras-mussen & Hjermind 2005). The fact that the Baltic

material varies greatly over short distances sup-ports this proposition. Even the domestic pottery from neighbouring locations may vary so strongly, in terms of morphology and decoration, that it is most sensible to interpret nearly all Baltic Ware as having been produced in an individual village, or maybe even in an individual household (Roslund 2001, 80-83).

However, it is easily determined that Baltic Ware and Slav Ware share a common trend. As demon-strated by the morphology of the pottery from the Fribrødre Å site, as well as by the decoration of this material, the Fribrødre Å assemblage is clearly related to late Slav ceramic forms, in particular the so-called furrowed vessels (Gurtfurchen Ware). It is most closely related to the so-called Warder group, which in the older archaeological litera-ture is perceived as forming part of the Fresen-dorf, Bobzin, Teterow and Vipperow groups, but which Kempke (1981; 2001, 241-242; also Roslund 2001, 150-152) so convincingly merged into a new group named after a fortified location in Eastern Holstein. The core area of the Warder Ware is the Slav coast west of the Oder, and it is dated rather broadly to the 11th and 12th centuries AD.

The vital question as to how this stylistic influ-ence occurred has been discussed by analysts on numerous occasions. The question was thoroughly analysed by Roslund in his dissertation (2001) and, although this research is based on Swedish finds, its results are immediately relevant to those from Denmark. The first contacts between Slav tribes and Danes took place in the Viking Age, where middle Slav vessels were imported throughout Eastern Denmark; examples of Menkendorf, Feld-berger and Fresendorf Wares, for instance, occur as far from the source as Århus.5 However, it is not until the end of the 10th century AD, and in par-ticular the beginning of the 11th century AD, that the Slav influence on Danish ceramic traditions becomes clearly visible. According to Roslund, this first influx of Baltic Ware may be due to Slav potters having been brought to Denmark as prisoners of war. In captivity, they then continued manufactur-ing pottery according to the traditions of their native countries. Gradually, this new and technologically more sophisticated ceramic tradition expanded into new areas and, as early as the beginning of the 11th century AD, small differences between the pottery produced north and south of the Baltic Sea become noticeable; the Slav ceramic tradition had now been transformed into Baltic Ware. Other forms of influ-ence are also recognised, such as Slav groups set-

486

tling in traditionally Danish areas, trade, and pos-sibly marriage between Danes and Slav women (Ros lund 2001, 231ff).

In summary, sherds from at least 66 vessels of the Baltic ceramic tradition has been recovered at the Fribrødre Å site. Most of these vessels have gently curved bodies. The rims are distributed almost evenly across inverted and everted types. The vessels are clearly related to late Slav ceramic forms, particu-larly those of the Warder group but, most likely, none of the analysed sherds represents imported goods. Compared to the pottery from other locations, the vessels seem to have smaller rim diameters, and the assemblage is characterized by the absence of rarer types, such as lids, lugs (handles), lamps, etc., as well as large storage vessels and bowls. In relation to other published assemblages with their more varied ranges of domestic pottery, this suggests that the Fri-brødre Å assemblage may represent a special case, as it may largely be composed of table wares. It is not possible to date the pottery more precisely than to the 11th and 12th centuries AD but it is thought that the vessels are largely contemporary and related to the activities at the shipyard.

Notes1. The vessels belong to rim types T-II:1a (Fig. 6:a-e),

T-II:1b (Fig. 7:a-f) and T-II:3b (Fig. 9:b-d).

2. The lids are WV 5 and WV 14 (see Klassen this volume), with a rim dia meter of 10 cm and 18 cm, respectively.

3. At Pedersborg, the figure was calculated on the basis of ornamented rim sherds, whereas, at Farver-gade, the figure was calculated on the basis of all decorated sherds.

4. Kempke (1984, 48), where this rim type belongs to his Group G. A very Danish-looking rim is presented on Plate 23:1, although this example is decorated in an unusual manner, and Plate 19:2 shows a vessel with similar decoration, but deviating rim. A vessel, which in terms of decoration is comparable to the Danish ones, is presented in Schuldt (1956, Abb. 58a), but he also shows vessels with related rim forms, which devi-ate in terms of decoration (ibid., Abb. 48g and Abb. 57). Also see Łosiński & Rogosz (1985, 195-197): Group J, maybe in particular Sub-Groups I and III. The latter are almost exclusively dated to the 12th century AD and the beginning of the 13th century AD.

5. Menkendorf (Andersen, Crabb & Madsen 1971, 79-80); Fresendorf (Damm 2005, 30); and Feldberger from the unpublished excavation FHM 4576 St Cle-mens Stræde 10.

ReferencesAndersen, H.H., P.J. Crabb & H.J. Madsen 1971: Århus

Søndervold – en byarkæologisk undersøgelse. København.

Boyhus, E.-M. og K. Snedker 1971: En bosættelse fra den tidlige Middelalder. Foreløbig rapport om udgravningen i Vejleby 1970. Lolland-Falsters Stifts-museum Årsskrift 1971, p. 5-16.

Brather, S. 1996: Feldberger Keramik und frühe Slawen. Studien zur nordwestslawischen Keramik der Karolin-gerzeit. Universitätforschungen zur Prähistorischen Archäologie 34. Bonn.

Damm, A. (ed.) 2005: Vikingernes Aros. Gylling.

Gebers, W. 1980: Ostseekeramik auf den dänischen Inseln. In: Kiel Papers ’80 from Fuglsang/Lolland. Siedlungsforschungen auf den dänischen Inseln und im westlichen Ostseeraum (ed. H. Hinz). Kiel, p. 139-168.

Hedeager, L., B. Poulsen & S.Å. Tornbjerg 1982: Land og by – en undersøgelse af østersøkeramikkens datering og spredning på Stevns. hikuin 8, p. 125-148.

Jensen, A.-E. 2004: Venner og Fjender – dagligliv ved Østersøen 700-1200. Lübeck.

Kempke, T. 1981: Die Keramik von Warder, Kreis Sege-berg, und ihre Stellung in Ostholstein. Offa 38, p. 289-321.

Kempke, T. 1984: Starigard/Oldenburg. Hauptburg der Slawen in Wagrien II: Die Keramik des 8.-12. Jahrhun-derts. Offa-Bücher 53. Neumünster.

Kempke, T. 2001: Slawische Keramik. In: Handbuch zur mittelalerlichen Keramik in Nordeuropa I-III (ed. H. Lüdtke og K. Schietzel). Neumünster, p. 209-256.

Kock, J. 1978: Brovold. En befæstet bebyggelse fra tidlig middelalder. Kuml, p. 193-222.

Liebgott, N.-K. 1978: Danske fund af møntdateret keramik c. 950-1450. København.

Liebgott, N.-K. 1979: Keramikfundene fra voldstedet Pedersborg ved Sorø. Aarbøger for Nordisk Old-kyndighed og Historie 1977, p. 118-171.

Liebgott, N.-K. 1982. Jernløsegård. En middelalderlig gårdtomt i Sønder Jernløse, Holbæk amt. Aarbøger for Nordisk Oldkyndighed og Historie 1980, p. 126-165.

Łosiński, W. & R. Rogosz 1985: Die Periodisierung der frühmittelalterlichen Keramik aus Szczecin. Przegląd Archeologiczny 32, p. 187-207.

Madsen, H.J. 1991: Vikingetidens keramik som his-torisk kilde. In: Fra Stamme til Stat i Danmark 2: Høvdingesamfund og Kongemagt (ed. P. Mortensen og B.M. Rasmussen). Jysk Arkæologisk Selskabs Skrifter XXII:2. Århus, p. 217-231.

Nielsen, H. 2002: Keramik – og andre slaviske ind-flydelser i det sydlige Vestsjælland i vikingetid og tidlig middelalder. In: Venner og Fjender. Dansk-vend-

487

iske forbindelser i vikingetid og tidlig middelalder (ed. A-E. Jensen). Næstved, p. 81-86.

Olsen, O. 1962: Sankt Ibs kirke i Vindebode. Et bidrag til Roskildes ældste historie. Fra Københavns Amt 1962, p. 61-87.

Pedersen, J.-Aa. 1980: Ergebnisse der archäologischen Tätigkeit im Rahmen des Nordischen Wüstung-sprojektes, besonders in Kippinge. In: Kiel Papers ’80 from Fuglsang/Lolland. Siedlungsforschungen auf den dänischen Inseln und im westlichen Ostseeraum (ed. H. Hinz). Kiel, p. 169-188.

Pedersen, J.Aa. 1989: De arkæologiske undersøgelser. In: Falsterundersøgelsen 2: Dokumentationsbind (ed. S. Gissel). Odense, p. 22-169.

Petersen, J.E. 1988: Farvergade i Næstved. Arkæologiske fund fra germansk jernalder og middelalder. Aarbøger for Nordisk Oldkyndighed og Historie 1987, p. 171-209.

Rasmussen, K.L. & D. Meier 2002: Om proveniensen af Østersøkeramikken fra Næs, Ravnse og Plügge. In: Venner og Fjender. Dansk-vendiske forbindelser i vikingetid og tidlig middelalder (ed. A.-E. Jensen).Næstved, p. 91-92.

Rasmussen, K.L. & J. Hjermind 2005: Bestemmelse af proveniens og brændingstemperatur på tidligmid-delalderlig keramik, lerklining m.v. fra Viborg og Spangsbjerg. In: Viborg Søndersø 1018-1030. Arkæologi og naturvidenskab i et værkstedsområde fra vikingetiden (ed. M. Iversen, D.E. Robinson, J. Hjermind & C. Christensen). Højbjerg, p. 423-437.

Roesdahl, E. 1973-74: Bundmærker på middelalderligt lertøj i Danmark. Kuml, p. 215-229.

Roslund, M. 2001: Gäster i huset. Kulturell överföring mellan slaver och skandinaver 900 till 1300. Lund.

Schmid-Hecklau, A. 2002: Slawenzeitliche Funde im Kreis Herzogtum Lauenburg. Studien zur Siedlungs-geschichte und Archäologie der Ostseegebiete 3. Neumünster.

Schuldt, E. 1956: Die slawische Keramik in Mecklenburg. Berlin.

Selling, D. 1955: Wikingerzeitliche und frühmittelalterli-che Keramik in Schweden. Stockholm.

Skaarup, J. 2005: Øhavets middelalderlige borge og vold-steder. Ringe.