Terminological issues: mats and basketry

Conventionally in the modern world, the term textile is applied particularly to woven and knitted fabrics and a range of other techniques, such as felting and crochet. In the specialized ethnographic and archaeological literature, however, the term textile sometimes encompasses all products which consist of interconnected basic components (Seiler-Baldinger Reference Seiler-Baldinger1994), though a more restrictive definition of textile excludes items made in basketry techniques (Jørgensen et al. Reference Jørgensen, Rast-Eicher, Wendrich and Hodder2021). Basketry includes mats made in twining, objects made in coiling techniques, nets, wickerwork and other twined objects. The extensive range of ancient basketry and textile techniques are amply shown by ethnological classification systems such as the works of Irene Emery (Reference Emery2009) or Annemarie Seiler-Baldinger (Reference Seiler-Baldinger1994) and—especially for basketry—Willemina Wendrich (Reference Wendrich1991). For prehistory, this variety of techniques is exemplified in the publications of the assemblages from the Swiss lake-dwellings of the Neolithic period (Médard Reference Médard2010; Reference Médard, Gleba and Mannering2012; Rast-Eicher Reference Rast-Eicher, Hüster-Plogmann, Schibler, Jacomet, Brombacher, Gross-Klee and Rast-Eicher1997).

Woven textiles and products of basketry, matting or twining techniques rapidly undergo deterioration once interred (Kronkright Reference Kronkright, Florian, Kronkright and Norton1990, 140). There are nonetheless some cases when these objects can survive the passage of time in archaeological contexts (Grömer Reference Grömer2016; Jørgensen et al. Reference Jørgensen, Rast-Eicher and Wendrich2023; Wild Reference Wild1988). The great majority of archaeological textiles are tiny fragments, often mineralized, preserved in connection with metal artefacts such as bronze, iron or silver. Organic materials can also survive in salt mines, in ice, in arid contexts, in waterlogged contexts, in bogs, as imprints on pottery or through carbonization (Pedelì & Pulga Reference Pedelì and Pulga2014).

Charred textiles, mats and basketry have better chances of survival. With charring incomplete, several chemical processes interact, leading to physical alterations. After carbonization, the charred and usually shrunken items preserve in carbonized form (see e.g. experiments by Grömer et al. Reference Grömer, Gleba, Van Horne, Lipkin, Ruhl and Wright2023). Although there are partial transformations, the microstructure of the object usually remains substantially intact, with techniques often still visible and fibres identifiable. Plant fibres are often more stable in a carbonized state.

We describe the item in question from Vésztő-Mágor as ‘matting’, though ‘basketry’ is technically the correct term for the item in question—contrary to what is understood colloquially, however, it does not represent a basket. Thus, to avoid confusion, ‘matting’ with its suggested interpretation was used.

There are many different matting techniques: some involve even twining or knotting. The following terminology is used to describe the item from Vésztő-Mágor (Fig. 4):

Figure 4.

Technical details of a stake-and-strand basketry with an active and a passive element in under 1/over 1 technique. (After Wendrich Reference Wendrich1991, fig. 26).

Context and micromorphological description of the discovery

The Bronze Age deposits exposed in the closed shelter form 0.75 m of stratigraphy beneath the bottom of the ploughzone (Fig. 5). The matting was found inside a structure, likely a house (Feature 2, in Profile C). This house is the latest structure in the profile, with 8–9 floor replasterings visible, each c. 2 cm thick. There is a clay platform ‘step’, a stratigraphically higher deposit, possibly dug into another room floor in the house, but slightly elevated over neighbouring rooms. The orientation of this laminar clay structure suggests an east–west orientation of the building, which likely extends north and west. A hearth feature, less than a metre away from where the matting was found, is a common element inside Middle Bronze Age houses. Hearths were used continuously in the same spot, with ash and charcoal deposits swept up and redeposited into sequential floor and sub-floor fill layers (Kovács & Vicze Reference Kovács, Vicze, Sulas, Lewis and Arroyo-Kalin2022; Kovács et al. Reference Kovács, Röpke and Anvari2024).

Figure 5.

Profile C drawing, photograph of the matting’s context, and visualization of the carbonized object, whose location is highlighted within the white box in the photo. A sample taken for phytolith and micromorphological analysis is also highlighted. (Image: K. Winter, P. Duffy and K. Saunderson.)

The carbonized matting was found on top of a replastered clay floor, connected to a wattle-and-daub construction, highlighted in a slide taken for phytolith and micromorphological analysis (Fig. 6, thin section of area b in Figure 7).

Figure 6.

Resin-impregnated micromorphology slab of sample VES 3 (EU 11-11, SS 3). Thin sections are marked with pencil and location of phytoliths samples (VES 15–18) are shown as well. (Image: P. Karkanas.)

Figure 7.

Image (XPL) of thin section VES 3b. The earthen construction has fallen upside down on the underlying floor. The back part of this construction is burnt. (Image: P. Karkanas.)

The clay floor in Figure 7 consists of an unsorted mixture of calcareous loess and rounded clasts of various soil material most likely derived from reworked remains of decayed earthen structures. It is also characterized by variable amounts of other anthropogenic inclusions such as fine bone, charcoal and sherds. This type of constructed floor is typical of the Late Neolithic deposits at the Vésztő-Mágor site (Parkinson et al. Reference Parkinson, Gyucha and Karkanas2018) as well as of other prehistoric tell sites in Hungary (Kovács et al. Reference Kovács, Röpke and Anvari2024; Röpke et al. Reference Röpke, Kovács, Petõ and Kienlin2024). The floor surface appears clean and sharp. It is capped by an overturned earthen construction rich in plant imprints and matter, particularly at its base (rear part), where it dominates the structure and appears to be burnt. Its composition differs from those of the underlying floors in that it is relatively well sorted and devoid of anthropogenic inclusions. Its face is finely replastered several times with a laminated fine-grained calcareous material (Fig. 6). As can be seen in the slab (Fig. 7), the construction appears quite long but relatively thin (c. 3 cm: see Fig. 6). The preservation of the rear side and the finely laminated plaster on the front face suggest that this is probably an intact piece of earthen construction preserved in its full thickness (c. 3 cm). This earthen construction is similar to re-plastered wall daub structures found in several Middle Bronze Age tell sites (Kovács et al. Reference Kovács, Röpke and Anvari2024). Above the burnt structure, the matting is visible as a mass of primarily charred whole reed pieces (see below) mixed with large fragments of Ulmus wood charcoal.

Description of the item in basketry technique

The matting attached to the earthen daub construction comprises a dark, fibrous, charred layer 130 cm in length and 3–4 cm in thickness in the section, and was partially block-lifted for analysis. However, because the carbonized object is very brittle, the block-lifted part was poorly preserved and only allowed some interlinking to be observed. The profile of the object in situ revealed a weave structure through the characteristic waved course of the active system along the profile (Fig. 5). However, the passive system is not as visible as the active system—these elements, likely a different, thicker, and stable material, such as twigs, were not as easily charred and did not preserve, outside of a single chunk identified as elm (Ulmus). The passive elements are represented by round, brown-coloured sediment spots where the active element bends, and in one spot is replaced by a root. Most of those structures occur every 1.4–1.6 cm (Table 1). As the active system is cut lengthwise in the profile, only one of these elements is visible. We assume further, parallel elements remain in situ.

Table 1.

Technical detail of the basketry item as identified along Profile C and from in situ blocks. (Data: K. Grömer & K. Saunderson.)

Phytolith analysis of carbonized remains

Phytolith analysis targeted the matting itself (VES 17), two layers below it (VES 15 and VES 16) (Fig. 7) and another set of floor layers 4 m away but in a similar stratigraphic context (VES 8 to VES 14). Photographs of samples and quantitative data for phytoliths can be found in Table S1 of the Supplementary Material. Reeds dominate the silica skeleton (SS) count at 48.5% of 233 silica skeletons of the matting item VES 17 (Table 2).

Table 2.

Relative contribution of silica skeletons in phytolith samples.

Grasses are also present in the phytolith assemblages but most probably belong to the surrounding matrix rather than the matting itself (Fig. 8).

Figure 8.

SEM images. (a) Grass phytolith; (b) epidermal phytolith structures from grasses, both from the matrix surrounding the charred matting. (Image: T. Schroedter.)

In two carbonated fragments, the arrangement of the vascular bundles in two rows is typical for Poaceae (Fig. 9a and b) (Schweingruber & Berger Reference Schweingruber and Berger2018). The stomata, rondels with cork cells and slender dentritics in another fragment (Fig. 9c) suggest a reed culm, or possibly grass husk. Another fragment belongs to Cyperaceae (sedges), probably from the genus Carex (Fig. 9d) (Schweingruber & Berger Reference Schweingruber and Berger2019).

Figure 9.

SEM images of diagnostic features from four fragments from the matting. (a) fragment a: stem of Poaceae, cf. Triticum sp.; (b) fragment b: Poaceae, cf. Triticum sp. (c) fragment c: epidermis, Poaceae, cf. Triticum dicoccum; (d) possibly Cyperaceae, cf. Carex. (Image and taxa identification: T. Schroedter.)

The Vésztő-Mágor site falls in the Lower Körös Basin, whose environment during the Bronze Age contained forest-steppe, wetlands and areas of stagnant water away from active river channels (Járai-Komlódi Reference Járai-Komlódi, Pécsi and Kordos1987; Kosse Reference Kosse1979). Depressions in the latter areas would have carried reed beds and clump sedges (Carex elata), and stagnant waters would have had rushes (Juncus subnodulosis) and fenland sedge (Carex davalliana). The use of reeds and sedges to make basketry and matting is well attested in the historic record of the Great Hungarian Plain (Ilia & Juhász Reference Ilia and Juhász1960; Szűcs Reference Szűcs1933).

Dating

The dating of the matting confirms it belongs to the Middle Bronze Age, and the associated ceramics would generally be described as representing the Ottomány/Otomani culture. Two ¹⁴C dates were run on samples of the matting itself (Table 3), yielding 1875–1745 b.c. (UGAMS-63227) and 1870–1695 b.c. (UGAMS-63227) at 68.3%, or –1885–1695 b.c. and 1880–1685 b.c. at 95.4%, modelled in OxCal v.4.4, using IntCal20 calibration curve (Reimer et al. Reference Reimer, Austin and Bard2020). These are consistent with the dating of the associated material culture and other dates from the site that are being prepared for publication.

Table 3.

Radiocarbon dates of the matting from Profile C.

Interpretation

To interpret the function of the object, we consider the technique and composition, shape and context. The structure of the object is rather unbalanced, with the distance between the passive strands being irregular. This may only have occurred with its collapse or with collapse on top of it. In one area, the passive systems are quite dense, indicating the flexibility of the active systems.

A container, i.e. a basket, seems unlikely due to the shape and length. Rather, it could have been a mat on the wall or the floor, perhaps covering the whole room or only partially. That the material is attached to daub suggests that it formed part of another construction such as a plastered ceiling or plastered wall that had collapsed into a horizontal position. Interestingly, the back of the construction is burnt, suggesting that the fire responsible for its collapse burned it only on one side (Fig. 7).

Matting suspended below a ceiling of light materials (possibly allowing in daylight), serving as a shelf, or even internal room dividers are other interpretations. The matting could have been used to protect an indoor area from pollution from the inside of the roof, for example, along with preserving warmth more efficiently in this area. It might have created some space between the actual room and the roof that could also have been used for storage of light items. This ceiling could have collapsed during a fire and then been covered by the ash layer above. A collapse of the ceiling or low wall would also explain why it is rather poorly preserved in the area of the ‘steps’ of the layer below. The white material several centimetres below on the bottom side of the object is likely a prepared clay floor onto which the matting fell (Fig. 7).

Comparative material from Vésztő-Mágor

The carbonized matting remains are not the sole well-preserved organics found during excavations at Vésztő-Mágor. While exposing the Late Neolithic Tisza layers in the 1970s, Katalin Hegedűs discovered several inhumations in contracted positions within coffins made of wooden planks (Fig. 10; Hegedűs & Makkay Reference Hegedűs, Makkay, Tálas and Raczky1987). These wooden coffins were noted in more than half of the 30 burials dating to the Late Neolithic, and in some cases the sides of the coffins had preserved planks measuring up to 25 cm in height. Excavators also identified the remains of matting used for wrapping the deceased. These burials are associated exclusively with the lower Tisza stratum, suggesting that either wooden coffins and matting (as shrouds) were used only in the earlier phase of Late Neolithic habitation at Vésztő-Mágor, or that the lower layers had better conditions for preserving non-carbonized organic materials compared to the upper ones (Makkay 2004, 41). This would be consistent with the discovery of unburned wood in the lowest deposits of the Tisza tell, Szeghalom-Kovácshalom, located 7 km away (Parkinson et al. Reference Parkinson, Gyucha and Karkanas2018), though both Middle Bronze Age tells at Békés-Várdomb 20 km away and Toboliu in western Romania also exhibit preserved wooden floors at shallower depths (Banner & Bóna Reference Banner and Bóna1974; Lie et al. Reference Lie, Cordoş, Găvan, Fazecaş, Kienlin and Gogâltan2018).

Figure 10.

Late Neolithic burials in preserved wooden coffins. The excavators also noted finding the remains of matting in which the bodies were wrapped. (Image: Archives of the Hungarian National Museum.)

Hegedűs also excavated Bronze Age contexts in Block IV, a trench c. 20 m from where the in situ exhibit would be built (Archives of the Hungarian National Museum). Her team exposed a large area of a house floor which contained what she interpreted as ashy layers of sedges (pernyeszerű, sás réteg) (Fig. 11) (c. 120 cm below surface). Having seen similar deposits during excavations in 2022–23, by macroscopic inspection we interpret these as phytolith layers (Fig. 12). One such section of matting appears to be folded over with an additional cordage of some kind wrapped on part of the surface. Initial examination of the weaving pattern looks similar to a twilled weave. They followed this layer in an extension block (Block VII, c. 105–135 cm below surface), continuing with identified phytoliths and rubble on a clay surface.

Figure 11.

‘Sedge floor’, Surface IV, in Block VII, excavated in 1973. An area approximately 2.5 m by 3.5 m is shown. For location, see Figure 1. (Image: Archives of the Hungarian National Museum/KJM.)

Figure 12.

Phytolith layer discovered in Profile F–E, in 2023, c. 2.5 m from the profile with the matting. The area shown is approximately 12×15 cm. (Image: A. Lingle.)

In a deeper and earlier Bronze Age context (c. 175 cm below surface), Hegedűs’s team opened a larger area and observed that the direction of the reported sedge layer was parallel to the orientation of the long axis of the house. The consistent location and preservation of the phytolith layer and macrostructure of their deposits suggested to them that they were not roof building material but were instead placed directly on the floor as mats. Details regarding whether they were loosely strewn, twined, or sewn into a mat, however, are not clear from the notes or the photographs.

Compared to the matting remains analysed in this study, the ‘sedge floor’ discovered during the Hegedűs excavations, and those identified in Profile F-E, seem to represent an alternative use of grasses, a different form of preservation, and a different taphonomic pathway. The grasses of the Bronze Age houses of Blocks IV–VIII were all directly on top of a prepared floor surface, extending at least 5×4 m, presumably the majority of the floor of a single room. These are reasonably interpreted as floor mats (see Cameron 2017 for possible analogues). Unlike the matting discovered in Profile C of the covered trench, it was preserved as a structured light grey phytolith layer, and not as black carbonized plant material. This seems to be the case for the other phytolith layers discovered in the Bronze Age contexts at Vésztő-Mágor from the 1970s as well; the remains of the mats also left in situ before preparing new house floors. The dominant phytoliths discovered in Profile C were identified as reeds, likely the active material, with elm serving as the passive element.

Broader comparisons

There are no objects similar to this example of complex matting remains with daub we found at Vésztő-Mágor known from other excavated Bronze Age tells from the Carpathian Basin. The matting includes grasses (Poaceae), but these probably belong to the surrounding matrix. Poaceae have been identified as common features on Bronze Age house floors at the site of Borsodivánka in northeastern Hungary (Röpke et al. Reference Röpke, Wilde, Fischl and Kienlin2016) and Százhalombatta on the Danube south of Budapest (Kovács Reference Kovács, Vicze and Sørensen2023; Kovács & Vicze Reference Kovács, Vicze, Sulas, Lewis and Arroyo-Kalin2022). Micromorphological studies from Százhalombatta do, however, report distinct imprints of matting from Level 3 of the tell, in comparatively cleaner areas, which may have been associated with sitting or sleeping (Kovács Reference Kovács, Vicze and Sørensen2023, 266–7; Richter Reference Richter, Andersson Strand, Gleba, Mannering, Munkholt and Ringgard2008). An extended phytolith layer was found contemporary with House G at Borsodivánka, but in contrast to Véstző-Mágor, was not in the house, but in between houses in an alleyway (Fischl & Kienlin Reference Fischl, Kienlin and Kienlin2024, 243).

Items made in basketry techniques are known in Europe dating to the Neolithic (Alfaro Giner Reference Alfaro Giner1984; Bazzanella et al. Reference Bazzanella, Mayr, Moser and Rast-Eicher2003; Médard Reference Médard2010; Rimkute Reference Rimkute, Andersson Strand, Gleba, Mannering, Munkholt and Ringgard2008), the Bronze Age (Rast-Eicher & Dietrich Reference Rast-Eicher and Dietrich2015), the Iron Age (for the latest collection, see Grömer et al. Reference Grömer, Gutjahr and Heiss2018), and the Roman period (e.g. Gostenčnik Reference Gostenčnik2001). In Anatolia, sedges (Scirpus sp.) were commonly used to provide matting at Çatalhöyük (Ryan Reference Ryan2011; Wendrich & Ryan Reference Wendrich and Ryan2012), likely obtained from the surrounding marshes (Rosen Reference Rosen and Hodder2005), similar to the phytolith layers found in the Hegedűs excavations at Vésztő-Mágor. Their spongy leaves and stems would be comfortable on floors to walk on, and sedges were also used for adult burial mats at Çatalhöyük (Ryan Reference Ryan2011). Reeds have been used more sporadically for the same purpose (Ryan Reference Ryan2011), as opposed to Aşıklı Höyük where reeds have almost exclusively been used for burial mats (Tsartsidou et al. under review). Among the techniques, matting, coiled basketry and wickerwork are still in use today in Hungary and Romania.

The stake-and-strand techniques identified from Vésztő-Mágor may have been employed for smaller items, such as baskets and other household objects. From Late Bronze Age lake dwelling sites in Zürich and Greifensee-Böschen in Switzerland, various baskets in stake-and-strand technique have been found (Rast-Eicher & Dietrich Reference Rast-Eicher and Dietrich2015, 149–50). Similar artefacts in the same technique come from Iron Age cult sites and graves such as the Býčí skála cave in the Czech Republic (Rast-Eicher Reference Rast-Eicher, Parzinger, Nekvasil, Barth and Schnurbein1995) or Kainach in Austria (Grömer et al. Reference Grömer, Gutjahr and Heiss2018). Household objects such as boxes or chairs in stake-and-strand technique are known both from actual objects but also from pictorial sources in the Roman period (Gostenčnik Reference Gostenčnik2001). The same technique was also used for construction details, such as fences and the inner structure of walls, as seen at the lakeside settlements around the Alps (Schlichtherle Reference Schlichtherle2016, fig. 26). A reconstruction of such a light structure within a house can be seen in a Neolithic house model at the open-air museum at Dispilio, Greece (Fig. 13). The upper loft floor, a platform of logs, may have served a similar purpose to the matting found at Vésztő-Mágor. The composition, however, would look more like the unplastered wall visible below on one side. Although no evidence of stake-and-strand technique was found at Bosodivánka, the excavators identified internal partition daub walls in House G, each 8–9 cm thick (Fischl & Kienlin Reference Fischl, Kienlin and Kienlin2024, 245). Due to their thickness, they do not expect they reached the roof but were only half-height room dividers. The basketry matting with daub identified at Vésztő-Mágor could have served a similar function, if not suspended horizontally as suggested in Figure 13.

Figure 13.

Example of the inner structure of a Neolithic house, based on archaeological evidence from Dispilio, Greece. Matting such as that discovered at Vésztő-Mágor, instead of a log beam platform, could serve a similar purpose. (Image: www.limneosoikismos.gr/gallery, accessed 24 July 2025.)

Similar horizontal basketry findings in buildings are nonetheless known, for example, from Early Medieval (ninth-century) Haithabu in northern Germany. Most of these were interpreted as collapsed wattle-and-daub walls. In contrast with Vésztő-Mágor, they still showed traces of thick daub on the walls rather than the 3 cm associated here. Some have also been described as mats in front of house entrances that offered a dry passage into the building. Both were built in the same structure. Such an outside mat seems less likely for the object from Vésztő-Mágor, as more sediment would be expected, as was observed in Haithabu (Schietzel & Bau Reference Schietzel and Bau2014).

In learning about the built environment at Bronze Age tells, we increasingly recognize that people often did similar things under different environmental constraints or made different building choices given similar environmental conditions. Currently it seems that choosing unweathered local sediment for floors is an example of consistent behaviour from the Danube to the Tisza region during the Bronze Age (Kovács et al. Reference Kovács, Röpke and Anvari2024), while use of wooden floors seems more restricted to the Körös river drainage (Fischl & Kienlin Reference Fischl, Kienlin and Kienlin2024; Lie et al. Reference Lie, Cordoş, Găvan, Fazecaş, Kienlin and Gogâltan2018). Internal partitions in Bronze Age structures appear at many settlements across the Great Hungarian Plain, though specific technological choices in constructing them may have been due to unconscious, inherited, habitual traditions rather than being overtly symbolically meaningful (cf. Lemonnier Reference Lemonnier1986). Using matting or basketry with daub to modify spaces with low dividing walls or ceiling storage may have been a less common practice that we observe only in the Körös region, though this conclusion may change as we become better attuned to the analysis of construction techniques through micromorphology, phytolith, and greater attention to the details of material remains at tells.