ºÚÁÏÊÓƵ

Introduction

Experimental archaeology encompasses a range of methodologies, from scientifically structured experiments with measurable variables to exploratory practices that focus on techniques and the broader processes of making (Coles, 1979; Currie, 2022). Proponents of scientific experiments champion their rigour, replicability, and control over variables as the most reliable methods for investigating historical technologies (Outram, 2008; Reynolds in Harding, 1999). However, such approaches face criticism for potentially neglecting the subjective and nuanced nature of traditional craft practices, especially when applied to ephemeral objects (Hurcombe, 2016; Hurcombe, 2008b; Ingold, 2023). Factors such as time, physical effort, and skills, which are critical components of craft, are often excluded from experimental frameworks. This can result in interpretations that oversimplify or distort historical realities (Coles, 1979; Currie, 2022; Hurcombe, 2014; Hurcombe, 2008b). This tension is particularly evident in the reconstruction of ephemeral technologies such as plant-based crafts, textiles, and basketry; perishable objects which remain underrepresented in the archaeological record due to their organic composition, vulnerability to decay, and routine disposal after use (Hurcombe, 2008a; Hurcombe, 2008b; Ingold, 2023; Fehon, 1978; Jacomet in Menotti and O'Sullivan, 2012). Hurcombe has described such artefacts as part of the "missing majority" of material culture (Hurcombe, 2014). Their absence not only obscures everyday practices but also limits understanding of the symbolic and ecological knowledge embedded in traditional material culture. This is particularly significant given that, beyond their structural utility, many of these objects embody aesthetic and symbolic features that express relationships with local environments, what Ingold (2023) refers to as "cultural nature connections".

Traditional cleft hazel baskets exemplify these issues. Once integral to daily life in Wales, they have largely disappeared, their fragile nature and utilitarian purpose leaving few physical traces (Jenkins, 2009). Deemed unremarkable, they were rarely documented or preserved, and their fragmented remains are difficult to identify archaeologically. Nevertheless, oral histories collected by St Fagans National Museum of History (Amgueddfa Cymru), along with trade directories and census records from the nineteenth and early twentieth centuries, attest to their widespread use and the presence of local basket makers across Welsh communities (St Fagans, 2025). Despite the scarcity of surviving examples, these sources document their everyday function and cultural significance within a now largely vanished material tradition. Crucially, knowledge of hazel basket making has survived, though not through direct familial transmission, but rather through the efforts of artisans who actively sought out the last remaining makers to learn and preserve these skills (Heritage Crafts Association, 2019; Heseltine, 1982; Jones, 1927). This fragile heritage kept alive by a handful of makers remains largely undocumented and understudied (Pybus, 2016; Revera, 2019).

To address these gaps, this research undertakes a comparative transdisciplinary reconstruction of a traditional Welsh hazel basket, combining a practical descriptive reconstruction by a volunteer group with ethnographic insights from traditional makers. While it does not follow a hypothesis testing model typical of experimental archaeology (Outram, 2008; Reynolds in Harding, 1999), it aligns with emerging approaches that emphasise embodied making, ecological literacy, and cultural transmission as valid and necessary modes of inquiry (Currie, 2022; Hurcombe, 2008b). By examining both the physical reconstruction process and the knowledge systems embedded in traditional practice, this research demonstrates how experiential and ethnographic methods can reveal aspects of ephemeral craft technologies often overlooked by formal experimental approaches. In doing so, it contributes to ongoing methodological debates in the ºÚÁÏÊÓƵ and EuroREA Journals (Martins, 2022) and underscores the importance of engaging with traditional crafts not only as technical skills, but as practices grounded in ecological awareness, embodied experience, and relational knowledge of materials and environments.

Methodology: Volunteer Descriptive Data

This study employed a transdisciplinary methodology, defined by the integration of academic and non-academic knowledge as co-producers of insight to address complex problems through collaborative inquiry (Hadorn et al., 2008). In this case, the research combined practical reconstruction with participant observation (Spradley, 1980), drawing on practice-based and ethnographic approaches (Ingold, 2013; Hurcombe, 2008b; Currie, 2022) to compare two distinct groups: a volunteer cohort engaged in experiential learning through basket reconstruction, and traditional makers whose skills are grounded in generational, place-based knowledge.

Between January and June 2020, the Swansea University Heritage Team coordinated a hands-on reconstruction project involving twelve volunteer participants. The group included amateur basket makers, coppice workers, historians, and one experienced willow basket maker. Over six months, the group collaboratively constructed twenty-four cleft hazel baskets using documented techniques, historical sources, oral accounts, and surviving examples.Each stage of the making process, harvesting, splitting, preparation, weaving, and finishing was documented through photography, video, and structured field notes. In parallel, volunteers participated in semi-structured, task-specific interviews at key moments, reflecting on decision making, challenges, and learning progression. Combined with participant observation, this generated insight into how experiential knowledge developed and was interpreted across the group (Spradley, 1980). The resulting descriptive data were analysed using a coding framework to trace activity sequences, material responses, problem-solving strategies, and decision-making patterns (Frankenberger, Badke-Schaub and Birkhofer, 1998; Hadorn et al., 2008). This approach enabled the identification of patterns in learning, adaptation, and reconstruction techniques.

Traditional Makers: Ethnographic and Thematic Data

To contextualise and contrast the experiential findings from the volunteer reconstruction, traditional knowledge systems, ecological understanding, and embodied practices were documented using an ethnographic approach (Hammersley and Atkinson, 2007; Spradley, 1980). As part of this, five in-depth, open-ended interviews were conducted between June 2020 and January 2021 with experienced Welsh hazel basket makers. To complement these interviews, the research team conducted in-person observations of four basket making sessions led by experienced makers to enable close attention to subtle but significant aspects of skilled practice. These encounters allowed for real-time clarification, contextual enquiry, and direct comparison with methods used by the volunteer group. Subsequent virtual interviews and continued exchanges provided opportunities to refine earlier observations and further explore techniques, timing, and material selection.

Unlike the descriptive coding applied to the volunteer data, the interview transcripts, field notes, and photographic documentation from the traditional maker fieldwork were analysed thematically using an inductive approach (Guest, MacQueen and Namey, 2012). This distinction reflects the differing nature of the datasets: the traditional maker material provided nuanced narrative depth and interpretive insight. Findings from the traditional makers were then systematically compared with those from the volunteer reconstruction. To enable meaningful comparison, a structured framework was developed that aligned the two data sets along three key dimensions: 1) material sourcing and ecological timing, 2) construction techniques and interaction with tools, and 3) aesthetic judgement and symbolic design. This matrix allowed a detailed assessment of where volunteer practices aligned with or diverged from traditional knowledge. These were interpreted in relation to wider scholarship on tacit knowledge, sensory learning, and vernacular material culture (Pink, 2015; Downey, 2010; Ingold, 2013.

Volunteer Reconstruction: Descriptive Findings

Stage 1: Harvesting and Sourcing Materials

Hazel was harvested during the winter months from a traditionally managed coppice, following seasonal recommendations recorded in the oral history evidence at St Fagans National Museum (St Fagans, 2025). Because the group's weekly schedule introduced a delay between harvesting and using the rods, the coppice worker advised that the materials should be stored outdoors in shaded conditions to retain their pliability. While not a hazel basket maker himself, his guidance was informed by ecological knowledge from hazel coppicing and proved partially effective, as storing rods in cold and shaded conditions helped maintain basic flexibility. However, these methods did not reflect the more nuanced and intentional storage and preparation practices employed by traditional makers. For the experienced basket makers, harvesting and storage were not incidental tasks but integral stages of the making process, requiring close observation and careful judgement. Harvesting was timed to coincide with seasonal conditions that maximised strength and flexibility. Once gathered, rods were sorted according to their intended function, with some stored upright and others laid flat depending on their future use. Many were also pre-shaped by hand during collection to anticipate their role in the finished basket. These subtle but significant actions reflected a deeply embodied understanding developed through sustained engagement with both material and environment. This divergence in technique and foresight highlighted a broader challenge faced by the volunteer group. Without access to this kind of tacit, embodied knowledge, participants lacked the sensory cues and experiential judgement needed to assess and prepare materials with confidence. Instead, decisions were made through extensive, often iterative group discussion, typically relying on the input of individuals with historical or loosely comparable expertise, such as coppice workers or willow weavers. This reliance on indirect knowledge became especially apparent when the group attempted to estimate the quantity and dimensions of hazel rods needed for basket construction. In the absence of intuitive, experience-based judgement, decisions were made through trial, approximation and group consensus rather than established practice.

This approach also shaped how material requirements were determined. Through hands-on experimentation and close inspection of surviving examples, the group eventually concluded that a single basket typically required between three and seven rods, each approximately 3 cm in diameter and 3 metres in length. While this estimate proved broadly accurate, it did not account for variations in material yield, such as the number of usable weavers that could be obtained from each rod. These inconsistencies affected both the efficiency and the uniformity of the making process.

Stage 2: Material Preparation-Readying the Weaves, Ribs, and Rim

The preparation of materials followed historically documented techniques for splitting and shaping hazel rods into weavers, ribs, and rim components (St Fagan, 2025). To produce weavers, a shallow incision was made at the top of the rod, cutting down to a single growth ring. This ring was then peeled back along the length of the rod as a thin strip of wood. Each rod yielded between eight and ten weavers, with approximately thirty needed to complete a single basket. Although the method was technically replicable, volunteers frequently experienced breakages. These were due in part to suboptimal storage, which caused the hazel to develop a plastic-like consistency, and in part to limited physical dexterity when handling the material. To reduce breakages, volunteers attempted to keep the weavers pliable by working outdoors in humid conditions or by soaking the rods prior to use. Both techniques were adopted on the advice of a traditional willow worker, who noted that willow responds well to rewetting and high humidity. However, while these measures offered some benefit, they provided only a partial solution. The advice, based on a different material, did not fully address the specific handling requirements of hazel, which must be selected with care, worked while fresh, or physically manipulated by hand to restore flexibility. As a result, breakages persisted, leading to repeated material failures. In response, volunteers developed a workaround by overlapping the ends of broken weavers and tucking them beneath existing ribs, a method not typically used by traditional makers, as it reduces the tensile strength and structural integrity of the basket.

In contrast, the preparation of ribs proved even more challenging for volunteers. Traditional techniques, are well described in evidence from St Fagans, involved splitting thicker rods into usable lengths by using a knife in place of a froe to cleave the wood along its grain (St Fagan, 2025). However, few volunteers were able to achieve this, and none could produce consistent results. Instead, they resorted to repurposed broken weavers that were sufficiently thick to serve as ribs, a method that deviated from traditional practice. Ribs and weavers were smoothed by drawing the hazel strips beneath a knife, a technique that volunteers could replicate successfully, though with limited finesse compared to traditional makers. For constructing the rim, three methods were attempted. The first, commonly used by traditional makers, involved splitting hazel rods to isolate the dense inner growth rings. This technique produces strong, tensioned rims that shape the basket's overall form, but it demands precise splitting, grain alignment, and physical dexterity. Despite multiple attempts, none of the volunteers successfully completed this method, as the material proved difficult to control without prior experience. The second method, using two curved but unsplit rods, was more accessible. Volunteers found it more achievable after repeated visits to collect and test rods with the right combination of strength and flexibility. This method, also used by traditional makers, allowed for improved balance and tension, although volunteers struggled with symmetry and control. The third method involved using a single curved hazel rod. It was the simplest to compete by volunteers but is rarely favoured by experienced makers, as it limits basket size and provides less control over the rim's structural integrity.

Stage 3: Weaving the Basket, Techniques and Structural Formation

Weaving began with three ribs bent into a shallow bowl shape and loosely positioned against the rim to allow for initial flexibility. Using a long, thin weaver, volunteers secured the ribs in place by creating a crosshatch pattern that locked the central structure. Weaving continued as additional weavers were threaded over and under the ribs, with their ends carefully tucked out of sight. While these foundational techniques are well documented and closely resemble those used in willow weaving, one aspect proved particularly challenging with hazel. At the end of each pass, the weaver must be twisted to reverse direction and maintain consistent tension around the ribs. This is an essential technique for shaping and tightening the basket. Many volunteers struggled to perform this movement, as hazel is less pliable than willow and requires greater precision and control. As a result, the weave often became uneven, with visible gaps and occasional material splitting (See Figure 1). The addition of ribs at regular intervals reinforced the basket's structure, achieved through the time-consuming process of manually adjusting the spacing of the ribs and weaves followed by verifying the alignment by repeatedly turning the basket. The central keel rib served as a structural strut, with supporting ribs added in pairs, while the final two ribs formed the handle which was intentionally left uncovered by weaving to define its shape and function. Achieving this shape? Structure? required precise eye judgement to position and stabilise the ribs, yet volunteers frequently misjudged the spacing, necessitating the addition of an extra rib on one side that negatively affected asymmetry and structural balance.

Excluding the initial attempt, during which time was not systematically recorded, constructing a typical 35 cm by 35 cm circular basket took the volunteer group an average of twelve hours. This included approximately two hours for material selection and preparation, followed by around five hours each for weaving and final assembly.

Discussion: Embodied and Ecological Understanding, and the Limits of Reconstruction

Traditional Ecological Knowledge

In comparison, traditional makers prioritised material quality over measurable variables, sourcing materials from specific locations and individual stools at specific times. They deliberately avoided managed coppices, which produced quick growing, brittle rods unsuitable for basketry. Instead, they favoured shaded, wild woodland stools where slower growth resulted in tightly packed growth rings that enhanced pliability and durability. This ecological understanding extended to seasonal timing. Rims were typically cut and spliced in summer, and weaves were either used immediately or stored in water troughs for prompt use. These practices contrasted sharply with those of the volunteer group, who relied on managed coppices and worked to a weekly meeting schedule that introduced delays between harvesting and use. Traditional makers' place-based knowledge of environmental factors such as growth patterns and seasonal conditions enabled them to select specific rods on stools in precise micro locations that they may have been observing for years. This expertise minimised weaver breakages, enhanced structural integrity, reduced warping, and supported a higher degree of aesthetic refinement.

In addition to their technical skill, traditional makers placed great importance on the artistic and cultural dimensions of their craft. They incorporated features such as retaining the outer bark on the basket's exterior and alternating weave depths and widths to create intricate patterns imbued with cultural significance (See Figure 2). These aesthetic choices embodied biophilic design principles, enhancing both strength and durability while fostering a symbolic connection to the wild woodland environment. For example, retaining bark on the basket's surface was a deliberate way of reflecting the material's natural origins, while also adding texture, strength, protection, and visual distinctiveness. In contrast, the volunteer group focused primarily on functional replication, often overlooking these structural and symbolic subtleties.

Embodied Knowledge and Adaptive Problem Solving

The challenges faced by volunteers underscore the critical role of embodied knowledge in traditional basket making. Traditional makers relied heavily on tactile feedback, such as running their hands underneath the basket to assess symmetry. Volunteers, lacking this experience, often relied on visual inspection and corrective adjustments, resulting in less refined, more time-consuming outcomes and frequent miscalculations that required ad hoc corrections. Physical interactions with materials, such as bending rods over the knee to prepare them for use or softening weavers by repeatedly wrapping them around wrists and fingers, are not documented in the historical record, yet they served not only practical construction purposes but also therapeutic functions. For instance, volunteers experimented with steaming to soften the hazel. They found this technique effective and assumed it might have been used historically, as it features in comparative practices like hedgelaying and was recommended by the traditional coppice worker. However, traditional makers explained that steaming was rarely used in their craft due to the difficulty and time required to set it up. Instead, they favoured manual techniques, not only for their practical effectiveness and sensory responsiveness, but also for the therapeutic value inherent in direct engagement with the material. The physical manipulation of rods involved subtle softening movements that helped overused body parts to relax. This divergence between the groups highlights the importance of embodied practice in traditional basket making, showing how sensory interaction shapes not only the crafting process but also sustains a holistic understanding of the material.

This contrast was especially evident when non-traditional makers attempted three different methods for constructing the rim. Although some descriptions exist in historical records and the structure can be reasonably inferred from surviving baskets, the main challenge was not accessing information but applying it without the embodied understanding essential to traditional practice. What seemed straightforward in textual accounts or physical examples often proved far more difficult in practice. The failure to complete the split-ring method highlights the importance of tacit knowledge, such as how to judge the direction of grain or apply pressure while bending, that cannot easily be conveyed through written guidance alone. Even in the more accessible methods, such as using curved rods, volunteers relied heavily on trial-and-error and repeated experimentation to source and shape materials. This contrasts with the intuitive adjustments made by traditional makers, who routinely assess and prepare rods according to their intended use, often beginning the shaping process during harvest. These observations reinforce the conclusion that rim construction is not just a technical step, but a sensory, adaptive practice embedded in broader ecological and cultural understanding (See Figure 3).

Construction Time and Structural Integrity

The construction of a 55 cm by 35 cm oval basket took traditional makers an average of three hours, compared to twelve hours for the volunteer reconstruction group. Traditional makers typically allocated about one hour to each key stage: material preparation, weaving, and final assembly. In contrast, the volunteer group required over five hours each for weaving and assembly, and two hours for material selection. A key factor in this proportional disparity was the approach to material sourcing. Traditional makers treated it as an integral part of the making process, selecting hazel from specific stools and locations at the optimal point in the growth cycle. This attention to timing and quality significantly reduced breakages and improved workability during later stages. For them, sourcing was not merely preparatory, but a craft practice itself.

Conclusion

The reconstruction of traditional Welsh cleft hazel basketry reveals the challenges of recreating ephemeral objects that embody deep cultural, ecological, and technical knowledge that is often inaccessible through conventional archaeological or historical methods. Volunteers encountered considerable difficulty in sourcing, preparing, and weaving hazel rods, largely due to a lack of embodied skill and ecological familiarity. In contrast, traditional makers drew on generational, place-based expertise to assess material quality and adapt to environmental conditions. This contrast highlights a clear disconnect between intuitive craft practices and reconstruction efforts based on observation or assumption. Approaches shaped by partial or inaccurate interpretations frequently compromised both the structural integrity and aesthetic coherence of the baskets.

These findings underscore the critical role of sensory engagement and embodied intuition in traditional craft, where tactile skill, environmental awareness, and continuous adjustment are essential. Artistic biophilic elements, such as weave patterns, depth variations, and retained bark, were integral to traditional designs, reflecting both functional needs and symbolic relationships with the natural world. These elements, often absent from the volunteer group's utilitarian focus, illustrate the limitations of reconstructions that lack cultural context. Ultimately, this study affirms that technical replication is insufficient for fully reconstructing ephemeral craft objects. A fuller understanding requires engaging with them as living practices, where making is inseparable from the maker's sensory, ecological, and embodied knowledge.