The perishable nature of textiles often limits archaeological research on ancient textile production, an activity that is critical to our understanding of the human past. In the absence of textiles, ceramic loom weights, use to aid in the weaving of cloth, provide some of the best and most abundant evidence for this time-consuming and labor-intensive activity in regions where the warp-weighted loom was utilized. In this dissertation, I investigate ancient textile production practices and examine the socio-economic impacts of this activity in southwestern Iberia (Spain and Portugal) during the late 4th and 3rd millennia BCE (3400-2000 BCE), through the study of loom weights. I present the results of three distinct but related research projects that contribute to our understanding of ancient textile production practices in this context and use my results to inform current debates on site function, socio-economic change, and regional interactions in archaeological research on the Chalcolithic in Southwestern Iberia. These projects include 1) experimental weaving and twining with crescent shaped loom weights, 2) a diachronic and contextual analysis of loom weights from the site of Perdigões (Alentejo, Portugal), and 3) a diachronic and contextual analysis of the loom weights from the site of Valencina de la Concepción (Andalucía, Spain).
My experimental research, presented in chapter 2, aims to address the doubts that surround the identification of ceramic crescent as loom weights in Iberia. To do this, I test the efficacy of using these crescents as loom weights in three methods of production, including vertical band weaving, warp twining, and warp-weighted weaving. My results suggest that the ceramic crescents would have been capable of serving as loom weights or as warp twining tools, despite their relatively light weights. I use the results of these experiments to inform my diachronic analyses of the loom weights from Perdigões and Valencina.
In Chapter 3, I analyze an assemblage of 3,153 loom weight fragments found in the central area of Perdigões, known as Sector Q (3400-2000 cal. BCE), examining their quantity, variability, distribution, ceramic composition, and fragmentation over time to assess the socio-economic and socio-political role of textile production at this notable ditched enclosure site. My results indicate that, in terms of size and typology, the loom weights in Sector Q remained remarkably consistent over time and appear largely similar to those found at other sites in the Alentejo. In terms of frequency, my results indicate that the presence of loom weights increased dramatically during the Middle and Late Chalcolithic and remained high during the transition to the Early Bronze Age while the southwest was experiencing a period of decline and many other sites, like Valencina, were abandoned. I argue that an investment in long-held regional textile traditions likely helped to support the continued occupation of this site during the second half of the 3rd millennium BCE.
In Chapter 4, I analyze 499 loom weights from the northern area of Valencina, including 317 from Cerro de la Cabeza (3300-2350 cal. BCE), 179 from Nueva Biblioteca (2900-2450 cal. BCE), and 3 from La Gallega (ca. 2480-2285 cal. BCE), examining quantity, variability, distribution, and fragmentation overtime and across settlement areas within this ‘mega site.’ I include a discussion of the Montelirio beaded garments, considering how these elaborate textiles might have been made in the context of this diachronic analysis, and use these distinct traces of textile production to contribute to current debates on the character of this ditched enclosure. While the Montelirio beaded garments attest that some textiles were created for symbolic and political purposes at Valencina, the loom weights from Cerro de la Cabeza suggest that textile production was more often undertaken to meet the needs of a long-term settlement with a fluctuating population. In terms of frequency, the quantity of loom weights at Valencina increases or decreases relative to the intensity of the settlement over time, with the number of tools starting off small in the Early Chalcolithic, reaching its peak in the Middle Chalcolithic, and declining in the Late Chalcolithic. Overall, my results indicate that textile production was an important component of a subsistence-oriented economy in the northern area of Valencina.
In the concluding chapter of this dissertation, chapter 5, I aim to inform current perspectives on site functions, regional interactions, and social organization in southwestern Iberia during the 3rd millennium BCE through the lens of textile production. I compare the results of my diachronic analyses and consider the various socio-economic roles that textile production played at both sites. At both Perdigões and Valencina, increasing evidence for textile production coincides with an influx of non-local materials, monumental ditch constructions, and increasing evidence for other production and consumption activities. While textile technology (i.e. crescents and plaques) was largely shared across these sites, I found more uniformity in subtype between sites like Perdigões and São Pedro in the Alentejo region, than across areas of excavation within Valencina. Based on this comparison, I argue that textile traditions were more closely maintained over time in the Alentejo and shared between ditched and walled enclosure sites in this region. This suggests the people living at these distinct site types were more likely part of the same community, rather than two culturally discrete groups. Furthermore, as the quantity of loom weights found at sites in the Alentejo appears to be considerably higher than at Valencina, this dissertation lends support to the idea that the Alentejo might have served as a center for textile production during the Chalcolithic. Finally, based on the combined results of all three of the research projects undertaken as part of this dissertation, I suggest that crescent and plaque loom weights might have been used in tandem to create textiles with a starting boarder earlier than we previously thought.