Mariana Pereira[i]

Loom elements have been considered traces of weaving in many Portuguese archaeological sites dated from the “Calcolithic”. However, they are in fact a very heterogeneous group, whose variety is seen not only in the number of perforations, in morphologies and weights, but also in the different contexts where they have been found. This diversity has not been clearly transmitted by the commonly used term “loom weights”. At the same time, the role that the types of loom elements would have has not been fully understood.
In this paper, the term loom element/component will be suggested as the most appropriate one to refer to all materialities possibly related to weaving. Furthermore, two studies from 2010 have contributed towards their better understanding: Pereira has shown, through the study of the loom elements from Castanheiro do Vento (Guarda) and in particular from the sub-circular structure A, that they might not always represent weaving.
From another perspective, Costeira has studied the loom elements from S. Pedro (Évora) and explored their relation with weaving by proposing a typological study that would help define their purpose.
Finally, since more research is required, a study guide for the loom elements will be explained, calling attention to the need for data to be more rigorously collected and compared if the loom elements are to be better understood.

Key words: loom elements/components; “Calcolithic” in Portugal; study guide.

Os elementos de tear têm sido considerados vestígios de tecelagem em muito dos sítios arqueológicos portugueses datados do “Calcolítico”. No entanto, eles são, de facto, um grupo heterogéneo, cuja variedade é transmitida não só pelo número de perfurações, pelas morfologias e pesos, mas também pelos diferentes contextos em que são encontrados. Esta diversidade não tem sido claramente transmitida pelo termo “pesos de tear”, que tem sido usualmente empregue. Ao mesmo tempo, o papel desempenhado pelos vários tipos de elementos de tear ainda não foi totalmente compreendido.
Neste artigo, o termo elemento/componente de tear será sugerido como sendo o mais apropriado para referir todas as materialidades possivelmente relacionadas com a tecelagem. Além disso, duas investigações de 2010 contribuíram para o seu melhor entendimento: M. Pereira mostrou, a partir de um estudo sobre os elementos de tear de Castanheiro do Vento, e em específico da estrutura sub-circular A, que eles podem nem sempre denotar a tecelagem.
Numa outra perspetiva, C. Costeira estudou os elementos de tear de S. Pedro (Évora) e explorou a sua relação com a tecelagem ao propor um estudo tipológico que ajudaria a definir a sua finalidade.
Finalmente, e uma vez que mais pesquisas são necessárias, será apresentado um guia de estudo para os elementos de tear. Neste, chama-se a atenção para a necessidade de haver mais dados rigorosamente recolhidos e comparados, pois somente assim eles poderão ser mais bem compreendidos.      
Palavras-chave: elementos/componentes de tear; “Calcolítico” em Portugal; guia de estudo.

I – Introduction

Within the studies related with Portuguese prehistory, loom elements (a variety of small baked clay objects with perforations) have been interpreted as evidence of weaving by the majority of researchers (such as Boaventura, 2002; Diniz, 1994; Valera, 2007; among others[ii]). This correlation has shaped the way archaeologists include these materialities in their perceptions of the communities of the past and of the archaeological sites, especially because they have been considered typical finds from the period between the end of the IV and the end of the III millennium BC[iii]. Despite this, the most common terminology used to define them (“loom weights”) reflects an unclear understanding of what kind of role loom elements would have inside weaving. In this paper, two approaches that have contributed to clarify their role will be presented and explained.
From one side, Pereira (2010)[iv] carried out a research that focused, for the first time, on the loom elements found, until 2009, in the archaeological site of Castanheiro do Vento (Guarda). The intention was to further develop the ideas presented by Gomes (2003), who, after studying the loom elements from the site of Castelo Velho (Guarda), suggested that different meanings might have been given to them depending on the context in which they were part.
From another point of view, Costeira (2010)[v] focused on the loom elements found in the archaeological site of S. Pedro (Évora) and studied them from different perspectives. Starting from the idea that they are remains of weaving, the intention was to build a typology that, based on a morphological and technological analysis, would contribute to the clarification of their function and serve as a model for future studies. At the same time, the focus was to understand, in spatial and chronological terms, the large amount of loom elements that were found.
Having in mind that, for the focused period of time, there isn’t enough information concerning weaving mechanisms and the use given to the various loom elements, these two contributions have helped deepen their study, opening the way for further investigations.

II – From “loom weights” to loom elements/components

The presence of loom elements in archaeological sites has been regarded as a sign for the past existence of weaving. However, when going through the many references concerning these materialities, it can be seen that there is still some confusion between the terms used to name the different types of loom elements and their purpose inside weaving. While some researchers specify, in their publications or reports, the types which were found, others use the term “loom weight” to refer to all the different loom elements, making unclear which type(s) were discovered. For this reason, the terms will be here clarified.
The materialities which are here defined as loom elements can be divided in three major groups: loom weights, ceramic tablets (or cards) and “crescents”.  The term loom weight has been incorrectly used to describe all the remains related with weaving, and whereas some researchers specify if they mean ceramic tablets, for example, others don’t clarify the type that is being mentioned. Loom weights should only refer to complete or fragmented baked clay objects that could be used as weights to keep the tension of the yarns. It is a very heterogeneous group, in which the morphologies range from the rectangular to the somewhat oval shapes, with normally two or four perforations. Compared to the other types of loom elements, loom weights are thicker and heavier, even with their different sizes (for example, the complete ones found in Castanheiro to Vento have between 350 g and 640 g.) 
In weaving, loom weights are used to maintain the tension of the warp in vertical looms (hence the name warp-weighted looms. See Image 1).
And even though these looms were never found in “Pre-historic” contexts in Portugal, their presence is assumed through the loom weights. Important contributions concerning the study of textiles and weaving for “Pre-History” were made by several authors, such as Barber (1991), Geijer (1982), Giner (1984) and Hoffman (1964). This last author has shown that some looms, such as the backstrap loom, don’t require any components besides wood and yarn (Barber, op. cit). In archaeological sites, however, there is no evidence either to suggest their use or to deny it. In fact, and due to this, the absence of loom elements cannot be considered a sign for the total inexistence of weaving, even if no information is available. The only possibility is, therefore, to explore all aspects regarding loom elements and their presence in the archaeological sites.
 Hoffman (op. cit), even though focusing on warp-weighted looms from the XX century, provided very useful information concerning these types of loom and described two main possibilities: one, in which the weights used would be chosen in relation to the quantity of yarns (and so in the same loom, the weights could have different weights (g.) and sizes); and another, in which the loom weights would have to be similar.
For this reason, different loom weights, when found together in archaeological sites, might indicate different looms, but they could likewise point to only one mechanism. And, since these weights were easily transported, their presence doesn’t always mean that a loom once stood there, or that there was a fixed weaving area. It is, therefore, difficult (or nearly impossible) to know the quantity of weights that might be associated with a loom, mainly because it would depend on the size of the fabric being woven. This heavily influences the interpretation of contexts in which they appear.
 Another unclear, yet very important characteristic of loom weights is the perforations. The motives for their variety are unknown, but this might be because there is no knowledge of the weaving techniques. Notwithstanding, loom weights have been found with two to four perforations, and these also have different sizes and formats, even within a single loom. It might be that there was no need for the perforations to be similar, particularly if one considers the possibility that the yarns were not directly attached to the weights, but could be attached to a yarn loop, which would go through the perforations (Pereira, 2009).
From the variety of perforations, one question arises: could loom weights with different number of perforations be used together? This is important given that different loom weights have been found together (see Image 2), and if they could not have been used in the same loom, then they could be evidence for various looms (still, the reason for why they were found together becomes, then, even more significant).   
The second type of loom elements is the group of ceramic tablets. These are square or rectangular tiles, of light weight, with several perforations: usually they range from two to four in total, but Costeira (op. cit) has described, among all the loom elements found in the site of S. Pedro, half of what may be a tablet with three perforations (in total it could have six). These tablets could have been used for what is now known as tablet weaving, given that they are too light to be used as weights.
Instead of helping maintain the tension of the yarns, like the loom weights, the tablets would be used to create patterns by twisting yarns (see Image 3). The diverse number of perforations could have meant that they were used for creating specific patterns, but again more questions arise: could different ceramic tablets be used together (having different number of perforations and different weights)? Even considering that all these loom elements were handmade, there was the knowledge to make them similar if there was a need to. So, can the variety indicate different techniques inside tablet weaving? Or were they given some other purpose? 
Whereas loom weights were supposedly used for warp-weighted looms, the tablets could be applied to various looms, both vertical and horizontal. They could have also been used together with the loom weights, in the same loom, given that they supposedly had different functions (such a loom is shown by Gomes, 2003). However, it is still not easy to understand if some objects were loom weights or tablets, or even if they might have been both (depending on the weaving mechanism). Kunst (2007), based on the research from Cardoso and Roque Carreira, mentioned that, although the majority of tablets found are of baked clay, they could have also been made of lighter materials, such as ivory, or even wood. For this reason and due to the weight of some of the ceramic tablets, it is still not certain if they were in fact used as tablets in weaving.
Finally, the third type of loom elements is known as “crescents”. Even though they are easily distinguished from the two other categories because of their crescent shapes, they are also very heterogeneous: they have different angles and diameters, some being more robust than others (see Photo 1 and 2 as examples). 
So far, all the identified “crescents” are made of baked clay and have two (sometimes three) perforations, one in each end. It is because of these perforations that they are associated with weaving, but this connection is the most uncertain amongst all types of loom elements. It is also not clear why they have different shapes and what that variety means.
In Portugal, the majority of “crescents” were found in the south of the territory and the fragment found in Castanheiro do Vento is the one found northernmost in the territory. Further research may show if this is an isolated case, but the entire field related with “crescents” is yet to be fully explored. Gonçalves (1989) has suggested that they might be related with a specific weaving technique, and this idea was further developed by Diniz (op. cit), but their role inside weaving is still uncertain. 
The three categories of loom elements not only encompass diverse objects, but also inside each category, the morphological variations are noticeable. Thus, the terms should not be confused, especially in archaeological reports and publications, where a precise description of what was found is required (considering that many researchers will not be able to directly access what was found). And while some differences still need to be clarified, it was shown that the term “loom weight” is not appropriate to designate the three different types, mainly because not all would be used as weights.
Consequently, the term already suggested by Boaventura (op. cit) and used by Pereira (2010) – “loom elements” – or the one used by Costeira (op. cit) – “loom components” – should be employed. In fact, Costeira (ibid) even considered the term loom component to be broader than that of loom elements. But if both terms are not confused with the different types that have been identified, then either of them are appropriate in comparison with the term “loom weights”.
Furthermore, these terms ease the identification of materialities which might be related to weaving (in the contexts of surface surveys, for example), but whose position in a loom is uncertain. But it is still vital to distinguish the different loom elements by associating them to one of the mentioned types (and justify), or if a new type is discovered, then to create a new category. In this way, the information available will be more uniform, and easier to compare and study.

III – Gomes’ approach and the context as a methodology of study

Loom elements, when found in archaeological excavations, are immediately considered to be a reference to the past existence of weaving and of looms (even the name given to these objects reinforces the connection). This correlation was, nonetheless, questioned by Gomes (op. cit), who de-constructed the association between loom elements and weaving, pointing out that their meaning might have depended on the context in which they were included.
During the study of loom elements from the site of Castelo Velho, Gomes (ibid) identified several contexts in which they were discovered: in some, their relation with the surrounding materialities points to a possible representation of weaving, but in others, the relations pointed otherwise.  They could have been associated with the construction of structures, their maintenance and “condemnation”; they were also connected with marking and organizing spaces; and their presence together with human bones led to their interpretation as elements related with their manipulation.
The idea that they might represent weaving was not criticized. On the contrary, Gomes (ibid) explored the way communities might have dynamically dealt with their own materialities, pointing out the complexity of the archaeological processes and interpretation. The researcher also questioned the identification of contexts by the presence of specific objects of which a priori interpretations are already made, such as identifying weaving areas by the simple presence of loom elements. The fact that there is no information concerning weaving mechanisms makes this inference very difficult and somewhat problematic. Even though it might be possible that the presence of loom elements represent the past existence of weaving areas and of looms, it does not mean that this is the case in all the archaeological situations. And in order to be able to verify this correlation, Gomes (ibid) followed a contextual approach, in which the interpretation of loom elements started from the study of the relations between them and the surrounding materialities/structures.
Albeit the suggestion that the significance of archaeological materialities can depend on the situation in which they are present and were used, a question concerning loom elements lingers: if they are not interpreted as only used in a loom, how to identify the contexts in which they might have played other roles?
In his study, Gomes (ibid) challenged the way loom elements should be interpreted, and this new approach influences the study of past communities. But his results did not have the expected effect in the archaeological community. In general, loom elements continued to be unquestionably related with weaving and with the previous existence of looms in the site. Additionally, based on their distribution, some researchers, such as Gonçalves (op. cit) or Valera (1997), have mentioned the existence of areas where weaving was carried out. Even though these authors might now have other ideas concerning the identification of weaving areas, it is highlighted that distribution patterns should be carefully analyzed and interpreted. They may not always be a direct reflection of a specific moment from human occupations, but a palimpsest of several actions (either temporally and physically associated, or not), in which the materialities could have been reused and given new meanings depending on the situations.
In the research that focused on the loom elements from Castanheiro do Vento, carried out in 2010, Gomes’ contextual methodology was applied. The aim was to verify whether it was possible to relate the loom elements to weaving, but also if they might have been related with other kind of past practices. Behind this study lied the interest to explore the theoretical framework pursued by archaeologists, in particular how they interpret the remains that are found and how they deal with materialities that are, in fact, diverse and have been located in different contexts (Pereira, 2010).
    It must be pointed out that the development of the notion of “contexts” (what they are and how they should be defined), and of their role in the study of past remains, has been focused by many researchers (such as Butzer, 1980; Culler, 2006; Gamble, 2001; Hodder, 1999; Shanks, Tilley, 1987; or Thomas, 2004). It was from their arguments that the main ideas underpinning the study of the loom elements from Castanheiro do Vento were created.
As Heidegger said (referenced by Vattimo, 1996) things can only be defined in relation to other things. “Contexts” are, in fact, composed of relations that are established between the remains that are found, either in a specific place or in the overall site. Gamble (op. cit) and Thomas (op. cit) pointed out that the questions archaeologists have and the way they interpret the materialities to answer those questions, is heavily influenced by their own background and context. Because researchers are conditioned by their reality, knowledge and concepts, Jorge (2003) and Valera (2007) added that contexts will never be fully understood. And as Hodder declared, “[o]ur context frames our definition of the archaeological context.” (1992: 184). This means that since the research was carried out to “test” if Gomes’ suggestions could be verified in other sites, the loom elements were considered, on purpose, the main materiality that had to be present in the contextual relations. It is acknowledged that “contexts” have a very fluid character and that the different remains could have been given several significances not just over time, but also within the same context.
The goal was, therefore, to explore three main questions, even if definitive answers were not expected: can the situation described by Gomes be verified in other sites? Are there any relations that support the connection between the loom elements found in Castanheiro do Vento and weaving? And how were these relations identified?

IV – The loom elements from Castanheiro do Vento
Coordinates (Jorge, et al, 2005; Cardoso, 2007):
·      Latitude: 41º 03’ 49’’ N;
·      Longitude: 07º 19’ 18’’ W;
·      Altitude: 730 m 
The research that focused on the loom elements found until 2009 in Castanheiro do Vento (Guarda) is open to further revisions and updates (Pereira, 2010), because the excavations, that started in 1998, have not yet been finished. From the 110 loom elements that were identified, 96 were part of an inventory[vi] that was carried out and where a general description was provided. Besides morphological details, the loom elements were also located within the overall excavated site and information concerning the other materialities from the same layer was included. The fields[vii] of the inventory were created having in mind the desire to understand not just the relation between the various loom elements, but also the one between them and the structures. The aim was to explore their connection to weaving.
Most of the loom elements were related to layer no. 3 of the site, as were nearly all the stone structures. Considering that this layer was dated from the beginning of the III to the middle of the II millennia B.C., it is acknowledged that the materialities might relate to different moments, and that there is no way of knowing if they were used at the same time. Therefore, any suggestion concerning their use was based on the morphological study. 
The main results of the study point to a very heterogeneous group with differences in sizes, morphologies, perforations and weights. Most of the complete loom elements have subrectangular shapes and vary in terms of weights (from 97g. to 639 g.) and sizes (from 7 cm to 14,5 cm). It is probable that the loom element measuring 4,5 x 7,5 cm and weighting 97 g. could have been used as a tablet (see Image 4).
Only two of the complete loom elements are less than 2,5 cm thick, but the importance of this measurement for weaving is still not clear (possibly it would not interfere with their use, but maybe less thick loom elements could have had specific uses).
It was important to confirm that the size, quantity, shape and location of the perforations is very different, not only within one single loom element (see Photo 3), but also between them. From the 23 complete ones, three have two perforations and the rest has four, but considering all 96, the majority has signs of at least two or three perforations (only four fragments have one perforation, possibly two if they were complete). The difference in the perforation’s sizes and shapes might be due to what was used to make them, but in some cases they are not even aligned, which means that similarity and symmetry were not always required characteristics for whatever practices they were involved in.   
Considering just the fragmented loom elements, the same heterogeneity is seen, pointing to various rectangular and oval shapes with different thicknesses and sizes. Being fragmented causes a change in morphology and in weight, but it is still unclear what consequences this would have in terms of their use. One possibility is that they could still be used as loom weights, for example, having in mind that some are still heavy when compared to other loom elements (for example, there is a fragment – 2/4 of the total size – that weights 270 g., more than some of the complete loom elements).
It is also possible to use lighter (or fragmented) and heavier weights together, but in practical terms it would be easier to make new ones or to weave with similar weights, since balancing the tensions of the yarn would not be so complicated. 
From the study of the fragments, it was observed that some broke in similar places and in similar ways. And when comparing the loom elements from Castanheiro do Vento with the ones from S. Pedro (Costeira, op. cit), this was again verified (see Photo 4 and Photo 5). This might show that they were involved in similar uses, but the possibility needs further exploring.
                   From all the loom elements, only seven have traces of decoration, either on one side or on two sides. The decorations are expressed by incisions that form zigzag and perpendicular lines, but also rectangular forms. Diniz (op. cit) has suggested that the existence of decoration might not have changed the use that was given to the loom elements, but that it might have changed their status. Valera (2007) also pointed out that the linear incisions could be somehow related with the representation of looms and of weaving. However, it is unclear if this decoration would compromise the position of the loom elements, or if it would even be relevant.
Concerning the relation between the loom elements and other structures, some were scattered throughout the site, but many were found in relation with structures, materialities and even with other loom elements (it should be mentioned that none of the decorated ones were found together). However, due to the lack of information concerning their specific location in relation to those structures and materialities, it was only possible to do an in-depth study of the ones that were identified inside the sub-circular structure A (previously studied by Vale, 2003).

a) The sub-circular structure A

The sub-circular structure has an area of 6, 38 m2 (Cardoso, op. cit) and a complex micro-spatial reality that was considered a context. Four moments were identified, yet no occupation phases could be distinguished (Vale, op. cit): the oldest is related with the foundations of the structure, which were built on a stratigraphic layer of five to 20 cm. Here, some ceramic fragments were found.
The second moment is related with a small stone structure made up of granite manual grain mills and two schist slabs, all vertically placed and stuck to the ground. This created a confined space within the structure, in which 32 ceramic fragments were found. Outside this small structure, 105 ceramic fragments were identified, as well as two stone flints, the upper part of a manual grain mill, a schist slab and four complete loom elements (see Image 5).
In the third moment, the space was filled with schist slabs, with almost no space between them, and the small stone structure was left visible. When the materialities of the fourth moment were being dug, and before reaching what was later identified as the third moment, 487 ceramic fragments were found, as well as two flints, a pebble stone, three lower parts of manual grain mills and four loom elements (see Image 6).
In the last moment (the fourth) the structure was again filled with large schist slabs, with some spaces between them. This moment was interpreted by Vale (op. cit) as an intentional closure and not just some stones randomly placed. As Cardoso (op. cit) also noticed, not only was the interior space of the structure reused and reorganized several times, but the materialities point to intentional actions. For example, all the ceramic fragments were not, as far as it is possible to tell, parts of the same vases and containers, but socially significant and manipulated as fragments.
From a total of eleven loom elements, it was only possible to give an estimated location of eight (Vale, op. cit; Pereira, 2010), since there was not enough information concerning the other three. The loom elements from the second moment were complete, and due to their morphology and weight (between 401g. and 490 g.), they could have been used together in a loom, if one considers their use inside weaving. However, by the position they were found in the structure (see Image 5), their interpretation is more complex.
There is no evidence that proves or denies the use of the space for storage, yet in practical terms, why would complete loom elements be stored inside a structure that was covered on purpose? The connection between these loom elements and weaving is always a possibility, but in this case, the presence of complete objects leads to the question of why they were placed in this space and buried if they could still be used. Perhaps after they were left inside this structure their connection with weaving changed. Unfortunately, there is also no way to know if the loom elements were placed inside the structure at the same time or not.
It was between the third and the forth moments that other six loom elements were found. From these, the complete ones were located in opposite places inside the structure (see Image 6) and they not only have different weights (457g. and 344g), but also different thicknesses (2,5 cm to 3,5 cm; and 4 cm to 4,5 cm). These differences do not exclude their use in weaving, but their context does not suggest so. It is possible that they were involved in the maintenance, construction or organization of the structure’s space, or they could even have been among the filling material. In both cases, they seem to be related with practices that don’t point directly to weaving, but with interventions done to the structure itself.
The loom element near the entrance is the only one, so far, that has both sides decorated (with linear incisions), and was found close to the biggest concentrations of ceramic fragments, indicating other practices besides weaving.
The fragmented loom elements were found in close relation to the inner wall of the structure, and since the ceramic fragments seemed to have gained a new status when placed inside this structure, so could have the loom elements (one of them was too fragmented to have been reused in weaving). The possibility that they were part of a group of materialities which were considered unfit for further use does not justify the complete loom elements. On the other hand, if this moment would be considered the result of storage, how could one explain the presence of fragmented loom elements, fragmented ceramic pieces (of which only two could be connected) and only the lower parts of manual grain mills?
In conclusion, the materialities inside the structure seem to relate to diverse practices, and while some loom elements might represent weaving, others seem to be more associated to the stone structure, as parts of the architecture and of the construction of the space.

V – Loom components from S. Pedro: a typological study
            The research carried out by Costeira (op. cit) in the site of S. Pedro (Évora) included the loom components uncovered in the excavations of the years 2004-2005, 2007, 2008 and part of the ones found in 2009, totalizing 1621 objects. Because the majority (96 %) is fragmented, it is estimated that the total number of loom components might be between 514 and 964. All of them were included in a typological study, based in an inventory where technical and morphological details were given[viii]. This study was created with the intention of improving the organization and uniformity of the data related with the loom components of S. Pedro, and it was also devised as a tool that could be used by other archaeologists in other sites.
When creating a typological organization, there is always the risk to have either too broad groups or too specific, mainly because the loom components were handmade and so each has its own size. On the other hand, the categories are very useful for specifying the varieties that might have existed (even if they were considered the same) and for strengthening the comparative basis for further research works (Costeira, ibid).
As Costeira (ibid) states, loom elements were indispensable for weaving, even if they could have diverse uses depending on the technique (the morphological diversity could be related with the various techniques). The existing problems surrounding their functionality are due to both the lack of knowledge concerning weaving mechanisms and to the function being a subjective interpretation based on today’s approach to the past.
Loom elements were considered to be one of the material expressions of the changes that happened with the so-called “Revolution of the Secondary Products”[ix]. In parallel, they are also seen as the most typical materialities from settlements of the “Calcolithic” period in the Southern Iberian Peninsula (Costeira, ibid).
In contrast, Pereira (2010) has pointed out that these changes might have occurred in different rhythms and over longer periods of time, taking into account the description of loom elements in sites dated from the Portuguese Initial and Late “Neolithic” (Calado, 2001). Although these are not common finds, taking them as indicators of a certain transition should be carefully considered, particularly because some looms don’t require any of the found loom elements, and so weaving might have been known earlier than shown by the archaeological excavations.
From the 1621 loom components, Costeira (op. cit) identified ceramic tablets (38%) and “crescents” (61%).
a)     Ceramic tablets

The ceramic tablets were organized according to their shapes (rectangular, oval and hyperboloid), and when compared to the possible ones from Castanheiro do Vento, a bigger variety is noticeable: concerning the number of perforations, they go from two (one in each side), three (two in one side and one in the opposite side), to four and probably six (three in each side) (see Photo 6 and Photo 7).
 The existence of loom elements with three perforations makes the reconstruction of fragmented ones more difficult, since it is unknown whether they would have three or four perforations in total. Nevertheless, this variety may show that there were different techniques, even if the majority has two or four perforations (229 tablets have two and 237 have four).
Some of the tablets are also lighter than the ones from Castanheiro do Vento, with 10 g. to 230 g. In terms of sizes, they are very heterogeneous: rectangular tablets have widths ranging from 1 cm to 6 cm, and in the oval shaped ones, the width is from 2 cm to 7,9 cm; values for thickness go from 0,5cm to 3,5cm (rectangular shaped) and from 0,8 cm to 2,9 cm (oval shaped); and they measure between 6 cm to 13,9 cm (rectangular ones). In general, the heavier and wider tablets have four perforations that have diameters varying from 0,1 to 1,3 cm (Costeira, ibid), and which are similar to the sizes of Castanheiro do Vento.
The hyperboloid shaped tablet is considered an exception for this site: it has two perforations, a width of 3,7 cm, it is 1,3cm thick and measures 11,4 cm. It weights 98 g. More information concerning the loom elements from other sites is required to understand whether this is a unique case or not.
 Following the given figures, Costeira (ibid) reached the conclusion that the measures of the ceramic tablets from S. Pedro are within those given by various other authors, such as Diniz (op. cit) and Boaventura (op. cit).
One important aspect is that, among the tablets that were found, some have traces of perforations, which were marked on the surface but not completed (see Photo 8 and Photo 9). According to the author, if the perforations were not done correctly, then the loom element could not be used, but this mistake could be corrected by doing more perforations (Costeira, op. cit). If ceramic tablets were being used for tablet weaving and for creating certain patterns with the yarns, then the position of the perforations could be important (even though not all the perforations had to be used). In Castanheiro do Vento, there was one loom weight that had two perforations on one side leading to only one on the other side (see Photo 10). But considering that the loom weights would require different characteristics, this would not interfere with weaving.

b)    “Crescents”

In relation to the “crescents”, four different types were identified, based on the shape of their diameters (the most common are the oval shaped, then the circular shaped ones, followed by subrectangular, and by only one robust oval), and all have two perforations, one in each side[x]. Overall, the sizes obtained by measuring one end to the other, range from 6 cm to 11,9 cm, even though the majority has from 8 cm to 9,9 cm; the most common width is from1 cm to 2,9 cm, but the minimum is 0,5 cm and the maximum 4,9 cm. In terms of thickness, the figures are between 0,5 cm to 2,5 cm (the majority is 0,5 cm to 1,4 cm thick); and concerning the weights, the minimum is 21 g. and the maximum is 80 g. (Costeira, op. cit).
The perforations from the tablets and from the “crescents” have similar diameters, and in case of the “crescents”, the figures are from 0,1 cm to 1,5 cm. The exception is the “crescent” with a robust oval diameter which is similar to the ones found in other sites, such as Monte da Ribeira, Perdigões and Mercador. It is 6,2 cm wide, 3,1 cm thick and weights around 244 g. Its only intact perforation measures from 1cm to 1,5 cm. It is the only one found in S. Pedro but, according to Costeira, the stratigraphic unit from where it was found does not provide a clear context (Costeira, ibid).
            Even though the typological classification of the “crescents” followed the shape of the diameters, the variety in forms and measures is very high. Therefore, the analysis that was conducted should be extended to other sites in order to get more comparable data. From another point of view, Costeira (ibid) refers the many similarities with the “crescents” found in other sites of the region, indicating that there might have been one common conceptual basis for their production and use.
Since the use of the “crescents” in weaving is uncertain, especially the position they would be used in, it would have been useful if their angles would be measured, even if the importance of this characteristic is uncertain. 
From all the loom components discovered in S. Pedro, only four are decorated (two tablets and two “crescents”): three have linear incisions (see Photo 11) and one has a zig zag incision. All of the motifs follow geometric patterns and, similarly to what was considered for Castanheiro do Vento, it is difficult to interpret what they might represent or mean. Also, without these incisions, none of the decorated loom components would be physically distinguishable from the others. From this observation, Costeira (ibid) suggests that the decoration might have been part of an “everyday gesture” that would not interfere with their use.    
Another important aspect related with the possible use of the loom components is the wear traces they might show. For the tablets, for example, it was argued that because of their constant use, the wear traces are not easily perceived. However, Costeira (ibid) acknowledged that it is a difficult task to distinguish wear traces from traces made during its production, and that there is no concrete way of identifying them.
It was equally admitted that it is complex to give a certain function to objects of which there is no information concerning their contexts of use. Yet, in her work, Costeira (ibid) supported the relation between these materialities and weaving. This is justified by the quantity of loom components found and their similarities in terms of morphology, characteristics and perforations. It was suggested, for example, that because the oval tablets are heavier than the rest, they could have been used as loom weights, and in this case, all their perforations would be used and so they would be in a horizontal position. The “crescents” present bigger interpretation issues, but it is also suggested that they could have had a similar function as the tablets, since their form would allow twisting the yarns. Nevertheless, the study of these materialities does not give access to its context of use or to the weaving mechanism, and so ethnographic studies are still the main reference (Costeira, ibid).  
In parallel to the typological inventory that was carried out, Costeira (ibid) identified the contexts where the loom components were found, with the purpose of getting information regarding their use. In terms of chronology, both types of loom components coexist during the entire occupation of the site, according to the study of the stratigraphic units. It was noted that, similarly to other archaeological sites from south of Portugal, the tablets are in somewhat larger numbers in the first part of the III millennium, whereas the “crescents” are from more recent periods of the III millennium. However, none of the types has appeared alone, so this sequence is still not proved.
On the whole, both tablets and “crescents” were found throughout the site, and having in mind the long occupation of S. Pedro, and the successive reoccupation, conversion and use of the structures and spaces, it was nearly impossible to identified defined contexts where the use of the loom components could be verified (Costeira, ibid).
 From the spatial analysis, it was noticed that most of the loom components have been found either alone or in small groups, of five or less. Some were found in groups of ten, and the groups above this number are even rarer. This shows, according to the author, that the loom components were considered inappropriate and discarded. Regarding the bigger groups, they were mainly found inside negative structures which were filled. In one of these structures, 85 loom components were found in the different filling layers. Most of the objects are fragmented (only five were found complete), which may indicate that this is the result of an accumulation of debris, and not of remains of looms (Costeira, ibid).
Inside another negative structure, 26 “crescents” were found: because most of them were complete and relatively close to each other, there might have been a loom. In the filling layers of another big structure found in the centre of the site, 39 loom components were found. Due to the stratigraphic complexity, and to the diversity of the objects, it is improbable that all the loom components have belonged to the same loom.  But their concentration in this structure allows her to suggest that they might have been used for weaving. Likewise, if postholes were found, Costeira would regard them as possibly being the remains of structures where weaving would take place (Costeira, ibid). 
In conclusion, even though most of the loom components were considered to have been discarded, their connection to weaving was supported, and the study of some contexts allowed Costeria (ibid) to suggest that evidences of weaving might have been found. 
V – A study guide for loom elements
It became evident that much information is still unknown regarding weaving and the role loom elements would have in it. And while technical and morphological details were questioned by Costeira (op. cit), Pereira (2010) has suggested that loom elements could have taken part in other practices depending on the context in which they were involved. To further decrease the uncertainties that surround the subject, more detailed information concerning loom elements is required. The lack of data may occur due to the complexities of an archaeological dig, or because investigators aren’t aware of what kind of information is significant and don’t mention it in their reports or publications.
Diniz (op. cit) has previously suggested a study guide for what she called “loom weights”. Now, the study guide presented in Pereira (ibid) will be again explained. The purpose is not to create a set of rules, but to contribute to the uniformity of the data and to ease possible comparative studies.  Since many questions are still being raised, the creation of key points might help clarify some of the existing problems, both on practical (why the differences in morphology, why the variety in the number and size of perforations, the relation between the crescents and weaving, among others) and interpretative levels (if they do not always denote weaving, to what other practices might they be related?). 
First of all, the use of the term loom element or component, instead of “loom weight”, is highly recommended when referring to the different types of materialities related with weaving. Afterwards, loom weights, ceramic tablets, “crescents” or other types might be identified, but the confusion that the term “loom weight” brought will be avoided. By clarifying the terms, it will be easier to understand what was found and what the archeologist describe, even if initially they are uncertain whether it is a loom weight or a ceramic tablet. This is also very helpful for surface surveys, when a quick description of what was found is required.
Secondly, when studying loom elements, there are at least three basic questions that should be answered:

1.   “Where” was it discovered - the location of the loom elements is extremely important for further understanding their place in the past community’s life. The layer, square or any other information useful for locating it should be mentioned. All loom elements without the information regarding their location will not be useful for the study of the site. Coordinates should be taken and, when possible, a small draft of its location can be made in the label accompanying the object.
2.   “What” was discovered with it – for a more in depth study, it is important to know what other materials/structures were found in the same layer/ area/ context. The loom elements could have been found near a structure or inside it, amongst ceramic fragments, near a fireplace, or in many other situations. These observations will not, most probably, answer the question of what was the role of the loom elements in the different contexts, but it might give hints that, together with other data, allows for further hypothesis.
3.   In which state” was it found – this field is usually dealt with after the excavation has taken place, and the following information should be mentioned:
a)   Is it complete or incomplete, and if so, where did it break. Possibly, if many loom elements broke in the same places, this might indicate similar uses.
b)  Morphology: it is yet unknown why there is such a big variety of loom elements in morphological terms. For a good analysis, a combination of Costeira and Pereira’s studies should serve as a starting point. One important aspect to consider is the perforations (their number, location and sizes).
c)   Measures of length, width and size – these will depend on the loom elements that are being studied. Since they are handmade, it is expected that the measures may not be precise, especially since they also depend on how and where they are taken from (the middle, the bottom or upper parts…). Even what is considered to be the width and the length may differ. One possible way is to designate them as the biggest side and smallest side, or to choose other criteria, just as long as they are explained. In the case of the crescents, their angle should be measured when possible, mainly because of their diversity. As these materials are still not well known, all the information is useful. Again, both the studies of Costeira and Pereira’s are good guiding tools.
d)  Weight: even when many researchers considered all loom elements to be loom weights, their weight is many times not mentioned. It is, however, what may help understand if, within one context, the loom elements could have been used together, or if they are too different.
These are some of the key ideas that should be considered when digging and studying contexts in which loom elements are found. The information that is registered will also depend on the aims of the researchers, since more information could be given (such as the clay composition and treatment, or decorative motifs and styles). Nevertheless, the most important aspect is that all information should be shared by the researchers as to provide a ground knowledge that will deepen the study of the loom elements.
VI – Final remarks
Throughout this paper it has been shown that the knowledge concerning materialities identified as loom elements is still not enough to understand their relation to weaving and the roles they might have had in it. Since they are considered to be typical finds related with the period between the end of the IV and the end of the III millennia BC (in Portugal), the interpretation of loom elements will influence the understanding that is made of the archaeological sites and of the past remains. Within this context, the results from two different approaches concerning their interpretation have expanded the way loom elements can be studied, but also shown that this is still a problematic topic.
From one side, Pereira (2010) continued the deconstruction started by Gomes (op. cit) and confirmed that, in Castanheiro do Vento, the connection between the loom elements and weaving is not always evident in the situations in which they were found. The relation is not being denied, but it may not always be the most evident one. The “context” was used as a method to explore the fluid character that materialities, such as loom elements, could have depending on the situations they were involved in. The existence of loom elements is usually related with settlements, but because they have been found in relation to human bones in CasteloVelho (Gomes, ibid), the idea that they could have been connected to other practices is strengthened, even if they were once used to weave. What these other practices are, besides being related with the construction of spaces, is still open to debate. As a suggestion, it would be valuable if other researchers were aware of this hypothesis, so that certain aspects may be noticed and new data could be used either to support or to refute it.
On the other hand, Costeira’s study (op. cit) of the loom components of S. Pedro has shown that even though they are very diverse, they still have many characteristics in common. Their interpretation as loom components used in weaving is justified by their attributes, such as the shape, dimension and perforations. Still, Costeira pointed out that it was not possible to confirm if the tablets and “crescents” found are loom components. This is due to the variety of contexts in which they were identified, being the majority related with debris contexts, which makes the task of understanding their use more difficult. Even so, these materialities were taken as related to weaving and as important finds connected with the so called “Revolution of the Secondary Products” (Costeira, ibid). However, both the term, and all that it encompasses, also needs to be further researched.
Several questions remained unanswered and provide future investigation topics: how to explain the different sizes, location and number of perforations? Why is there such a big morphological variety and could it have been linked with different weaving techniques? Also, could different loom elements, with different weights, be used together? What would be the position of the loom elements in a loom?
Most importantly, by comparing both studies, it is noticed that the amount of loom elements in the sites is very different. What does this mean? In fact, how many loom elements are needed to support the existence of weaving in a settlement? Although it is a very important question, it is nearly impossible to answer it, considering that loom elements could vary according to what was being woven, for example.
Other future studies could relate to the fact that some loom elements from Castanheiro do Vento and from S. Pedro show similarities in the way they are fragmented. This might indicate that whatever they were being used in would create certain breaking patterns, but more comparisons need to be made. It is also important to further investigate the relation between “crescents” and weaving, having in mind the need for angle measurements and contextual analysis.  In parallel, more information is also required if the decoration motifs are to be fully studied: the contexts need to be well registered and it would be useful to compare the motifs of the loom elements with the ones from the other materialities of the same sites.
            Taking all of this into consideration, it can be said that this is a topic that still needs much more exploring. For that reason, a study guide was suggested and, if employed, will provide some of the needed data. 
VII – Reference List
BARBER, E. J. W. (1991) Prehistoric textiles: the development of cloth in the Neolithic and Bronze Ages with special reference to the Aegean. United Kingdom: Princeton University Press. ISBN 0-691-00224-X
BOAVENTURA, Rui (2002) O sítio calcolítico do Pombal (Monforte): uma recuperação possível de velhos e novos dados [Online]. Portugal: Instituto Português de Arqueologia. Trabalhos de Arqueologia 20. ISBN 972-8662-03-3. Available at: [Last accessed on the 15th of April, 2012].
BUTZER, Karl W (1980) Context in archaeology: an alternative perspective In Journal of Field Archaeology [Online]. University of Boston: JSTOR. Vol. 7, nº 4; Pp. 417-422. Available at: [Last accessed on the 15th of April, 2012].
CALADO, Manuel (2001) Da Serra d'Ossa ao Guadiana. Um estudo de pré-história regional [Online]. Portugal: Instituto Português de Arqueologia. Trabalhos de Arqueologia 19. ISBN 972-8662-02-5. Available at: [Last accessed on the 15th of April, 2012].
CARDOSO, João C. Muralha (2007) Castanheiro do Vento (Horta do Douro, Vila Nova de Foz Côa) – Um recinto monumental do IIIº e IIº milénio a.C.: problemática do sítio e das suas estruturas à escala regional. University of Oporto. Unpublished PhD dissertation.
COSTEIRA, Catarina Isabel dos Reis (2010) Os componentes de tear do povoado de S. Pedro (Redondo, Alentejo Central), 3º milénio a.n.e. University of Lisbon. Unpublished Master Thesis. Available at:
CULLER, Jonathan (2006) On deconstruction: theory and criticism after Structuralism. 1st edition in 1983.Cornell University: Routledge. ISBN 0-415-04555-X.
DINIZ, Mariana (1994) Pesos de tear e tecelagem no calcolítico em Portugal In Trabalhos de Antropologia e Etnologia, 1st Congress on Peninsular Archaeology. Oporto: Sociedade Portuguesa de Antropologia e Etnologia,vol. 3-4. Actas IV. Pp. 133-149. ISSN: 0304-243X.
FORBES, R. J. (1987) Studies in ancient technology. Leiden/New York: E. J. BRILL. Vol. IV. 2nd edition. ISBN 90 04 08307 3
GAMBLE, Clive (2001) Archaeology the basics. Routledge. ISBN 0-203-15804-0 (ebook)
GEIJER, Agnes (1982) A history of textile art: A selective account. London: Pasold Research Found in association with Sotheby Park Bernet. ISBN 0856670553
GINER, Carmen Alfaro (1984) Tejido y cestería en la Península Ibérica: historia de su técnica e industrias desde la prehistoria hasta la romanización. Madrid: Instituto Español de Prehistoria. ISBN 84-00-05710-4
GOMES, Sérgio (2003) “Contributos para o estudo dos “pesos de tear” de Castelo Velho de Freixo de Numão (V.N. Foz Côa): exercícios de interpretação do registo arqueológico.” University of Oporto. Unpublished Master Thesis.
GONÇALVES, Vítor S. (1989) Megalitismo e metalurgia no Alto Algarve Oriental: Uma aproximação integrada. Lisbon: Instituto Nacional de Investigação Científica. Vol. 1. ISBN 972-667-109-4
HODDER, Ian (1992) Theory and practice in archaeology. Routledge. ISBN 0-203-67382-4 (e-book)
HODGES, Henry (1989) Artifacts: An introduction to early materials and technology. London: Duckworth. 3rd edition
HOFFMAN, Marta (1964) The warp-weighted loom. Robin and Russ Handweavers.
JORGE, Vítor Oliveira (2003) Das sete vidas dos objectos In Revista da Faculdade de Letras da Universidade do Porto. Oporto: Faculty of Arts, Ciências e Técnicas do Património. 1st Series, vol. 2. Pp. 843-864. ISSN 1645-4936.
JORGE, Vítor Oliveira; CARDOSO, João Muralha; VALE, Ana; VELHO, Gonçalo;
PEREIRA, Leonor (2005) Sítio Pré-Histórico de Castanheiro do Vento (Horta do Douro, Vila Nova de Foz Côa): principais conclusões das escavações de 2005 In Portugalia. Nova Série. Vol. XXVI. Pp. 41- 52
KUNST, Michael (2007) – Zambujal (Torres Vedras, Lisboa): relatório das escavações de 2001 [Online] Revista Portuguesa de Arqueologia, Dialnet, vol. 10, no.1. Pp. 95-118. ISSN: 0874-2782. Available at: [Last accessed on the 15th of April, 2012].
PEREIRA, Mariana Pinto Leitão (2010) “Pesos de tear” e “elementos de tear” na Pré-história Recente Portuguesa: repensar o processo arqueológico. University of Oporto. Unpublished Master Thesis. Available at:
PEREIRA, Mariana Pinto Leitão (2009) Weaving in the III millennium B.C. – Contribution for a future archaeological experiment In II Jornadas de Jóvenes en Investigación Arqueológica. Madrid: OrJia. Vol. 2. Pp. 517-521. ISBN 978-84-7956-094-2.
SHANKS, Michael; TILLEY, Christopher (1987) Social Theory and Archaeology. United Kingdom: Polity Press. lSBN 0 7456 0184 7
THOMAS, Julian (2004) Archaeology and Modernity. London: Routledge. ISBN 0-203-49111-4 (e-book)
VALE, Ana (2003) Castanheiro do Vento: Contributo para o estudo dos resultados das primeiras campanhas de trabalhos (1998-2000). University of Oporto. Unpublished Master Thesis
VALERA, António Carlos (1997) O Castro de Santiago (Fornos de Algodres, Guarda): Aspectos da calcolitização do Alto Mondego Lisbon: Câmara Municipal de Fornos de Algodres.
VALERA, António Carlos (2007) – Dinâmicas locais de identidade: estruturação de um espaço de tradição no 3º milénio a.C. (Fornos de Algodres, Guarda).Portugal: Município de Fornos de Algodres /Terra de Algodres
VATTIMO, Gianni (1996) Introdução a Heidegger. Lisbon: Instituto Piaget. ISBN 972-771-049-2

[i]; Young Researcher of CEAUCP
[ii] Cardoso, 1994; Gonçalves 1989, 2003; Kunst, 2007; Rocha, 2001; Schubart, Ferreira, Monteiro, 1969; or Soares, Silva, 1987.
[iii] Most commonly known as “Calcolithic” and “Recent Pre-history”
[iv] Master Thesis completed in the University of Oporto, and entitled “Pesos de tear” e “elementos de tear” na Pré-história Recente Portuguesa: repensar o processo arqueológico” («Loomweigths» and «loom elements» in the Portuguese Recent Prehistory: rethinking the archaeological process).
[v] Master Thesis completed in the University of Lisbon, and entitled “Os components de tear do povoado de S. Pedro (Redondo, Alentejo Central), 3º milénio a.n.e.” (The loom components from the settlement of  S. Pedro (Redondo, Alentejo Central), 3rd millennium BCE
[vi] From the 110, three fragments went together with other three loom elements; and eleven were not found or did not have the basic information required for the study, such as the layer and the exact location where they were found. This is why only 96 were included in the inventory.
[vii] The inventory was composed of the following fields: location, photograph; a drawing of all the sides and showing the location of the perforations; other materialities found in the same layer; if the loom element was complete or not; dimensions of the width, length and thickness; morphology; number of visible perforations and their size; clay composition and predominant colours; traces of fire; surface treatment; and decoration description.  
[viii] The fields of the inventory include the identification (the stratigraphic unit, the number and the state of the loom elements); the morphology; metric characterization; the number of perforations and sizes; decorative motifs, location and  technique; traces of uses; and other complementary information
[ix]  This term, and the changes it refers to, should be carefully considered and further explored
[x]  Because a “crescent” with two perforations in one side was identified by Calado – 1995; 2001 – Costeria decided to create a second category as to allow for further types to be added to her typology. So far, Calado’s “crescent” is the only one with these characteristics.


VII - Images

Image 1: Example of a warp-weighted loom. The numbers point to: (1) the main rod where the warp was attached to; (2) the vertical rods that support the upper one; (3) the rod where yarns from the warp are wrapped (it is the movement of this rod that allows the yarn to be woven. The weft yarn is inserted in the space created by the movement of the rod, which is shown in the images of the left); (4) the rods that separate the yarns; (5) and the loom weights, that help stretch the yarns, and keep their tension equal (image taken from Forbes; 1987: 209).

Image 2One context from Castelo Velho, where different loom elements were found (image adapted from Gomes, 2003).

 Image 3: How ceramic tablets, with two perforations, could be used for twisting yarns and for creating a “twisted” fabric (image taken from Hodges, 1989: 136).
Image 4Possible tablet from Castanheiro do Vento. It weights 97 g. (image drawn by Pereira, 2010).

Image 5Location of the loom elements found in the second moment (image taken from Pereira, 2010, Annex VI)

 Image 6Location of the loom elements found between the third and fourth moments (image taken from Pereira, 2010, Annex VI)

Photo 1 and Photo 2“Crescents” from S. Pedro (photos taken from Costeira, 2010: 98, 100).

Photo 3Different location and format of perforations in one loom element (picture taken by Pereira, 2010).

Photo 4 and Photo 5loom elements from Castanheuro do Vento (left) and S. Pedro (right) showing similar breaking patterns. (photo 1 taken by Pereira, 2010; photo 2 taken from Costeira, 2010: 97, Annexes).

Photo 6 and Photo 7: Tablet fragment with three perforations on one side, and another with a total of three perforations (pictures taken from Costeira, 2010: 95, 96, Annexes)

Photo 8 and Photo 9Tablet fragments with remains of two incomplete perforations (pictures taken from Costeira, 2010: 92, 97, 

 Photo 10: Loom weight from Castanheiro do Vento, where two perforations lead to only one perforation in the other side (photo taken by Pereira, 2010).

Photo 11Decorated tablet from S. Pedro (photo taken from Costeira, 2010: 103, Annexes)



Lídia Baptista*
Sérgio Gomes**

Abstract: In this article we will discuss the architecture of “negative structures” in Alentejo (South of Portugal) using two case studies: Vale das Éguas 3 and Horta do Jacinto. In this discussion, we will take the architecture of negative structures as a practice within people create a dialogue between the “pit” and “its fill”, recreating the space and the place of things. 

Keywords: Negative Structures; South of Portugal; Late Prehistory 

Resumo: Neste artigo discutimos a arquitetura das estruturas em negativo do interior alentejano (Sul de Portugal) a partir de dois exemplos: Vale de Éguas 3 e Horta do Jacinto. Nesta discussão, a arquitetura é compreendida enquanto uma prática que cria um diálogo entre as estruturas em negativo e os seus enchimentos, no qual é recriado o espaço e o lugar das coisas. 

Palavras-chave: Estruturas em Negativo; Sul de Portugal; Pré-história Recente

In the last five years we have been working at Alentejo (in the south of Portugal) digging several prehistoric sites (Baptista and Oliveira 2008; Baptista 2010; Baptista and Gomes 2012a). These  sites have been identified due the construction of a water pipeline promoted by EDIA SA. Most of the contexts are negative structures. These kind of features are traditionally interpreted in terms of the contents of their fills: as domestic discard areas (whenever there are sherds, bones or other kind of fragmented things); as storage structures (in the examples where there are seeds); or as burial architecture (in those cases where there are human skeletons) (Baptista and Oliveira ibid.). Regarding these interpretations there is a dichotomy between something that is fill (from which we define a function to the structure), and the pit itself which is understood as a kind of architecture. In this paper, we aim to discuss such dichotomy, arguing that when we look at these features within a conceptual framework which takes architecture as a practice and not as a “building record”, the pit and its fill are both elements of architecture (McFadyen 2006). 
We choose two examples in order to do the discussion. The first example is a 3rd Millennium structure identified at the site Vale de Éguas 3 (Salvador, Serpa, Beja); such example, as we will see, allow us to make a link between architecture and pottery fragmentation practices. The second example is a 2nd Millennium structure identified at the site Horta do Jacinto (Beringel, Beja); this second example allow us to discuss the links between the deposition of human and animal corpses and the building of space inside the “pit”. In this way, the discussion that we will present, instead of looking for a function through the analysis of the fills, will focus on the practices that those fills entail. In other words, the analysis will look for the “web of actions” (Jorge 2011) within these structures were constructed. 

1. Vale de Éguas 3

Vale das Éguas 3 is a 3rd millennium site located in the south of Portugal at Salvador (Serpa, Beja) (Figure 1). As mentioned, it was identified, in 2009, during the construction of a water pipeline1 promoted by EDIA SA. These works allowed the identification of six pits containing different kinds of materials (Figures 2, 3 and 4) (Cunha, Baptista and Gomes 2010). These contexts suggest different practices for the fragmentation of pottery, and how sherds can be related back to the architectural practice of pit-making itself (Baptista and Gomes 2012b). In this group of structures, Structure 2 stands out from the rest of the group because of its high number of sherds and its contexts2. Considering this, we will focus our analysis on the case of Structure 2. In such analysis we aim to consider some issues related to fragmentation and deposition of materials during Late Prehistory (Chapman 2000, Pollard 2001, Jones 2002; Garrow, Beadsmoore, Knight 2005; Jorge 2005, Valera 2010, McFadyen in press, for example).

Structure 2

The upper fill of Structure 2 was a clay deposit [200] with small stones clustered at its top. After we dug it, we identified [201], a similar deposit but lighter in colour. During its excavation, we defined an horizontal plane regarding the occurance of a half of a "prato de bordo espessado” (Figures 5 and 6). The remaining deposits filling the pit [202, 203 and 204] were very similar, and their individualization was connected with colour variations. In these deposits, several sherds from the same vessel were recovered. Note that, unlike the "prato de bordo espessado" identified in [201], the fragments were scattered throughout the deposits (Figure 7).
The kind of fragmentation, and the integrity of the sherds, is very different. On the one hand, the “prato de bordo espessado” is represented by half of the vessel, and during its deposition the integrity of this fragment was preserved in a specific deposit of soil. On the other hand, the spread of the fragments of the vessel from [202, 203 and 204] are apparently random. However, during the refitting of the container, we could see that within the apparent randomness of the distribution of these fragments, there was one aspect which ordered the distribution: there was a tendency to concentrate the sherds of the base of the vessel in [202] and the rest of vessel in [203 and 204] (Figure 8). Thus, it is noteworthy that in the same structure, a common practice of fragmentation contains two variants that express distinct dynamics of the same practice.
In the case of the “prato de bordo espessado”, the integrity of the fragment (and the formalization of the level at which it occurs) suggests a scenario in which the process of fragmentation may have participated in a social dynamic that Chapman (2000) calls enchainment. Such an interpretation juxtaposes the fragmentation process to the production dynamics of a social bond, in which the fragmentation of objects becomes a way of multiplying the invocation of people, places and moments. A biography of the fragment, and not the object, is thus one of the protagonists of this network of participants in the chain. Chapman (ibid.) distinguishes the process of chaining from another process in which the representation of the object is constructed by the “accumulation” of meanings. Objects, and not their fragments, are transformed into elements that produce a network of relationships. Within this the unity of the object is the way to bring in other participants of the social network. This distinction, proposed by Chapman (ibid.), becomes especially relevant in the case of the vessel from [202, 203 and 204] since their occurrence articulates two possibilities. Indeed, if we consider that the spread of ceramic fragments is due to post-depositional phenomena of a vessel put in an inverted position, Structure 2 brings together unities whose biographies are constructed and socialized in a very different way. However, considering that the spread of the fragments corresponds to different moments of filling, we can multiply the scenarios of its process of fragmentation. In this case, a process whereby there is the multiplying / fragmenting of a vessel, divided according to the order of its configuration (the base parts in one place, and the upper body parts in another), in different contexts which further emphasizes them not as a unit, but as things that were once a part of the same element. Regarding all these possibilities, Structure 2 allow us to highlight the complex relationship between deposition, architecture and fragmentation.

2. Horta do Jacinto

Horta do Jacinto is a 2nd millennium site located in the south of Portugal at Beringel in Beja (Figure 1). It was identified in 2007 during the construction of a water pipeline3 promoted by EDIA SA. These works allowed the identification of two pits (Figures 9 and 10) containing different kinds of materials (Baptista, Cunha and Gomes 2010; Baptista, Gomes and Costa 2012; Ferreira 2008). In this paper we will focus on Structure 1, where we found the “deposition” of a human and an animal. By looking at these depositions we will discuss how these two “materials” entailed different practices, and how this was connected with contrasts in the organization of space inside the pits.

Structure 1

Briefly, we can systematize the filling of Structure N.º 1 in the following way:
1- a stone level sealing or capping the other fills (Figure 11);
2 - a deposit of soil that has been shaped reproducing the configuration of the pit in order to receive a 9-12 years old individual; the individual has been placed sitting on the ground of this structure made with the deposit (Figure 12);
3 - a horizontal stone level built at the base of the structure, in which has been placed a 9-16 months Sus sp5 (Figure 13);

4 - The internal organization of the pit is connected with the configuration of how the bodies were deposited (Figure 14). Considering this we should highlight that: 

  • the animal's body is in a level below the plane of the human body; there is an opposition between the horizontality of the animal body and verticality of the human one;
  • the architectural device where the animal was placed was built with stones, the structure in which the human body was deposited was constructed with sediment;
  • the horizontality of the stone level contrasts with the verticality of the pit, unlike the space where the human body was deposited which mirrors the contour of the pit, adjusting space to the dimensions of the body.

By considering the relationships between what is put inside the pit and its internal configuration, it seems that once we look at the feature as architecture then the shape of the pit is as important as what is inside of it, and not just as a way to think about a functionality but to understand it as an architectural device.    
We would like to add another aspect regarding the materials and refitting. Indeed, during the study of the materials, we notice that there was only fragments of artefacts. By analyzing the refits, it would seem that the majority of the kinds of fragmentation could be explained by post-depositional phenomena. However, there is one artefact that entails another kind of discussion. It is between two pieces of a grinding stone (Figure 15): one found at the stone level that closes the structure (at the top), and another from the stone level at the base of the pit. In this case, fragmentation can be understood as a process within an architectural practice. A process that gives sense to the process of filling the pit. 
To recap, Structure 1 is constituted from actions that bring together soil, sculpted shape and bodies. Furthermore the distribution of fragments of objects keep memory at work in these spaces: they create a material tension between how things were and how things are now. Architecture here is not just a building activity; architecture constructs other possibilities for how humans, animals and artefacts relate to each other in the world.

In the case of Vale de Éguas 3 we discuss how fragmentation and deposition of pottery connects with architecture. In the case of  Horta do Jacinto we discuss how different bodies and artefacts entailed different space organizations. In both cases we can see that the analysis of the fill, rather than a way to explain a function of a structure, can be a way to ask about how space is constructed. A way to look for a “web of actions” (Jorge 2011) within the meaning of the structures was constructed. By analysing these structures as proposed by L. Mcfayden (2006), we are moving from an inquiry which privileges the creation of a discourse based on its functionnality. Instead, we are looking for the multiple ways in which people create space. In this way, we are trying to develop an approach which, rather than a “building record” exercise, tries to go further in the understanding of how architecture is a practice within people negotiate their worldly conditions of living.

This article is the result of two posters we presented at the Theoretical Archaeology Group held at Bristol in 2010. We would like to thank Stephanie Wynne-Jones and Joshua Pollard (University of Bristol) for giving us the opportunity to present our work. We also want to present our deepest gratitude to all our colleagues and friends which work with us making this paper possible: Lesley McFayden, Susana Oliveira Jorge, Claúdia Costa, Lurdes Oliveira, Nelson Vale, Rodry Mendonça, João Molha, Rui Pinheiro, Zélia Rodrigues, Bárbara Carvalho, Sara Luz, André Saraiva e Francisco Barros.  


Baptista, L. (2010). “The Late prehistory of the watershed of the Ribeiras of Pisão and Álamo (Beja, South Portugal): a research program”, Journal of Iberian Archaeology 13: 69-79.

Baptista, L. Oliveira, L., (2008) “Análise dos modelos interpretativos dos recintos de fossas e/ou fossos da Pré-história Recente do Sudoeste Peninsular”, comunicação apresentada no 1º Congresso Português de Arqueologia Empresarial, Lisboa, 2008.

Baptista, L., Cunha, L., Gomes, S. (2010). Bloco de Rega do Pisão. Intervenção Arqueológica em Horta do Jacinto, Arqueologia e Património Lda., Beja.

Baptista, L., Gomes, S. (2012a). Minimização de Impactes sobre o Património Cultural decorrentes da execução do Bloco de Rega de Brinches-Enxoé – Fase de Obra. Relatório Final Global, Arqueologia e Património Lda., Beja.

Baptista, L., Gomes, S. (2012b). “Vale de Éguas 3 (Salvador, Serpa): fragmentação de materiais cerâmicos e enchimento das estruturas”, Actas do V Encontro de Arqueologia do Sudoeste Peninsular, Almodôvar, Coordenação de Manuela de Deus, Almodôvar: Município de Almodôvar: 597-603. 

Baptista, L., Gomes, S., Costa. C. (2012). “As dinâmicas de deposição e construção no sítio pré-histórico de Horta do Jacinto (Beringel, Beja)”, Actas do V Encontro de Arqueologia do Sudoeste Peninsular, Almodôvar, Coordenação de Manuela de Deus, Almodôvar: Município de Almodôvar: 585-595. 

Chapman, J. (2000). Fragmentation in Archaeology: people, places and broken objects in the Prehistory of South-Eastern Europe, London, Routledge.

Cunha, L., Baptista, L., Gomes, S. (2010). Bloco de Rega de Brinches-Enxoé. Intervenção Arqueológica em Vale de Éguas 3, Arqueologia e Património Lda., Beja.

Ferreira, T. (2008). Horta de Jacinto, Bloco de Rega do Pisão, Beja, Estudo de Análise Antropológica,  Styx Antropologia.

Garrow, D., Beadsmoore, E., Knight, M. (2005). “Pit clusters and the Temporality of Occupation: an Earlier Neolithic Site at Kilverstone, Thetford, Norfolk”, Proceedings of the Prehistoric Society 71: 139-157.

Jones, Andrew (2002). Archaeological Theory and Scientific Practice, Cambridge, Cambridge university Press.

Jorge, S. O. (2005). O Passado é Redondo. Dialogando com os sentidos dos primeiros recintos monumentais, Porto, Edições Afrontamento.

Jorge, S. O. (2011). “Castelo Velho, Alto Douro: The Site as a Web of Actions”, comunicação apresentada no Workshop - A Joint Consideration of the Study of Prehistory in Britain and Portugal: Towards a Critical Understanding of Time, Space, Practice and Object in the Prehistoric Past, em Janeiro de 2011, na FLUP.

McFadyen, L. (2006). “Material culture as architecture - Neolithic long barrows in Southern Britain”, Journal of Iberian Archaeology 8: 91-102.

McFadyen, L. (forthcoming) “The breakage and post-breakage histories of the pottery at Castelo Velho”, in Jorge, S. O. (ed.) Castelo Velho. Porto: Ministério da Cultura, Portugal.

Pollard J. (2001). “The aesthetics of depositional practice”. World Archaeology, vol.33(2), pp.315-333, Stable URL:, accessed 24/01/2009.

Valera, A. (2010). “Marfim no recinto calcolítico dos Perdigões: “Lúnulas”, Fragmentação e ontologia dos artefactos”, Apontamentos 5, NIA: 31-42. 

Valera, A., Costa, C. (2010). “Animal paws in funerary contexts in southern Portugal and the segmentation problem”, poster apresentado ao 11th International Council of Archaeozoology, Paris, 23-28 August 2010.  

Sem comentários: