Introduction
The engraved stone plaques of Late Neolithic and Copper Age Iberia (3200–2200 bc), with their delicately incised geometric designs and ‘eyes’ that occasionally peer out from them, have enchanted prehistorians since the 1800s. The Portuguese prehistorian and medical doctor Augusto Filippe Simões (1835–1884) wondered whether they might be ‘amulets or insignias or emblems or cult objects’ (Simões, Reference Simões1878: 53). Recently, the plaques and other idol-like objects were the focus of the international exhibition Mobile Images of Ancestral Bodies: A Millennium-Long Perspective from Iberia to Europe (Bueno Ramírez & Soler Díaz, Reference Bueno Ramírez and Soler Díaz2021). The plaques’ aesthetic and intellectual appeal is enduring.
Typically found as grave goods in tombs in south-western Iberia (Figure 1), the plaques are hand-sized objects (150 mm) most often made from slate. They are primarily found in southern Portugal but also known in south-western Spain. In rare occasions, they are found in settlements (Andrade et al., Reference Andrade, Costeira, Mataloto, Collado Giraldo and García Arranz2015). More than 1800 plaques from over 300 sites have been incorporated in the Engraved Stone Plaque Registry and Inquiry Tool (ESPRIT) online database (https://iberian.its.uiowa.edu; Lillios, Reference Lillios2021), which is updated as plaques are excavated, published, or otherwise made known. At the time of writing, ESPRIT contained records for 1826 plaques.
Figure 1. Distribution of engraved plaques and key sites mentioned in text. 1: Comenda da Igreja; 2: Freixa; 3: Escoural; 4: Anta Grande do Zambujeiro; 5: Perdigões; 6: Cebolinho 1; 7: Barcarrota; 8: Pijotilla.
The plaques participated in the social world of Late Neolithic and Copper Age agrarian communities who lived in fortified settlements, aggregated in ditched enclosures, and constructed monumental collective tombs for their dead (Lillios, Reference Lillios2008). These communities maintained long-distance trade connections within Iberia as well as with people in North Africa, western Europe, and the central Mediterranean. They crafted diverse objects in clay, stone, copper, and gold, some likely produced by specialists. The presence of exotic and finely crafted goods, variability in type and scale of mortuary architecture, and differential forms of treating the dead suggest that some form of social ranking existed. The clearest evidence comes from the opulent graves at Valencina de la Concepción (Sevilla, Spain). There, at the Montelirio tholos, at least fifteen females (or likely females) were found covered in cinnabar and with ivory and amber objects, rock crystal arrowheads, and hundreds of thousands of beads that were once part of robes (Fernández Flores et al., Reference Fernández Flores, García Sanjuán and Díaz-Zorita Bonilla2016). Near the tholos, the so-called Ivory Lady was found with, among many items, an African elephant tusk (Cintas-Peña et al., Reference Cintas-Peña, Luciañez-Triviño, Montero Artús, Bileck, Bortel and Kanz2023). No plaques were found at Montelirio, so if plaques were markers of social identity or esteem, they were not the only objects that did so.
For this period in Iberia, archaeologists debate the degree of ranking of individuals, families, or lineages, whether they experienced differential access to food, other important resources, or power, and how social distinctions were perpetuated. In bioarchaeological studies carried out by Díaz-Zorita (Reference Díaz-Zorita Bonilla2017), differences were found between the health status and diet between individuals interred in megaliths and those in non-megalithic structures. She found that people buried in megaliths consumed more protein than those in non-megalithic tombs. This may suggest that megaliths, with their high labour involvement, materialized certain elite identities. Thus, although many archaeologists view megaliths in Iberia as expressions of collectivism among groups without hierarchy, or at least hierarchies that did not translate into exploitation (Díaz-del-Río, Reference Díaz-del-Río2021), the bioarchaeological data suggest that some groups or individuals may have distinguished themselves or experienced their social world with special privileges. And indeed, there may have been different social logics of hierarchy (achieved vs inherited status) at play in different regions or which changed over time. This article considers the role that the engraved plaques may have played in creating the social world of ancient Iberians.
Archaeologists have proposed numerous interpretations for the plaques’ iconography and function. The most widely held idea, or at least the one with the greatest longevity, is that they represented a deity, specifically a Mother Goddess (Gimbutas, Reference Gimbutas1991; Gonçalves, Reference Gonçalves1999; Andrade, Reference Andrade, Collado Giraldo and García Arranz2015). A second is that they were images of ancestors (Bueno Ramírez, Reference Bueno Ramírez2010). A third is that they were heraldic emblems for a class of the dead (Lisboa, Reference Lisboa1985; Lillios, Reference Lillios2002, Reference Lillios2008). Recently, Negro et al. (Reference Negro, Blanco, Rodríguez-Rodríguez and Díaz Núñez de Arenas2022) have proposed, using experimental studies, that the plaques were not primarily funerary offerings but rather the art of children who were representing owls. If, however, children had made these plaques, those plaques with owl-like imagery should be much more common, given that children tend to be the largest demographic in any society. In fact, owl-like plaques make up only about four per cent of all plaques. Given the scant evidence for this idea, it will not be considered further here.
The plaques were part of an extensive and variable class of objects made from bone, stone, ivory, and clay that ancient Iberians used in their ritual lives (Bueno Ramírez & Soler Díaz, Reference Bueno Ramírez and Soler Díaz2021). There is no reason to think that all the plaques (or these ritual objects) had one function or meaning given their formal and spatial variability, long period of use (approximately 700 years), biographies that sometimes included modification, and the large area in which they are found, roughly 300 × 400 km, or 120,000 km2. Eight types have been identified based on their form, composition, number of perforations, number of sides engraved, and design elements (Figure 2). An additional type shows signs of modification (recycled) and tends to be found in settlements (Gonçalves et al., Reference Gonçalves, Pereira and Andrade2003; Lillios, Reference Lillios and Lillios2010). These recycled plaques are not included in our analyses since their original form or type is not always identifiable.
Figure 2. Plaque types and frequency. a: Cebolinho 1; b: Olival da Pega; c: Vega del Peso; d: Comenda da Igreja; e: Cueva da la Mora; f: Marquesa; g: Idanha-a-Nova; h: Valencina de la Concepción.
Sources: a, b: Leisner & Leisner, Reference Leisner and Leisner1951: pl. XXXIV, 16; pl. XXVIII, 29; c, d, e, f: Leisner & Leisner, Reference Leisner and Leisner1959: pl. 53, 1, 4; pl. 27, 1, 66; pl. 53, 7, 11; pl. 4, 5, 8; g: Leisner, Reference Leisner1998: pl. 75, 2 [a-g: reproduced by permission of Deutsches Archäologisches Institut]; h: Fernández Gómez & Oliva Alonso, Reference Fernández Gómez and Oliva Alonso1980: fig. 6, reproduced by permission of author.
When closely associated with articulated skeletons, plaques are found on the chest or by the side of the body. The sex and/or age of associated individuals could only be determined in a few instances; in these cases, they were found with females and males (Table 1). Not all the dead were buried with a plaque. When the minimum number of individuals is known, it is almost always larger than the number of plaques. The tomb of Santa Margarida 3 (Gonçalves, Reference Gonçalves2003) is unusual in that it has almost the same number of plaques as individuals, with twenty-two plaques and an MNI of twenty-five (Gonçalves, Reference Gonçalves2004: 61). The plaques’ dating is challenging given that they are often found in disturbed commingled contexts, and the soils in which most are found, in the Portuguese Alentejo, are acidic. The few dates of associated remains point to their use between 3200 and 2500 cal bc (Gonçalves, Reference Gonçalves, Bueno Ramírez and Soler Diaz2021). Since the Alentejo is also where the greatest variety of plaque types is found and has an abundance of slate outcrops, it seems reasonable to assume that plaque-making began there. Unfortunately, we know little about the production of the plaques, other than at the Águas Frias site on the Guadiana River between Portugal and Spain, where, in 2003, plaque roughouts were found, some sketchily engraved and some finished (Gonçalves, Reference Gonçalves2013).
Table 1. Plaques found with sexed individuals.
That only a minority of the plaques (4 per cent) have clear biomorphic traits, such as ‘eyes’ (or oculi) or a ‘nose’ (or ‘beak’), presents a challenge to the notion that all plaques represent a deity, such as a Mother Goddess. These plaques include the Biomorph Simple and Biomorph Whiskered types (Figure 2 g and h). The overwhelming majority (70 per cent) are of the Classic type— the focus of this article. Classic plaques have a ‘ top’ and a ‘base’ separated by a horizontal line or bands (Figure 3). The bands are filled with hachures or small triangles, or are unfilled. The top can have straps (horizontal or vertical) or other motifs, such as inverted ‘v's, rays, and a collar. In rare cases (n = 10), Classic plaques have ‘eyes’ or oculi (Figure 3g), like the Biomorph plaques. The base is typically decorated with six design motifs, which can be found on their own or in combination with other motifs, and are most often organized along horizontal registers. These motifs include triangles, checkerboard, vertical bands, chevrons, zigzags, and herringbone. That these motifs are all textile weaves may suggest that the plaques were inspired by clothing or other woven goods that have not survived, except in rare cases (Lillios, Reference Lillios2008: 131–33).
Figure 3. Classic plaques. a: Cebolinho 1; b: Santiago; c: Aljezur; d: Brissos 6; e: Comenda; f: Olival da Pega 1; g: Pessilgais 2.
Sources: a, e, f: Leisner & Leisner, Reference Leisner and Leisner1951: pl. XXXIV, 16; pl. XI, 72; pl. XXVIII, 37; b, d: Leisner & Leisner, Reference Leisner and Leisner1959: pl. 34, 1, 3; pl. 22, 1, 36; c: Leisner, Reference Leisner1965: pl. 131, 52; g: Leisner & Leisner, 1969: pl. 7, 20, 1. Reproduced by permission of the Deutsches Archäologisches Institut.
Analyses of the formal organization of the Classic plaques, their spatial distribution, design, engraving style, formal and stylistic patterning by tomb, and other indicators of social differentiation, suggested they were genealogical records for a special class of the dead, or a kind of heraldry (Lillios, Reference Lillios2008). Given the available evidence, it seemed reasonable to suggest that they were a form of writing, namely semasiography, or writing without words (Boone & Mignolo, Reference Boone and Mignolo1994), with the base design referencing a lineage or clan, and number of registers recording the genealogical distance separating the deceased from a founding ancestor. This was supported by preliminary studies showing that Classic plaques with a lower number of registers were more concentrated in the Alentejo (the likely ‘birthplace’ of the plaques and possibly the ancestral home of associated lineages or clans), while those with a higher number of registers were distributed over a wider area (as would be expected as later generations moved away from their homeland). Statistical analyses were not conducted then, and evidence pointing to the importance of the number of registers was qualitative. For example, on one plaque from Dolmen das Conchadas, there was an unfinished sketch on one side, and on the other completed side, a different number of registers, suggesting that the draft was an error that was corrected (Leisner and Leisner, Reference Leisner1965, Tafel 27, 64, ESPRIT #160). In other plaques, either ample space remained to create another register, but was not, or a register was tightly squeezed in. These design features indicated that qualitative and quantitative approaches to the plaques are needed.
While analyses that would definitively assess the genealogical model, such as aDNA, are not currently available for associated individuals, they could become possible through excavation and analysis of sites with plaques with well-preserved skeletal remains. In the meantime, other proxy methods could provide possible insights. García Rivero and O'Brien (Reference García Rivero and O'Brien2014) tested the genealogical model using phylogenetics by analysing a sample of all plaque types—not only Classic plaques. They concluded that the plaques’ formal patterning is more likely indicative of a south-western Iberian ‘common ideological background’ and rejected the genealogical hypothesis. This study, however, misunderstood the original hypothesis, which was proposed only for Classic plaques, and did not consider any of the plaques’ qualitative features.
Given that the genealogical model was proposed over fifteen years ago and there are now nearly double the number of plaques in ESPRIT, this seems an opportune moment to evaluate whether the genealogical hypothesis can still be substantiated. To do so, we employ statistical methods. We pose two questions that assess the relationship between two variables—geography and tomb size—and two design components—the number of registers and number of bands:
1) Is there a statistically significant relationship between the number of registers of Classic plaques and their geographic distribution? This relationship had been addressed in an earlier study but with fewer plaques known at that time and without the use of statistics.
2) Is there a statistically significant relationship between the number of bands separating the top and base of a plaque and the size of the tomb in which that plaque was found? The number of bands was noted to be highly variable (with most plaques having no bands, and a few with many bands), so it seemed worth investigating whether bands may have marked some form of social identity or status. To assess this question, we selected tomb size as a known and independent variable that could be viewed as a proxy for the social status of the dead or community that constructed the monument.
Registers and Plaque Distribution
Classic plaques are characterized by a base organized by registers (most often horizontally) with iterations of geometric designs. That registers were an important design feature is suggested by qualitative evidence, including corrections and uneven spacing. For these reasons, the number of registers was recorded in ESPRIT. The Classic plaques analysed for this study are those with triangle (n = 428), chevron (n = 102), zigzag (n = 79), and checkerboard (n = 48) designs, totalling 657 plaques (see Supplementary Material: Table S1). For those with herringbone and vertical band designs, it was less evident how the registers should be counted; as these are also very rare design types (herringbone: n = 11; vertical bands: n = 8) that would be difficult to compare statistically to the other more common design types, they were not included in analyses.
If the registers of Classic plaques recorded genealogical information, we would expect to see a correlation between the number of registers, assuming this reflects a temporal dimension, and genealogical relationships. Specifically, we would expect to see: 1) fewer plaques with a low number of registers, with an increase in the number of plaques with more registers; and 2) plaques with low numbers of registers more tightly clustered in the Alentejo (their presumed area of origin) and those with higher numbers of registers more widely dispersed throughout south-western Iberia. Translated into human social behaviour, we would expect the number of people from earlier (older) generations of an important ancestor to be fewer and living (and buried) closely together in the Alentejo, where the first plaques are most likely to have been made, and those of later generations to be more numerous and more widespread.
The relationship between the number of plaques and number of registers is indicated in Figure 4; these are shown aggregated and by each of the four design elements. As the plot shows, the number of plaques initially increases with the number of registers, but after a certain number of registers, the frequency of plaques decreases for all design types. This could be due to a lack of space on the plaques (it becomes increasingly difficult to make a plaque with more registers), represents what was considered an aesthetically pleasing design, or reflects the memory of genealogies (or their importance) decreasing over time. Alone, these results are not illuminating, but they are not inconsistent with a genealogical interpretation.
Figure 4. Number of plaques by number of registers.
To assess the question of the Classic plaques’ spatial distribution by register number, we used an unpaired two-samples Wilcoxon test to compare the means of register numbers of plaques from Évora vs plaques not from Évora (Hogg & Tanis, Reference Hogg and Tanis2006). For this, we aggregated the Classic plaques by all four designs as well as looked at patterns by design element. Évora is the district in the Alentejo where the largest number of plaques was found and is generally accepted by archaeologists as the engraved plaques’ ‘heartland’. These two sample groups were roughly similar in size and, thus, could be analysed statistically.
The results of these analyses are shown in Table 2. In the case of the triangles, checkerboard, and zigzag designs, the plaques from Évora are likely to have smaller numbers of registers. For those with chevrons, the pattern is inversed (plaques from Évora have a higher mean number of registers than those outside Évora). Statistical significance was, however, only detected among the triangle plaques, as sample sizes were not sufficiently large for the other three designs.
Table 2. Mean number of registers.
* Significant p-value is labelled with an asterisk.
The distribution of the plaques by register number (aggregated) can be seen in Supplementary Material: Figure S1. In this series of maps, the distribution of the plaques expands outward from a core region in Évora (and Beja, for a plaque with one register) with increasing number of registers, and then shrinks again back to Évora.
These analyses, in combination, while not conclusive, support geographic patterning in the distribution of the Classic plaques by number of registers, which is consistent with them serving a role in genealogical (or temporal) reckoning.
Site Size and Plaque Design
In a further series of analyses, we examined internal design patterning among the Classic plaques in relationship to the contexts in which they were found. Communities of Late Neolithic and Copper Age Iberia housed their dead collectively in a range of contexts. These included dolmens (with or without corridors), tholoi (corbelled structures), hypogea (rock-cut tombs), caves, rockshelters, and pits associated with ditched enclosures. ESPRIT has records for plaques recovered from 303 sites (although the precise provenance of 32 sites, associated with 143 plaques, is unknown, as they were published only with region or country information). Plaques of known (and relatively specific) provenance (n = 1683) have been found in all types of burial contexts and some settlements (Table 3). The vast majority were recovered in megaliths (dolmens and tholoi) (81.4 per cent).
Table 3. Frequency of sites and plaques, by site type.
Given the labour involved in construction—from quarrying stones to building the chamber and corridor (if there was one), transporting them to the construction site, erecting them, and covering them with rubble and earth—megaliths have long been viewed as expressions of investment and social hierarchy. The bigger the tomb, the more labour. How much labour could be mobilized may be related to demographic factors (how many people were available) and/or social and ideological factors (nature of leadership, religious beliefs). Renfrew (Reference Renfrew1983) posited that the enormous labour required for constructing monuments like Stonehenge, at thirty million person-hours, must have entailed the existence of a central authority, like a chief, although this view is no longer widely accepted. More recently, archaeologists have turned to energetics and task analyses to generate more nuanced understandings of the labour involved in building megaliths in Ethiopia (Zena, Reference Zena2021) and Spain (Barrientos & García Sanjuán, Reference Barrientos and García Sanjuán2021). It is generally recognized that monuments not only reflect a particular social landscape or form of organizing labour but that they actively shaped that landscape and the identities of the people who engaged with them. Once a collective tomb was built, people would have visited it repeatedly to bury their (new) dead and would thus have interacted with the previous dead and their associated objects, triggering memories of the past and actions that sedimented histories and guided future actions. Although in this study we only consider megaliths (dolmens and tholoi), labour was certainly also involved in carving rock-cut tombs (hypogea). Given the difficulty in comparing the labour associated with carving hypogea to that of erecting megaliths, and the rarity of plaques in hypogea (5.8 per cent), we do not include hypogea (nor caves, rockshelters, or tumuli) in this analysis. It should also be noted that the stones of megalithic structures were sometimes moved during the site's use, i.e. their size could have changed over the monument's life.
To explore megalith size as a proxy for the status of the individuals buried within them (and by inference, the function of the plaques associated with them), we compiled information on the sites’ size. The monographs of Georg and Vera Leisner were important sources (Leisner & Leisner, Reference Leisner and Leisner1943, Reference Leisner and Leisner1951, Reference Leisner and Leisner1956, Reference Leisner and Leisner1959; Leisner, Reference Leisner1965, Reference Leisner1998), as was the doctoral dissertation of Leonor Rocha (Reference Rocha2005), who studied megaliths in the Alentejo excavated by Manuel Heleno (1894–1970). In general, three measurements were reported: chamber diameter, corridor length, and chamber height. Since not all megaliths are passage graves and therefore do not all have corridors, corridor length could not be used for comparison. Of the remaining two measurements, more tombs had their chamber diameter reported (n = 117) than their height (n = 71) (see Supplementary Material: Table S2). Chamber sizes range from 1 m2 at Amieiro 5 (Cardoso et al., Reference Cardoso, Caninas and Henriques2003) to 36 m2 at Anta Grande do Zambujeiro (Soares & Silva, Reference Soares and Silva2010). Chamber heights range from 0.7 m at Monte da Velha 2 (Soares & Arnaud, Reference Soares and Arnaud1984) to 5 m at Anta Grande do Zambujeiro (Soares & Silva, Reference Soares and Silva2010). To maximize the number of sites that could be compared for this study, and because chamber height is a strong indicator of labour (the taller and hence probably the heavier the stone to achieve the chamber's height, the more labour involved), we first confirmed the correlation between chamber height and chamber size. Since chambers are more accurately described as polygonal than perfect circles, chamber size was calculated as an area (by multiplying the dimensions of the two axes reported, if different, or by squaring the diameter). It was assumed that if a correlation were found between chamber area and height, chamber area could be used as a proxy for labour and to compare the labour involved in the tombs’ construction. Using Pearson's product-moment correlation test, we indeed found a significant correlation between chamber size and chamber height (p < 0.001; correlation coefficient = 0.54). With this knowledge, we were able to generate a dataset of 117 tombs with known areas. Using a boxplot, five tombs were recognized as outliers, or particularly large tombs: Olival da Pega 1 (22.4 m2), Boiça 1 (26.5 m2), Escoural (36 m2), Vale de Rodrigo 1 (36 m2), and Anta Grande do Zambujeiro (36 m2).
What would mobilize a community to build such enormous tombs as Anta Grande do Zambujeiro? When a biographical approach is taken to site histories, it often becomes apparent that labour-intensive monuments were often constructed in persistent places, i.e. those that already had a sacred or special quality (Schlanger, Reference Schlanger, Rossignol and Wandsnider1992). This seems to have been the case at Stonehenge (Jacques et al., Reference Jacques, Phillips and Lyon2017) and is likely for the largest of the Iberian megaliths. Geoarchaeological studies have shown that some of the stones used to build Iberian monuments originated some distance from the tomb, ranging from a few hundred metres to twenty kilometres (Kalb, Reference Kalb1996, Reference Kalb2011; Lozano et al., Reference Lozano, Ruiz-Puertas, Hódar-Correa, Pérez-Valera and Morgado2014; Boaventura et al., Reference Boaventura, Moita, Pedro, Mataloto, Almeida, Nogueira, Boaventura, Mataloto and Pereira2020). Thus, in its early career, tomb size would have been a variable dependent on the sacredness of the space and available labour. Once built, the sacredness or potency of the tomb would have accrued as more people were buried and the site increased in mnemonic density (Lillios, Reference Lillios2015). Particularly large and sacred tombs would have become the preferred resting places for elites, and by virtue of their size they were also able to house larger numbers of the dead. Perhaps not surprisingly, there is a positive correlation between the size of a tomb and the number of plaques found in it (using Spearman's rank-order correlation test; p = 0.014). Thus tomb size can be viewed as both a dependent variable (in the tomb's early history) as well as an independent variable, once the monument was built.
Tomb Size, Bands, and Collars
Turning to the relationship between some of the design elements of Classic plaques, if these plaques were simply part of a shared iconographic system among communities in south-western Iberia (as García Rivero & O'Brien, Reference García Rivero and O'Brien2014 suggested), there should be no correlation between their designs and labour proxies, such as tomb size. In our study, we assessed this hypothesis.
A key feature of Classic plaques is the separation between the top and base fields by either a horizontal line or a set of bands, which can be un-hachured and/or hachured (or filled with tiny triangles). As Table 4 shows, most have no bands and just a horizontal line (69 per cent; n = 594), while another twenty per cent (n = 175) have one band (either un-hachured or hachured). The maximum number of bands is fourteen. To see if there is a relationship between the number of bands and tomb size, a Spearman's rank-order correlation test was applied, because the numbers of bands in the sample are not normally distributed. The results indicate a significant positive correlation between the two variables (correlation coefficient = 0.14, p-value < 0.001), although the correlation is relatively weak. The highest number of bands was on a plaque from Anta Grande do Zambujeiro (Figure 5a), one of the largest tombs in Iberia. Two other plaques from Anta Grande were in the ‘top ten’ of those plaques with the largest number of bands (one has ten bands, another has seven) (Figure 5c and 5g). Other plaques with numerous bands (i.e. more than seven) come from Escoural (Figure 5d), another extraordinarily large tomb with one of the greatest number of plaques found (at least ninety), and the very large tomb of Comenda da Igreja (21.6 m2) with a plaque featuring seven bands (Figure 5i). Two others come from the ditched enclosures of Perdigões and Pijotilla (Figure 5b and 5e), which have been considered special aggregation places with a cosmological significance (Valera et al., Reference Valera, Žalaitė, Maurer, Grimes, Silva and Ribeiro2020). The remaining ‘top ten’ with large numbers of bands include one from Cebolinho 1 (16 m2; seven bands) (Figure 5f) and Freixa (11.25 m2; seven bands) (Figure 5h). The size of Barcarrota, which yielded a plaque with seven bands (Figure 5j), could not be determined.
Figure 5. Classic plaques with seven or more bands. a: Anta Grande do Zambujeiro; b: Perdigões; c: Anta Grande do Zambujeiro; d: Escoural; e: La Pijotilla 2; f: Cebolinho 1; g: Anta Grande do Zambujeiro; h: Freixa; i: Comenda da Igreja; j: Barcarrota.
Sources: a: photograph by Eduardo Estellez/Adobe Stock, b: photograph by K. Lillios, photograph by permission of Antonio Valera; c: http://www.matriznet.dgpc.pt/MatrizNet/Objectos/ObjectosConsultar.aspx?IdReg=1038032; Reproduced by permission of Museus e Monumentos de Portugal, E.P.E. d: http://www.matriznet.dgpc.pt/MatrizNet/Objectos/ObjectosConsultar.aspx?IdReg=128935 , Reproduced by permission of Museus e Monumentos de Portugal, E.P.E.; e: Hurtado Pérez, Reference Hurtado Pérez1979, photograph by permission of V. Hurtado; f: Leisner & Leisner, Reference Leisner and Leisner1951: pl. XXXIV, 15; reproduced by permission of Deutsches Archäologisches Institut; g: Gut, Reference Gut1990, vol. I, pl. 6b; Reproduced by permission of Deutsches Archäologisches Institut; h: http://www.matriznet.dgpc.pt/MatrizNet/Objectos/ObjectosConsultar.aspx?IdReg=140479 ; Reproduced by permission of Museus e Monumentos de Portugal, E.P.E.i: photograph by K. Lillios, courtesy of Museu Nacional de Arqueologia; j: Leisner & Leisner, Reference Leisner and Leisner1959: pl. 52, 11, 12; reproduced by permission of Deutsches Archäologisches Institut
Table 4. Frequency of Classic plaques by number of bands.
Another intriguing pattern is that all those plaques with high numbers of bands (seven or more) have a collar. However, collars are generally rare in Classic plaques (18 per cent). To assess whether band number and chamber area are significant predictors of the presence of a collar, a logistic regression model was used. The presence of a collar was the response variable, while the number of bands and the size of the chamber area were the explanatory variables (Supplementary Material: Table S3). The results show a significant positive correlation (correlation coefficient = 0.631) between the number of bands and the presence of a collar, meaning that plaques with more bands are more likely to have a collar. Such a correlation was noted in earlier analyses using a smaller dataset (Lillios, Reference Lillios2008: 164). On the other hand, there does not seem to be a significant relationship between chamber area and the presence of a collar.
It was also noted that almost all plaques with large numbers of bands (and a collar) had a zigzag base design. A logistic regression model using the presence of zigzag design as the response variable and band number and the presence of a collar as explanatory variables (Table 5) indicate that there is a significant positive correlation between the number of bands and the odds of having a zigzag base design.
Table 5. Results of the logistic regression of presence of zigzag design.
* Significant p‐values are labelled with an asterisk.
Given that Classic plaques with a zigzag design and numerous bands tend to be found in the largest tombs, and that collars are strongly correlated with high band numbers, plaques with both a collar and many bands appear to have been markers of elite individuals. What these designs represent is not at all clear. It may be that the zigzag marked a special lineage or clan (religious specialists?). It may be relevant that a fairly large percentage (37 per cent) of the Biomorph Simple plaques (often viewed by scholars as deities, owls, or hybrid beings) also bear a zigzag design.
Discussion and Conclusion
Our work reassesses the genealogical hypothesis for the Classic plaques using a series of statistical analyses to evaluate whether significant patterning exists among their design, their geographic distribution, and the size of the tombs in which they were found. Our study indicates a positive relationship between tomb size and the presence of plaques with high numbers of bands. There is also a relationship between the number of bands, the presence of a collar, and the base design. What the bands and collars specifically denote is, however, unclear. Perhaps bands recorded the number of children, territorial claims, military conquests, or something else considered remarkable enough for Late Neolithic and Copper Age people to record on that plaque (and the person associated with it).
One design element we have not yet discussed are the straps on Classic plaques (Figure 3), which can be horizontal or vertical. For the two Classic plaques with sexed individuals, the one with horizontal straps (from Cova das Lapas) was found with a young male, while the one with vertical straps (from Monte Canelas) accompanied a female (Table 1). Neither site was included in the analyses discussed above, as Cova das Lapas is a cave and Monte Canelas is a hypogeum. The other two plaques with sexed individuals are a Biomorph plaque (from Porto Torrão, with a female) and one too fragmented to assign to a type (from Santa Margarida 3, with a female). An intriguing possibility is that the straps on Classic plaques indicate biological sex (horizontal = male; vertical = female), especially given the similarity of the plaques’ designs with textile weaves and the highly gendered nature of clothing. This proposition clearly requires further investigation with a much larger sample of sexed individuals.
Within our hypothesis, there are four possible interpretations of the Classic plaques: 1) they were definitely not genealogical records; 2) they were possibly genealogical records; 3) they were likely to have been genealogical records; and 4) they were definitely genealogical records. Their formal patterning, which is not random nor simply the outcome of a shared iconographic template, rules out the first interpretation. Given the absence of aDNA linked to individuals with plaques, it is not possible to assess the fourth option. With currently available evidence, which supports statistically non-random associations between the plaques’ designs and their spatial distribution, the third interpretation, that they were likely to have been genealogical records, is our best explanation for Classic plaques.
If Classic plaques communicated specific information about the identity of certain individuals, they would be some of the oldest form of known writing or recording systems in the world, as they are contemporary with early Mesopotamian writing, although some work suggests that proto-writing systems existed as early as the Upper Palaeolithic (Bacon et al., Reference Bacon, Khatiri, Palmer, Freeth, Pettitt and Kentridge2023). As a means to record kinship, the engraved plaques may have been a textual component of the ideological arsenal used to legitimize the status and attendant social privileges of individuals or kin groups in the Iberian Late Neolithic and Copper Age and to perpetuate these over time.
The end of the era of the Iberian plaques was ushered in by the emergence and dispersal of the so-called Beaker culture (Olalde et al., Reference Olalde, Mallick, Patterson, Rohland, Villalba-Mouco, Silva and Dulias2019). Intriguingly, all the geometric motifs of the engraved plaques are found on Iberian Bell Beakers. Perhaps the memories of clan histories persisted through the late third and early second millennia bc in Iberia, although reconfigured in new media. The engraved plaques appear to us to be the experiments of ancient Iberians in recording histories. If they were a form of writing, they could be viewed as another nail in the coffin of ‘prehistory’, a concept many would argue is dead (Schmidt & Mrozowski, Reference Schmidt and Mrozowski2013). Clearly much more work remains to be done on the plaques to elucidate their formal patterning and understand the organization of their production.
To conclude, much of the debate around the Iberian plaques, which is centred around whether they were heraldic, or were related to ancestors, or represented deities, obscures more than reveals. Individual plaques clearly had social lives (Kopytoff, Reference Kopytoff and Appadurai1986), and as a class of objects, perhaps sacred texts, their use or meaning no doubt changed over time. When they were first created, Classic plaques may have been accorded to individuals whose genealogical history was central to the mythos of a community, but who, over time, became important ancestors, or even deities themselves.
Acknowledgements
Funding for the statistical analyses was provided by the College of Liberal Arts and Sciences at the University of Iowa. Data used for analyses are available upon request. We are grateful to João Zilhão, Jeffrey Quilter, and Daniel Rubin for helpful comments, Ted Fitzgerald for maintaining ESPRIT, Cate Frieman and Madeleine Hummler for their insightful editorial skills, and three anonymous reviewers for their excellent suggestions.
Supplementary Material
To view supplementary material for this article, please visit https://doi.org/10.1017/eaa.2024.34.
Open access
