International Pots of Mystery: Using PXRF spectroscopy to identify the provenance of botijas from 16th Century sites on Peru's north coast
Introduction
Over the last few decades increased miniaturisation of analytical instrumentation has had a significant impact upon the archaeological community, in particular developments related to portable X-ray fluorescence (PXRF) spectroscopy. Non-destructive, highly portable, commercially available and much cheaper than laboratory-based XRF instruments, PXRF spectrometers have been widely employed for the chemical characterisation of obsidian and metal, and increasingly ceramics. Although pottery presents a number of difficulties for such analysis, studies involving PXRF spectrometers applied in a manner mindful of methodological issues can provide useful insights into a range of issues around production, distribution and consumption.
In Peru, recent studies have used PXRF analysis to explore questions related to ceramic production and exchange (del Solar-Velarde et al., 2014, del Solar-Velarde et al., 2016; Pena, 2013; Sharratt et al., 2019; Vanvalkenburgh et al., 2018; Wackett, 2017). del Solar-Velarde et al., 2014, del Solar-Velarde et al., 2016 used PXRF analysis in conjunction with scanning electron microscopy (SEM), optical microscopy and cathode luminescence to study ceramics from the site of San José de Moro, specifically, Late Mochica fineline bichrome and polychrome, and Coastal and Highland Cajamarca-style ceramics. All three groups were shown to be chemically distinct, supported by results across techniques, indicating the use of different recipes and raw materials and suggesting that inhabitants incorporated local and non-local ceramics into their burial traditions. Furthermore, examination of pigments using PXRF analysis suggested possible incorporation of highland materials and techniques. Other studies have also underscored the need to use PXRF in conjunction with other analytical techniques. Sharratt et al.'s (2019) PXRF-based study of 70 tinajas across eight wineries in the Moquegua Valley led to the chemical discrimination of the majority of ceramics from the Valdivia winery, and suggested differences in ceramic production network participation between upper and lower valley wineries; however, due to the limitations of PXRF analysis, laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) was suggested as a possible next step to further refine chemical groups. Similarly, Wackett's (2017) and Vanvalkenburgh et al.'s (2018) studies in the Moquegua and Zaña Valley, respectively, have also indicated the need for supporting techniques to improve discrimination, with only a few groups discerned from otherwise chemically similar sample sets.
In cases where PXRF studies alone show a lack of chemical differentiation within archaeological ceramic populations, such as some of those described above, numerous factors may be at play, including those related to instrumentation setup and sample selection (such as element selection, calibration setup, matrix effects, limits of detection and sample preparation), similar source materials and potting techniques, as well as the possibility that all samples were in fact made by the same production centre. In these cases, use of another analytical technique is suggested to ascertain if said similarity is a result of PXRF instrument setup related issues alone. In situations such as single valleys/small areas of production and distribution, where the geology and production methods are similar, the use of other types of analyses may well enable greater insights into ceramic production and distribution than PXRF analysis due to issues such as element selection and limits of detection. In this vein, the aforementioned studies also highlight that where ceramics were produced in multiple regions with distinctly different chemistries, and where those chemical differences are able to be measured by PXRF spectrometers, successful identification of at least production region should be possible.
This paper presents attempts to determine the provenance of botijas from the north coast of Peru at two colonial sites: San Miguel de Piura and Carrizales. Botijas are one of the most common and widely distributed ceramic types recovered from Spanish colonial archaeological sites (terrestrial and maritime) in the Americas, a truly global storage vessel made and found across the Atlantic and Pacific regions. Moreover, pre-existing studies provide a good experiment methodological starting point, presenting information on production and associated chemistry by region, dates and distribution, all of which can be used to inform the experimental set up. In this paper, the authors explore the benefit of such foreknowledge and test the limits of analysing botijas in the field, particularly whether or not it is possible to discern regionally-specific production centres. Moreover, through the analysis of provenance-based botija distribution patterns across multiple contexts at multiple sites, this study also seeks to explore the economic, social, religious and political factors that contributed to the circulation and use of these vessels and how these factors intersect with place of manufacture.
Section snippets
Approaching pottery with a PXRF spectrometer
The increased popularity of PXRF spectroscopy over the last ten years has not been without its growing pains as the wider archaeological community has explored and developed its understanding of the methodological requirements and associated limitations and advantages of this portable instrumentation (Killick, 2015; Shackley, 2010; Shugar, 2013; Speakman et al., 2011). In regards to ceramic studies, careful attention must be paid to experimental methodology. For archaeologists unfamiliar with
Botijas
Botijas (olive jars/botijuelas) are commonly found in Spanish colonial archaeological sites across the New World. Large, coarse and wheel-made, these ceramic storage vessels arrived in the Americas with the onset of Spanish colonisation and were soon produced in the colonies (Avery, 1997; Goggin, 1960; Marken, 1994: 41–138). These vessels generally possess ringed mouths, narrow necks and rounded bodies tapering to rounded bottoms, although flat bottom botijas have been identified (James, 1988;
The sites
San Miguel de Piura (Piura la Vieja) and Carrizales represent two very different site types. Although both sites were abandoned by the late 16th Century and were important to the contemporary maritime trade operating along Peru's north cost, the former was the first city established by the Spanish under Diego de Almagro and the latter, an indigenous reducción. Having been built and occupied by different communities under different circumstances, comparisons of material culture patterns at these
Experimental methodology
The sherds in this study derive from the adobe structure within the church at San Miguel de Piura, specifically from cuadrículas 88, 92 and 119 (excluding surface material), and from a variety of contexts within Conjunto 123 at Carrizales, specifically those surveyed and excavated during 2012, 2014 and 2015 (Fig. 2). A total of just over 380 botija sherds were identified at San Miguel de Piura and 443 sherds from Carrizales.
Visual identification
In the San Miguel de Piura collection, all sherds were visually identified as Spanish-made, except one that appeared to be possibly a New World product based on body colour, although the temper indicated a Spanish origin: 88-10-a-6. Of the rims, three are possible Type 5 rims (119-16-a-24, 92-8-a-2, 92-3-a-4) and three Type 3 semi-triangular/triangular style rims (119-23-a-10, 92-2-a-2, 92-2-a-12, Marken, 1994: 51, 60). Rim 92-3-a-4 seems to flare more so than the rest, and 92-2-a-2 seems
Discussion
The results of this study show that regions and even countries of origin can be determined for botija sherds based on the analysis of data collected with PXRF spectrometers, when they are appropriately calibrated and applied. In the cases of San Miguel de Piura and Carrizales, Old and New World sources were determined for botijas, with groups discerned largely by Ca, Cr and As concentrations. These chemically-based provenance assignments not only support visual identifications but also allow
Conclusions
This study successfully identified botijas produced in different regions based on chemical fingerprints determined using data obtained with a PXRF spectrometer. Botijas are one of the most ubiquitous ceramic types found on sites across the Americas and there is great potential socio-economic, political and religious information in the identification of not only the ceramic type but in the provenance of individual sherds. Analysis of botijas using PXRF spectrometers is clearly fit-for-purpose
Acknowledgments
The authors wish to thank the XRF laboratory, SSEAU, MWAC, University of New South Wales for generously allowing the use of their PXRF instrument, as well as the Ministerio de Cultura (Dirección Desconcentrada de Cultura Piura) for enabling this project, including the use of their offices in Piura for analysis. Research on the San Miguel de Piura collection was authorised by the permission RD 018-2016-DGM from the Peruvian Ministry of Culture. The authors also wish to thank Rabeya Akter (ICP
References (71)
- et al.
Portable XRF analysis of archaeological sediments and ceramics
J. Archaeol. Sci.
(2015) - et al.
Identifying New World majolica from 16th–18th Century sites on Peru’s north coast
J. Archaeol. Sci. Rep.
(2018) The awkward adolescence of archaeological science
J. Archaeol. Sci.
(2015)- et al.
Geochemical characterisation of north Asian glazed stonewares: a comparative analysis of NAA, ICP-OES and non-destructive PXRF
J. Archaeol. Sci.
(2012) - et al.
Sourcing ceramics with portable XRF spectrometers? A comparison with INAA using Mimbres pottery from the American Southwest
J. Archaeol. Sci.
(2011) Trayectoria de la Cerámica Vidriada en el Perú
- et al.
Handheld XRF analysis of Maya ceramics: a pilot study presenting issues related to quantification and calibration
Mark Wainwright Analytical Centre Report: Analysis of Sample by LA-ICP-MS
(2016)Informe Final del Proyecto Investigación Arqueológica San Miguel de Piura; Primera Fundación Española en el Perú (1534), Temporada 2008
(2009)Informe Final del Proyecto Investigación Arqueológica San Miguel de Piura; Primera Fundación Española en el Perú (1534), Temporada 2011. Informe inédito. Lima: [s.n.]
(2012)
Fundación, esplendor y colapso de la iglesia de San Miguel de Piura, primer templo del Perú
Boletín Arqueología PUCP
Informe Final del Proyecto de investigación de colecciones y fondos museográficos Estudio de la cerámica asociada a la iglesia de San Miguel de Piura (1534–1578). Informe inédito. Lima: [s.n.]
Pots as Packing: The Spanish Olive Jar and Andalusian Transatlantic CommercialActivity, 16th–18th Centuries
1922, La Crónica del Perú
Los Trece de la Fama
Two pre-Columbian Pottery Productions of the Peruvian North Coast: An Archaeometrical Approach to Understand Technical Traditions of the Mochica and Cajamarca Cultures
Characterization of pre-Columbian artefacts in situ through handheld portable X-ray fluorescence spectrometry: the case of ceramics from the Mochica site of San José de Moro (Peru)
Herit. Sci.
“If you eat their food…”: diets and bodies in early colonial Spanish America
Am. Hist. Rev., June
The Body of the Conquistador: Food, Race and the Colonial Experience in Spanish America, 1492–1700
Local and European transport jars in Panamá Viejo
Principal component analysis for compositional data with outliers
Environmetrics
Effects of elevated levels of lead in ceramics on provenancing studies using non-destructive PXRF: a case study in Byzantine Cypriot glazed ceramics
X-Ray Spectrom.
Tendencias de consumo en México durante los períodos colonial e independiente
The Spanish Olive Jar: An Introductory Study
A reassessment of the chronology and typological framework of the Spanish olive jar
Hist. Archaeol.
Neutron activation analysis of colonial ceramics from southern highland Ecuador
Archaeometry
Neutron activation analysis of Inca and colonial ceramics from central highland Ecuador
Archaeometry
On the Edge: A Study of Spanish Colonisation Fleets to the West Pacific and Archaeological Assemblages From the Solomon Islands
Sourcing olive jars using U-Pb ages of detrital zircons: a study of 16th century olive jars recovered from the Solomon Islands
Geoarchaeology
Sherds on the edge: characterisation of 16th century colonial Spanish pottery recovered from the Solomon Islands
Archaeometry
Zooarchaeology and changing food practices at Carrizales, Peru following the Spanish invasion
Int. J. Hist. Archaeol.
Inside the Reducción: crafting colonial foodways at Carrizales and Mocupe Viejo, Zaña Valley, Peru (1570–1700)
Int. J. Hist. Archaeol.
Andalusian Ceramics in Spain and New Spain: A Cultural Register From the Third Century B.C. to 1700
Pottery From Spanish Shipwrecks, 1500–1800
Interpreting Cultural and Sociopolitical Landscapes in the Upper Piura Valley, Far North Coast of Perú (1100 B.C.- A.D. 1532)
Cited by (6)
Indigenous labor and the circulation of majolica in the colonial Andes
2023, Journal of Archaeological Science: ReportsCrafting a colony: Geochemical insights into 1000 years of state craft in the Moquegua Valley, Peru
2023, Journal of Archaeological Science: ReportsSelecting reproducible elements in non-destructive portable X-ray fluorescence analysis of prehistoric and early historical ceramics from Korea
2023, Journal of Archaeological Science: ReportsCitation Excerpt :Although traditional methods such as instrumental neutron activation analysis (INAA) (Glascock and Neff, 2003) and inductively coupled plasma mass spectrometry (ICP-MS) (Kennett et al., 2001) have played a pivotal role and significantly contributed to advances in provenance studies, archaeologists have longed for the development of non-destructive methods. In this regard, the use of portable X-ray Fluorescence (pXRF) analysis has been a welcome development, and the applications of this technique to provenance studies have substantially increased in number during the last decades (e.g., Ikeoka et al., 2012; Tykot et al., 2013; Frahm 2014; Ceccarelli et al., 2016; Emmitt et al., 2018; Frahm 2018; Kelloway et al., 2019; Simsek et al., 2019; Eslami et al., 2020; Tykot and Martin 2020). However, the reliability of chemical characterization of ceramics by pXRF has been debated.
Introducing the BRICC (Bricks and rocks for Instruments’ ceramic calibration) sets: Open-source calibration materials for quantitative X-ray fluorescence analysis
2022, Journal of Archaeological Science: ReportsCitation Excerpt :We hope that the program will return at some point, but in the meantime, there remains no inexpensive source of geological SRMs. Harry Rowe (then at the Texas Bureau of Economic Geology, now at Premier Oilfield Group) and colleagues (Rowe et al., 2012; Rowe and Kaiser, 2012; Kaiser et al., 2013) devised a “mudrock” calibration, developed for one brand of pXRF instruments, that has since been employed to analyze ceramic artifacts (e.g., Ligman, 2016; Cook et al., 2017; López-García et al., 2018; Shalvi et al., 2019; Krueger et al., 2020; Michałowski et al., 2020; LeMoine and Halperin, 2021). Note that “mudrock” is something of a catch-all term for all sorts of fine-grained clastic sedimentary rocks, including shale, slate, siltstone, and mudstone, among others.