Lessons From Ceramic Petrography: A Case of Technological Transfer During the Transition From Late to Inca Periods in Northwestern Argentina, Southern Andes

1 Introduction

The investigation of pottery production organization in archaeological contexts has been undertaken by numerous researchers over time, who have developed various theoretical and methodological approaches (Arnold, 1971, 1975, 1985, 1991, 1993, 1994, 2000a; Bishop & Neff, 1989; Blackman et al., 1993; Costin, 1991; Costin & Hagstrum, 1995; Druc, 1996, 2000, 2009, 2011; Druc & Velde, 2021; Gosselain, 1992, 1999, 2000; Gosselain & Livingstone-Smith, 2005; Hagstrum, 1985; Peacock, 1984; Rice, 1981, 1987; Roux, 2019; Sinopoli, 1991; van der Leeuw, 1977, 1984; Vandiver, 1988). These approaches have not always been mutually exclusive in their theoretical and methodological trajectories. To the contrary, they have facilitated exploration and assessment of diverse aspects or subjects related to the organization and scale of ceramic production within various geographic and cultural contexts, encompassing both archaeological and ethnoarchaeological realms (compare Arnold, 2000b; Costin, 1991; Druc, 2009; Rice, 1981, 1987; van der Leeuw, 1984, 1993; Vandiver, 1988). Over the past three decades, both ceramic petrography and compositional analyses of archaeological ceramics using Instrumental Neutron Activation Analysis (INAA) have constituted a significant interdisciplinary field of study bridging the realms of human sciences (anthropology and archaeology) and physical sciences (chemistry and physics). This has facilitated a focused investigation into the technology and origin of ceramic artifacts, as well as the ceramic raw materials (clays and temper) utilized in their creation (Asaro & Adan-Bayewitz, 2007; Blackman & Bishop, 2007; Cremonte, 2014a, b; De La Fuente, 2004, 2011a; De La Fuente et al., 2015; Druc, 2014, 2016; Druc et al., 2017, 2020; Druc & Uribe, 2018; Glascock, 1992; Hancock et al., 2007; Harbottle & Holmes, 2007; Hughes, 2007; Ixer et al., 2014; Kilikoglou et al., 2007; Mommsem & Sjoberg, 2007; Mommsem, 2012; Neff, 2000, 2002; Newton et al., 2007; Plá & Ratto, 2007; Riehle et al., 2023; Yellin, 2007).

This article examines and deliberates on the outcomes derived from extensive ceramic petrography, morphological studies, and archaeometric analyses – magnetic susceptibility/thermoluminiscence (SUS/TL) – conducted on pottery from the Late and Inca periods at various archaeological sites located in the southern sector of Abaucán Valley (Department of Tinogasta, Province of Catamarca, Northwestern Argentina; Figure 1). Furthermore, we explore the choices made by potters in terms of technology and their impact on production during the transitional phase between these two pre-Hispanic periods.

Figure 1

Southern sector of Abaucán Valley showing the sites mentioned in the text (Dept. of Tinogasta, Catamarca, Argentina). References: (1) CV2, (2) CV5, (3) CV6, (4) Barranca de la Puntilla-BP-, (5) SaCat04, (6) SaCat14, (7) Río Colorado, (8) SaCat15, (9) SaCat13, (10) SaCat12, (11) SaCat02, (12) Costa de Reyes No 5 CR5B-, and (13) GPS042.

2 Shared Knowledge, Shared Practices: Communities of Potters and Chaîne Opératoire

Communities of practice emerge from shared activities, learning, and knowledge exchange, fostering a collective identity (Wenger, 1998). These practices may be explicit or implicit, forming a common pool of skills, narratives, and tools. Membership is shaped by mutual interaction rather than social categories.

Ethnographic studies highlight how potters collaborate across various stages of pottery-making, such as clay selection, processing, and firing (Gosselain, 1992, 1999, 2000; Sillar, 1997, 2000, 2009). Many techniques are evident in finished pottery, allowing replication, while others require direct demonstration due to their complexity (Budden & Sofaer, 2009; Potts, 2022). Practical demonstrations and guidance become crucial for transmitting intricate motor skills and technical knowledge, both individually and across societies.

Mastery of forming techniques relies on both cognitive knowledge (connaissance) and experiential learning (savoir-faire) (Gandon, et al. 2014, 2018, 2021; Harush, et al. 2020). This distinction parallels the contrast between discursive and performative knowledge (Budden & Sofaer, 2009). Scholars drawing on Mauss’ concept of the “complete person” argue that potters physically shape themselves through their craft (Lehmann, 2018). As apprenticeship occurs within a cultural context, motor skills become vehicles for both technical expertise and sociocultural identity.

The concept of “chaîne opératoire,” from French archaeology and anthropology, refers to the sequence of steps in artifact production (Leroi-Gourhan, 1964, 1973). It serves as a key analytical tool for understanding technology, social organization, and cultural transmission (Gallay et al., 1996; Roux, 2019). In archaeology, it helps reconstruct past technologies and cognitive processes by analyzing production sequences (Tixier, 1967). Chaîne opératoire also aids in studying technological innovation and transmission. Variations in production sequences reveal experimentation, adaptation, and the spread of techniques across cultural networks (Gelbert, 2003). By tracing these operational sequences, archaeologists gain insight into human interactions and the interconnectedness of ancient societies (Roux, 2016).

3 Late and Inca Periods in Northwestern Argentina (NWA): Chiefdoms, Power, State Rule, and Pottery Production

The Late Period (circa 950–1450 AD) in NWA was conventionally characterized as a phase marked by substantial regional growth, heightened socio-political complexity, disparities in wealth, economic hierarchy, and internal conflicts (warfare) (González, 1977; González & Pérez, 1972; Ottonello & Lorandi, 1987; Raffino, 1991; Tarragó, 2000). This era witnessed the emergence of regional chiefdoms in distinct geographical regions, each associated with specific valleys. This led to centralized authority, controlled labor forces, increased social stratification, specialized craftsmanship, and the establishment of large fortified archaeological sites strategically situated in defensive locations (cf. Leoni & Acuto, 2008; Nielsen, 2007). Notably, the Belén and Santamaría cultures emerged as Late Period chiefdoms marked by heightened socio-political complexity, a strong emphasis on agriculture and pastoralism – evident through extensive stone masonry settlements and intensified agricultural infrastructure – and the practice of specialized craft production primarily evidenced through pottery making (González, 1977; González & Pérez, 1972; Ottonello & Lorandi, 1987; Raffino, 1991; Tarragó, 2000) (cf. González, 2004 for insights into metallurgical production). A significant aspect of the Late Period pertains to mortuary practices, specifically the presence of funerary urns used for infant burials and, in exceptional instances, adult burials as well (Berberían, 1969; González, 1977; González & Pérez, 1972). These burials involved intricate ritual mortuary practices, with varied patterns of elaborate decoration adorning the external surfaces of funerary urns. This decorative aspect is a distinguishing feature of Late Period societies (Nastri, 2008; Sempé & García, 2007; Wynveldt, 2007, 2008). The archaeological sites are characterized by large clustered residential compounds, where households serve as the fundamental units of spatial organization (Leoni & Acuto, 2008). Some of these sites, such as Belén and Santamaría, incorporate Inca architecture within distinct sectors, as seen in Fuerte Quemado (middle Calchaquí Valley, Catamarca), Hualfín (Hualfín Valley, Catamarca), and La Paya (northern Calchaquí Valley, Salta) (Ambrosetti, 1907–1908; González & Díaz, 1992; Orgaz, 2008; Raffino, 1991). Although the Sanagasta culture was not initially classified as a chiefdom, it featured sizable settlements with mudbrick (tapia and adobes) constructed rooms. Like Belén and Santamaría, Sanagasta also practiced burying infants in decorated funerary urns and had agricultural terraces (González, 1977; González & Pérez, 1972). While Late Period pottery production is not fully comprehended, ongoing research suggests household-level production with certain degrees of dimensional and paste standardization for specific ceramic forms (De La Fuente, 2011a; Páez, 2010).

The Inca state emerged as a macro socio-political entity that exerted dominion over a vast territory for nearly a century (c. AD 1480–AD 1532), extending its political and economic control from northern Ecuador to southern Chile (Espinoza Soriano, 1970, 1975, 1987; Murra, 1975, 1980; Pease, 1973; Rostworowski, 1978, 1983). This expansive territory was divided into provinces, each subject to distinct systematic policies governing the production of goods, leading to specialized craftsmanship (coca growers, miners, potters, weavers, farmers, fishers, etc.) at both local and state levels (D’Altroy, 1992; Espinoza Soriano, 1970, 1975, 1987; Murra, 1975, 1980; Pease, 1973; Rostworowski, 1978, 1983; Spurling, 1992). Labor constituted the primary Inca tribute, collected from local groups, and according to ethnohistory, communities were taxed on a household basis in accordance with the population recorded during periodic censuses (Le Vine, 1987; Murra, 1975, 1980, 1982).

In the realm of ceramic production, the Incas augmented output through local potters and colonists known as mitmaqkuna, creating enclaves for labor (cf. D’Altroy, 1992; D’Altroy et al., 1994, 2000; Espinoza Soriano, 1970, 1975, 1987; Hayashida, 1999; Murra, 1978, 1980; Spurling, 1992). Inca pottery exhibits remarkable uniformity in both form and decoration across their expansive territory (Costin, 1986, 1991; Costin & Hagstrum, 1995; D’Altroy & Bishop, 1990; Meyers, 1975; Morris & Thompson, 1985), although instances of local styles integrated with Inca designs can be observed at a regional level (cf. D’Altroy, 1992; Donnan, 1997; Hayashida, 1994, 1999; Kriscautzky, 1999; Spurling, 1992; Williams, 1999, 2000). A pivotal aspect of Inca ceramics pertains to their distribution and function within the empire, with research focusing on their manufacture and consumption (Bray, 2003; Cremonte, 1991, 1994; D’Altroy & Bishop, 1990; see also Williams et al., 2016).

Research examining the production and distribution of Inca ceramics in the central Andes. D’Altroy and Bishop’s (1990) study of Inca ceramics in the upper Mantaro Valley employed INAA on 173 samples from Cuzco, Lake Titicaca, the upper Mantaro Valley, and Tarma. Analysis of 115 samples supported regionally focused production and distribution under Inca rule, with centralized control over raw materials and multiple manufacturing centers creating overlapping consumption patterns. Evidence also indicates the transport of certain ceramics between provinces and Cuzco, including sumptuary shallow plates. On the other hand, Hayashida (1999), Hayashida et al. (2003) examined pottery production in the Leche Valley, showing a localized system. Petrographic and INAA analyses suggest workshops supplied their administrative centers with minimal inter-site exchange, though some compositional overlap between La Viña and Tambo Real warrants further study.

Recent compositional analyses using petrography, INAA, and LA-ICP-MS confirm a localized model of ceramic production in the Central Andes, where potters utilized local materials to fulfill both household and imperial demands (Alconini, 2014; Bray & Minc, 2011, 2016, 2020; Burger et al., 2019; Chacaltana-Cortéz et al., 2023; Cremonte et al., 2015; Davenport, 2019, 2020; Kosiba et al., 2023; Minc et al., 2016; Pïscitelli et al., 2016; Szilágyi & Szakmány, 2009; Szilágyi et al., 2012; Williams et al., 2016, 2019).

In the NWA region, ceramic styles are differentiated between Inca/Cuzco, Inca Provincial, and local styles (Calderari & Williams, 1991; Williams, 2000). While the prevalent hypothesis suggests that certain Inca settlement types, referred to as “administrative centers,” were key pottery production centers with goods distributed from these centers to each region for local and state consumption (D’Altroy & Bishop, 1990; D’Altroy et al., 1994), contradictory evidence arises from other Inca settlements, categorized as tambos (a type of site along the Inca road), where various pottery production stages have been documented (cf. Alconini, 2013; Bárcena & Roman, 1986–1987; Hayashida, 1999, pp. 340-341; Hosler, 1996; Orgaz, 2008; Spurling, 1992).

Noteworthy compositional research by Ratto and colleagues over the past two decades in the medium and higher sectors of the Abaucán valley (Tinogasta, Catamarca) indicates that pottery production was integral to a local conservative tradition dating back to the Formative period and extending into Inca times, with Batungasta archaeological site serving as a focal point (De La Fuente, 2011a; Plá & Ratto, 2003, 2007; Ratto, 2013; Ratto & Orgaz, 2002–2004; Ratto et al., 2002, 2004, 2006).

As highlighted by Williams (2000, 2010) and Cremonte and Williams (2010), the distribution of Cuzco ceramics was constrained to specific geographic regions, with many Inca ceramics in the NWA region intended for regional consumption (Cremonte & Williams, 2010, Figure 2; Williams et al., 2016). However, compositional evidence suggests that certain small and lightweight vessel types, such as shallow plates, were transported to distant regions far from the Inca capital of Cuzco (Cremonte & Williams, 2010; D’Altroy & Bishop, 1990; D’Altroy et al., 1994; Lazzari et al., 2009; Williams, 2010).

Figure 2

Pottery sherds analyzed in this research: (a) Sanagasta rim, infant funerary urn, (b) Belén body, bowl, (c) Belén rim-body, bowl, (d) Belén body, urn, (f)–(h) Inca body, aryballous, (i) Belén body, infant funerary urn, (j) and (l) Sanagasta rim, bowl, (k) Sanagasta rim, infant funerary urn, and (m)–(p) Sanagasta rims, bowls.

4 Geographical, Geological, and Pre-Hispanic Cultural Contexts: The Southern Sector of Abaucán Valley

The southern sector of the Abaucán Valley encompasses an inter-mountain region at the border of Catamarca and La Rioja provinces. It stretches from the Batungasta archaeological site (27° 53′19″ – 67°43′15″) in the north to the inter-province border between Catamarca and La Rioja provinces in the south. The western boundary is marked by Cerro Negro de Rodríguez-Sierra de Famatina, and the eastern boundary is defined by the Sierras de Zapata-Vinquis (Geological Map 14 d, Tinogasta, and Geological Map 13c, Fiambalá) (Sosic, 1972). Figure 1 illustrates the study area and its key archaeological sites.

Geologically, the study area is divided into two provinces: (1) the Sierras Pampeanas (present in Sierra de Zapata, Copacabana, and Fiambalá) and (2) the Famatina or transpampean province (found exclusively in Sierra de Narváez) (Figure 1).

To the east of the study area lies the Zapata Mountain Range, composed of a large granitic formation with fine, medium, and coarse grain. This formation includes feldspar, quartz, muscovite, biotite, and microcline minerals. Some isolated outcrops of aplite are also present (Sosic, 1972; Toselli et al., 1992).

The southern foothills of Sierra de Fiambalá, located in the north of the study area, are dominated by the “Los Ratones Granite” Formation, characterized by coarse-grained granites containing alkaline pertitic feldspars, quartz, plagioclase (PF), and biotites (Hongn et al., 2010; Rubiolo et al., 2003).

The Sierra de Copacabana features a complex geological composition comprising three formations: (1) a central formation associated with the archaeological sites, composed of granites with a porphyritic texture and a matrix of quartz, PF, microclines, biotite, muscovite, titanite, zircon, and apatite (Fauqué & Caminos, 2006); (2) a metamorphic formation composed of dark-colored gneiss, eye gneiss with feldspar porphyroblasts, amphibolites, and quartz-feldspathic schists with light and dark banding (Fauqué & Caminos, 2006); and (3) a metamorphic formation containing mylonitic rocks such as mylonitic schist, blastomylonites, mylonitic gneisses, breccia belts, cataclastic microbreccias, prothomylonites, and mylonites (Fauqué & Caminos, 2006).

Finally, the Sierra de Narváez comprises six geological formations primarily composed of sedimentary rocks (Cisterna, 1992; Fauqué & Caminos, 2006; Sosic, 1972). These formations include the Suri Formation, Vinchina or Tambería Formation, Crestón Formation, Patquía-La Cuesta Formation, El Abra Formation, and Costa de Reyes Formation. These formations consist of sandstones, conglomerates, shale, evaporites, and pyroclastics.

In our survey and study, we have examined 25 archaeological sites, of which 10 feature surface architecture and grinding artifacts. An additional five sites are characterized by the present agricultural structures. Among the ten sites with surface architecture, one is assigned to the Middle Period (c. AD 500 – AD 900), two sites belong to the Inca period (c. AD 1480 – AD 1530), and seven sites date to the Early Period (c. 600 AC – AD 500). The remaining sites belong to the Late Period and display abundant surface ceramic materials (De La Fuente, 2011a, b). One Inca site, known as Costa de Reyes No 5, was partially excavated during the 1960s and classified as a tambo linked to the Inca Road system. In addition, around ten pottery kilns of various forms associated with fired walls and signs of vitrification were detected near this archaeological site (De La Fuente & Vera, 2016, 2023, Vera et al., 2019). Previously research carried out during the last 10 years has allowed us to technologically characterize the ceramic materials (De La Fuente, 2011a; De La Fuente & Carreras, 2010; De La Fuente et al., 2010a, b). Primary and secondary forming techniques of vessels also were studied through macroscopic and microscopic analyses (X-ray Industrial Radiography and ceramic petrology) (De La Fuente, 2011a, b). In addition, experimental replications of specific ceramic forms, such as bowls, were conducted to gain a comprehensive understanding of the forming techniques involved. Finally, SEM-EDS and Raman microespectroscopy analyses were employed to examine the paintings and pigments used by ancient potters for vessel decoration (De La Fuente et al., 2010b). Firing technology and temperatures were studied in detail for a specific sample (De La Fuente & Vera, 2023; Rasmussen et al., 2012).

To establish a relative chronology for the area, we have dated with thermoluminiscence 68 ceramic fragments from 9 archaeological sites, mainly Inca and Late Period sherds (Sanagasta culture). This approach, particularly effective for Late Period ceramics, has enabled us to divide the Late Period into three distinct time blocks (De La Fuente et al., 2010a; Vera et al., 2019, Table 1). Provenance analyses utilizing INAA on an extensive sample of sherds and clay deposits pointed out that Late Period pottery production involved several different sources located in the main water streams, with some pottery from this period being imported into the area. In contrast, Inca pottery demonstrates a more tightly defined compositional group, primarily associated with Inca Provincial sherds alongside some Late Period fragments, suggesting a controlled use of clay sources in the region (De La Fuente et al., 2015). Ceramic petrography used in this context has helped us to understand the technological choices done by potters in these two different cultural periods, and how them have partially disrupted through time (De La Fuente et al., 2015; Vera et al., 2019). Also, ceramic petrography data have contributed to highlight the chemical patterns observed in provenance analyses (De La Fuente et al., 2024).

5 Ceramic Sample and Methods

Ceramic petrography was conducted on 93 pottery sherds originating from the Late and Inca periods, sourced from 8 archaeological sites located in the southern sector of the Abaucán Valley (Figures 1 and 2). Sherd main characteristics like style, site, and cultural period are detailed in Table 1.

Thin sections were meticulously prepared at the Laboratory of Petrology and Ceramic Conservation within the School of Archeology at the National University of Catamarca. These sections were then analyzed using an Enosa M-80-P2 polarized microscope, with magnifications ranging from 40× to 80×. The analysis focused on three crucial elements: (1) the matrix, (2) nonplastic inclusions, and (3) cavities.

Matrix analysis encompassed several attributes, including refractory character (the material’s ability to transmit light), classified as isotropic or anisotropic; chromatic aspect, referencing the Munsell color table; and matrix background, describing its composition and the mineral elements 15 μm or less (such as quartz, micas, or their combinations).

For nonplastic inclusions, attention was devoted to their types (larger than 15 μm), distinguishing between minerals and rocks. This differentiation relied on reference samples from the Laboratory of Petrology and Ceramic Conservation, as well as rock and mineral manuals. Other factors examined were granulometry, orientation, degree of sphericity, and distribution of inclusions. Cavities were scrutinized based on their presence, orientation, proportion, and shape.

Quantitative analysis of paste composition involved the JMicrovision v1.3.4 program, which facilitated modal distribution observation of inclusions, cavities, and matrix through a Point Counter (with a minimum count of 300 points per cut), generating percentage data. Granulometry data for nonplastic inclusions were also collected using the same program.

Morphological studies were conducted using the previously recorded information and partially published elsewhere (73 completed vessels) (De La Fuente, 2011a). The methodology used for traceological analysis is based on the classification proposal of manufacturing traces by Garcia Rosselló and Calvo Trías (2013). It consists of two stages: the first involves the characterization of surface and internal marks and the second involves interpretation.

First, a distinction was made between direct and indirect traces (family), followed by an analysis of the qualities or properties of each mark based on 12 attributes (shape, texture, appearance, trend, disposition, distribution, structure, location, surface location, margins, section, association, overlap, and dimension). Each mark was assigned a specific variable for characterization (Garcia Rosselló & Calvo Trías, 2013, p. 132).

The identified traces in the sample included:

1. Concavity and convexity variation: Recognized by uneven thickness forming a wall with a wavy profile, where concave and convex zones alternate (Garcia Rosselló & Calvo Trías, 2013, p. 148).

2. Surface appearance: The visual appearance of the outer layer of the vessel’s surface (Garcia Rosselló & Calvo Trías, 2013, p. 195).

3. Burrs: Clay residues protruding from the vessel’s surface as a relief (Garcia Rosselló & Calvo Trías, 2013, p. 188).

4. Bands: Shiny ribbons visible only when the vessel is subjected to oblique light (Garcia Rosselló & Calvo Trías, 2013, p. 208).

5. Grooves: Hemispherical depressions observed on the ceramic surface (Garcia Rosselló & Calvo Trías, 2013, p. 175).

6. Edges and projections: Lines resulting from the intersection of two surfaces. It can form a simple angle on its outer part or a projection on a uniformly profiled surface (Garcia Rosselló & Calvo Trías, 2013, p. 159).

The second stage of the analysis involves an inferential process with interpretative intentions for each trace, derived from the presentation and analysis of the described traces. The following categories were used: Phase (Ph): refers to the different stages through which clay passes in relation to water loss due to manipulation throughout the crafting process. Phase I: natural state clay; Phase II: mixing and preparation; Phase III: plastic state clay; Phase IV: initial drying; Phase V: leather-textured clay; Phase VI: secondary drying; Phase VII: dry state clay; Phase VIII: firing; Phase IX: cooling; Phase X: finishing. All these traces identified are interpreted into two different frameworks with several levels of inclusivity: (1) the technological framework process (TFP): refers to the different manufacturing processes that a vessel goes through, organizing the manufacturing sequence, such as primary modeling, primary surface treatment, secondary modeling, and secondary surface treatment; and (2) the detailed technological process (DTP): the set of technical actions aimed at transforming raw materials, shaping, and modifying the surface of a vessel.

Provenance studies were conducted using magnetic susceptibility/thermoluminescence (SUS/TL) techniques on a selected sample of 68 sherds. Samples were selected following stylistic criteria, and they come from eight archaeological sites from the area under study. Sherds are from: (1) Costa de Reyes N° 5 (17 sherds), (2) SaCat13 (13 sherds), (3) SaCat12 (5 sherds), (4) GPS042 (6 sherds), (5) BP (7 sherds), (6) CV2 (11 sherds), (7) CV5 (5 sherds), and Río Colorado (4 sherds). Sherds were cleaned, and the present slips, if any, were carefully removed either from the external and internal sides of each sherd. Magnetic susceptibility measurements utilized approximately 1 g powdered samples with a KLY-2 susceptibility meter. The TL/OSL-system TL-DA-12, manufactured at Risø National Laboratory in Denmark, was employed for TL measurements.

Rasmussen’s method (2001) discriminates the provenance of archaeological pottery by measuring two intrinsic physical “fingerprints” of the clay – magnetic susceptibility and TL sensitivity – and classifying them in a two-dimensional plot. Susceptibility reflects the amount and proportion of iron-bearing minerals (mainly hematite, with minor magnetite) in the original clay and subsequently in the fired pottery sample. On the other hand, the TL signal is integrated between 202 °C and 235 °C (the range that maximizes discrimination between clay sources) and normalized to a 10 mg sample mass. Thus, variations in mineral composition and trapped-electron density cause clays from different deposits to cluster in separate regions, allowing reliable assignment of origin even from samples as small as 0.01 g.

6 Results

6.1 General Characteristics of the Petrographic Sample

Our point counting study revealed that, across all analyzed cases, the matrix predominates, with an average of 75.95%, a standard deviation of 7.53%, and ranges from a minimum of 54.33% to a maximum of 91.67% (Figure 3a).

Figure 3

(a) Relationships between matrix, nonplastic inclusions, and cavities and (b) distribution of nonplastics inclusions in the whole sample (n = 93).

Second, nonplastic inclusions account for 15.10% of the total, a median of 14.08%, a standard deviation of 7.14%, and measurements ranging from a minimum of 2.67% to a maximum of 37.16%. Finally, cavities represent the smallest proportion, with an average of 8.88%, a median of 8.33%, a standard deviation of 5.04%, and measurements spanning from a minimum of 0.60% to a maximum of 22.38% (Figure 3a).

6.3 Ceramic Styles and Petrography

6.3.1 Inca

The petrographic sample associated stylistically with the Inca consists of 27 thin sections from 6 archaeological sites, CR5 (n = 11), GPS042 (n = 7), SaCat 02 (n = 1), SaCat 13 (n = 6), SaCat 14 (n = 1), and SaCat 15 (n = 1) (Figure 4a).

Figure 4

Percentages of nonplastic inclusions identified for each ceramic style and distribution per site: (a) Inca, (b) Sanagasta (Late Period), and (c) Belén (Late Period).

In total, 17 types of nonplastic inclusions were identified, with a prevalence of felsic minerals such as CQ at 40%, followed by PF at 8.28%, and a minor presence of PQ at 1.08% (Figures 4a, 5a, 6a, 7b, c, e, and 8a, e, f; Table 2). Plutonic igneous rocks (IgFr(p)) were identified at 11.86% and groggrog (Gr) at 7.19%. In smaller quantities, muscovite (M) accounted for 5.12% and volcanic rocks (IgFr (v)) at 4.64%. Nine other types of nonplastic inclusions were observed in proportions equal to or less than 4%.

Figure 5

Photomicrographs of ceramic thin sections of (a) Inca and (b) and (c) Late Period ceramic sherds, showing grog temper and different mineral and rock fragment inclusions. XPL (left) and PPL (right). Magnification 40×.

Figure 6

Photomicrographs of ceramic thin sections of (a) Inca, and (b) and (c) Late Period sherds, showing grog temper and different mineral and rock fragment inclusions. XPL (left)and PPL (right). Magnification 40×.

Figure 7

Photomicrographs of ceramic thin sections of Inca sherds (b), (c), (e), and (f) and Late Period sherds (a) and (d). Different types of “grog” inclusions are shown together with igneous and metamorphic rock fragments. XPL. Magnification 40×.

Figure 8

Photomicrographs of ceramic thin sections of Inca sherds (a), (e), and (f) and Late Period sherds (b), (c), and (d). Grog and sand tempered pottery sherds. XPL. Magnification 40×.

Table 2

Comparative petrographic data for ceramic styles (n = 93)

	Inca (n = 27)	Sanagasta (n = 45)	Belén (n = 15)	Non-determined (n = 3)	Diaguita -Inca (n = 1)
CQ	40.00%	45.07%	43.54%	14.29%	11.54%
PQ	1.08%	2.52%	3.35%	0.00%	0.00%
PF	8.28%	10.86%	11.36%	3.26%	5.38%
KF	1.23%	0.53%	0.50%	0.00%	0.00%
M	5.12%	5.47%	9.62%	0.00%	0.00%
B	4.11%	4.85%	5.11%	1.22%	0.00%
Ca	1.06%	1.84%	0.00%	0.40%	0.00%
IgFr (p)	11.86%	9.68%	10.82%	12.25%	6.92%
SedFr	1.97%	4.63%	0.88%	23.27%	0.00%
MtFr	1.48%	0.74%	0.00%	0.00%	0.00%
IgFr(v)	4.64%	3.39%	8.04%	0.00%	34.61%
Ad	2.24%	0.55%	1.17%	3.27%	33.83%
Gr	7.19%	1.07%	1.19%	32.65%	0.00%
ArcInc	3.34%	4.16%	0.84%	8.57%	0.00%
VG	2.00%	0.02%	0.00%	0.00%	0.00%
An	3.74%	3.24%	3.00%	0.82%	0.78%
Pir	0.64%	0.52%	0.45%	0.00%	0.78%
MO	0.00%	0.05%	0.00%	0.00%	6.16%
Indet.	0.00%	0.81%	0.14%	0.00%	0.00%

6.3.2 Sanagasta

For Sanagasta, the sample comprises 45 thin sections from 11 sites (BP [n = 4], CR5 [n = 16], CV2 [n = 6], CV5 [n = 4], CV6 [n = 2], GPS042 [n = 2], RC [n = 2], SaCat03 [n = 4], SaCat04 [n = 1], SaCat12 [n = 1], and SaCat13 [n = 3]; Figure 4b).

Nineteen types of nonplastic inclusions were observed, with a majority of felsic minerals (CQ [45%], PF [11%], and PQ [3%]) and plutonic rocks (10%) (Figures 4b, 6b, c, 7d, 8b–d; Table 2). In smaller quantities, muscovite (5%), biotite (5%), sedimentary rocks (5%), clay inclusions (4%), volcanic rocks (3%), and andesites (3%) were identified. The remaining nine types of inclusions were observed in proportions equal to or less than 1%.

6.3.3 Belén

Belén-style ceramics consist of 15 fragments from 7 archaeological sites (CV2 [n = 4], CR5 [n = 3], SaCat13 [n = 3], GPS042 [n = 2], BP [n = 1], CV5 [n = 1], and SaCat12 [n = 1]; Figure 4c).

In total, 15 types of nonplastic inclusions were identified, highlighting felsic minerals such as CQ (44%), PF (11%), PQ (3%), plutonic rocks (11%), muscovite (10%), volcanic rocks (8%), and biotite (5%) (Figures 4c, 5b, c, 7a, 8; Table 2). In quantities less than 4%, the remaining eight types of inclusions were identified.

The whole sample was further examined by multivariate statistics. A cluster analysis was performed using Minitab 18 package (Ward method, Complete linkage, CI = 0.768). Figure 9 presents the cluster obtained showing two main petrographic groups (A and B) and 7 petrographic subgroups (Group 1–Group7). Table 3 presents detailed information on each group composition. Petrographic groups are composed as follows:

Figure 9

Cluster analysis of petrographic data, showing the seven groups formed.

Table 3

Data for petrographic groups formed by cluster analysis (n = 93)

	A			B
	G1 (n = 14)	G2 (n = 11)	G3 (n = 8)	G4 (n = 4)	G5 (n = 22)	G6 (n = 20)	G7 (n = 14)
CQ	48.87%	46.32%	17.71%	34.37%	45.69%	45.30%	43.34%
PQ	0.36%	1.79%	0.00%	8.09%	5.23%	1.57%	0.00%
PF	11.60%	11.62%	3.87%	9.62%	12.45%	12.37%	5.73%
KF	0.00%	1.85%	0.00%	0.97%	0.26%	2.32%	0.00%
M	6.22%	8.65%	0.00%	14.97%	8.72%	4.04%	0.32%
B	10.32%	4.26%	0.45%	6.43%	3.93%	6.25%	0.32%
Ca	0.92%	0.17%	4.47%	0.00%	0.12%	1.13%	1.27%
IgFr (p)	6.30%	5.12%	7.26%	11.38%	7.87%	5.48%	33.14%
SedFr	0.12%	0.00%	25.32%	0.00%	0.00%	0.00%	0.96%
MtFr	0.60%	1.28%	1.64%	0.00%	0.23%	0.66%	0.32%
IgFr(v)	3.75%	4.73%	5.75%	13.23%	4.23%	4.88%	4.75%
Ad	0.00%	0.00%	7.11%	0.00%	0.00%	1.16%	5.90%
Gr	2.50%	4.47%	14.29%	0.93%	2.31%	3.90%	0.96%
ArcInc	1.67%	1.62%	8.34%	0.00%	5.41%	3.50%	2.07%
VG	0.00%	2.06%	0.00%	0.00%	0.00%	1.88%	0.00%
An	5.79%	4.97%	0.58%	0.00%	3.13%	5.07%	0.29%
Pir	0.98%	1.12%	0.13%	0.00%	0.43%	0.48%	0.00%
MO	0.00%	0.00%	1.17%	0.00%	0.00%	0.00%	0.00%
Indet.	0.00%	0.00%	1.93%	0.00%	0.00%	0.00%	0.64%

Group 1: Composed of 14 samples, originating from the SaCat12 site (n = 2), GPS042 (n = 1), Costa de Reyes No 5 (n = 4), CV5 (n = 2), Barranca de la Puntilla (n = 2), CV2 (n = 1), and SaCat03 (n = 2). The sample consists of ceramic fragments in the Sanagasta (n = 9), Belén (n = 2), and Inca (n = 3) styles. Among the nonplastic inclusions, there was a high presence of felsic minerals such as CQ (48.87%) and PF (11.60%), as well as biotite minerals (10.32%). In lesser quantities, muscovite minerals (6.22%), plutonic rocks (6.30%), and amphibole (5.79%) were identified. Other inclusions were found in proportions below 5%.

Group 2: Composed of 11 samples from the SaCat13 (n = 3), SaCat14 (n = 1), SaCat15 (n = 1), GPS042 (n = 2), Costa de Reyes No 5 (n = 1), Barranca de la Puntilla (n = 2), and CV5 (n = 1) sites, consisting of fragments from the late period (Sanagasta [n = 4], Belén [n = 2]), and Inca (n = 5). This group is characterized by the abundance of CQ (46.32%) and PF (11.62%), while muscovite (8.65%) and plutonic rocks (5.12%) were observed in lower proportions, with other types of inclusions below 5%.

Group 3: With a total of eight specimens, all from the Costa de Reyes No 5 site, this group includes a greater number of late period fragments such as Sanagasta (n = 6), as well as a minor presence of an Inca fragment and a Diaguita Chileno fragment. This group stands out for the high presence of sedimentary rocks (25.32%) and grog (14.29%), in addition to CQ (14.71%). In proportions between 5% and 10%, there were identifications of plutonic rocks (7.26%), volcanic rocks (5.75%), andesites (7.11%), and clayey inclusions (8.34%).

Group 4: Group 4 consists of only four cases, from the SaCat13 (n = 2), SaCat12 (n = 1), and CV5 (n = 1) sites. Besides the high presence of CQ (34.37%) and PF (9.62%), the most representative inclusions are vulcanites (13.23%), muscovite (14.97%), and plutonic rocks (11.38%), while PQ (8.09%) and biotite (6.43%) were observed in smaller quantities.

Group 5: Represented by 22 specimens from CV6 (n = 1), Río Colorado (n = 2), Costa de Reyes No 5 (n = 5), SaCat13 (n = 3), GPS042 (n = 5), CV5 (n = 1), CV2 (n = 3), SaCat03 (n = 1), and SaCat04 (n = 1) sites. Most of the fragments stylistically belong to the late period (Sanagasta [n = 14] and Belén [n = 5]), with a smaller quantity from the Inca period (n = 3). Similar to Group 2, this group also exhibits a high representation of felsic minerals, with CQ at 45.69% and PF at 12.45%, while plutonic rocks (7.87%), muscovite (8.72%), clayey inclusions (5.41%), and PQ (5.23%) are present in lower proportions.

Group 6: Group 6 consists of 20 cases from CV6 (n = 1), SaCat13 (n = 4), GPS042 (n = 3), Costa de Reyes No 5 (n = 4), CV5 (n = 2), Barranca de la Puntilla (n = 1), CV2 (n = 4), and SaCat03 (n = 1) sites. Unlike the previous groups, there is a predominance of fragments classified stylistically as Inca (n = 11), although late period styles are also present (Sanagasta [n = 8] and Belén [n = 1]). In this case, CQ minerals (45.30%) and PF (12.37%) prevail, while biotite (6.25%), plutonic rocks (5.48%), and amphibole (5.07%) are identified in smaller proportions.

Group 7: Finally, Group 7 is composed of 14 fragments, all from the Costa de Reyes No 5 site, with Sanagasta styles (n = 8) being the majority, followed by Inca (n = 4) and Belén (n = 2). In this case, a high presence of CQ (43.34%) and plutonic rocks (33.14%) was observed, while andesites (5.90%) and PF (5.73%) were present in minor proportions.

6.4 Provenance by SUS/TL

Provenance analysis was carried out by applying a novel approach using SUS/TL to a selected sample of 68 sherds (Rasmussen, 2001).

The first results indicate that SUS/TL method can differentiate samples at an inter-basin level, for example, between the Abaucán river basin and the Zapata River basin (Figure 10a–e). Samples from sites CV2 and CV5, which are located at Zapata River basin, are easily geographically discriminated from the Abaucán river basin samples. Several Late Period samples from sites like SaCat13 and SaCat12 are discriminated at the inter-basin and site level (Figure 10d and e). Also, the same situation is observed when samples from CR5 are plotted with these two parameters (Figure 10c).

Figure 10

SUS/TL provenance analysis: (a) the whole sample, (b) identification of Zapata River basin sites sherds, (c) CR5B vs CV5 sites sherds, showing a clear separation, (d) SaCat13, CV5, and CV2 samples, clearly discriminated in the SUS/TL plot, (e) plot of sherds from intra-basin sites geographically located at west of Copacabana Mountain, and (f) sherds plotted according to the ceramic styles, showing discrimination between Late Period and Inca samples. In violet color are the two Diaguita Inca sherds.

At an intra-basin level, for example, within the Abaucán basin in the southern sector of the Abaucán Valley (Las Higueritas River, Colorado, Costa de Reyes, de la Costa), it is difficult to distinguish the origin of pottery at the site level, although two groups and several outliers can be visualized (Figure 10e). Table 4 presents all the SUS/TL data.

Table 4

SUS/TL data for provenance analysis (n = 68)

Lab sample number	Sample number	SUS	1σ	TL	1σ	Lab sample number	Sample number	SUS	1σ	TL	1σ
KLR-7572	DLF-002	18.12	0.04	2.5	0.7	KLR-7606	DLF-123	26.07	1.3	85	6.3
KLR-7573	DLF-005	5.51	0.13	20.8	4.2	KLR-7607	DLF-127	3.3	0.1	3.3	0.6
KLR-7574	DLF-006	19.59	0.17	7.6	2.3	KLR-7608	DLF-131	18.8	0.6	5.3	1
KLR-7575	DLF-015	7.28	0.02	3.5	0.2	KLR-7609	DLF-132	3.02	0.04	7.3	1.9
KLR-7576	DLF-011	1.74	0	9.5	1.3	KLR-7610	DLF-134	15.79	0.46	6.2	1.9
KLR-7577	DLF-032	3.11	0	8.9	2.2	KLR-7611	DLF-139	9.45	0.16	7.3	1.2
KLR-7578	DLF-033	3.83	0.01	9.5	0.6	KLR-7612	DLF-140	10.05	0.31	6.9	1.5
KLR-7579	DLF-038	5.15	0.01	5.1	1.5	KLR-7613	DLF-141	5.08	0.03	6.3	1.3
KLR-7580	DLF-039	8.2	0.02	13.2	4	KLR-7614	DLF-143	20.5	0.7	2.7	1
KLR-7581	DLF-052	1.77	0.01	6.5	0.7	KLR-7615	DLF-148	15.53	0.43	5.9	1.6
KLR-7582	DLF-053	10.36	0.01	9.9	0.6	KLR-7616	DLF-155	2.32	0.07	5.6	0.9
KLR-7583	DLF-054	10.9	0.01	5.3	0.7	KLR-7617	DLF-156	2.77	0.07	19.7	8.9
KLR-7584	DLF-056	15.55	0.02	7.4	0.6	KLR-7618	DLF-158	51.4	3.51	9.5	1.2
KLR-7585	DLF-057	11.37	0.56	12.6	2.8	KLR-7619	DLF-160	15.21	1.06	33.9	3.5
KLR-7586	DLF-059	2.34	0.05	5.5	0.6	KLR-7620	DLF-167	48.48	3.9	34.9	4.9
KLR-7587	DLF-063	10.49	0.42	7.2	2.7	KLR-7621	DLF-175	1.16	0.03	2.9	0.9
KLR-7588	DLF-065	8.03	0.14	10.5	1.7	KLR-7622	DLF-178	3.32	0.05	4.1	0.9
KLR-7589	DLF-066	1.32	0.06	4.4	1.5	KLR-7623	DLF-180	11.37	0.21	0.7	0.5
KLR-7590	DLF-067	13.66	0.01	3.9	0.5	KLR-7624	DLF-181	16.33	0.98	12.1	1
KLR-7591	DLF-081	15.29	0.65	4.2	1.7	KLR-7625	DLF-184	9.54	0.26	1.6	0.4
KLR-7592	DLF-075	4.98	0.09	19.4	4.8	KLR-7626	DLF-183	2.36	0.03	6.3	1.2
KLR-7593	DLF-070	21.41	0	5.9	2.2	KLR-7627	DLF-188	8.59	0.34	16.8	1.9
KLR-7594	DLF-084	9.88	0.47	3.5	0.8	KLR-7628	DLF-194	7.62	0.26	7.4	3
KLR-7595	DLF-087	4.31	0	2.6	0.6	KLR-7629	DLF-199	3.63	0.12	1.3	0.3
KLR-7596	DLF-089	1.46	0.03	9.7	1.7	KLR-7630	DLF-200	3.03	0.06	2.3	0.8
KLR-7597	DLF-094	3.71	0.11	4.7	3.7	KLR-7631	DLF-203	29.89	0.73	3.5	0.7
KLR-7598	DLF-095	11.81	0.42	5.3	1.9	KLR-7632	DLF-204	6.22	0.07	1.3	1
KLR-7599	DLF-098	6.45	0.16	2.3	0.4	KLR-7633	DLF-211	4	0.05	4.5	0.9
KLR-7600	DLF-104	11.29	0.27	3.3	0.7	KLR-7634	DLF-212	12.04	0.25	2.6	0.3
KLR-7601	DLF-105	5.86	0.11	3.2	1.1	KLR-7635	DLF-216	15.26	0.57	1.3	0.5
KLR-7602	DLF-108	3.03	0.07	2.6	0.6	KLR-7636	DLF-220	1.35	0.02	3.7	1
KLR-7603	DLF-109	13.76	0.34	4.2	0.4	KLR-7637	DLF-226	2.85	0.02	6.4	1.3
KLR-7604	DLF-110	11.14	0.39	6.5	2.8	KLR-7638	DLF-227	2.8	0.14	1.7	0.4
KLR-7605	DLF-111	27.66	0.03	0.9	0.1	KLR-7639	DLF-228	14.17	0.24	13.4	2.8

When we plot the samples based on cultural periods, we observe the following: (1) the Diaguita-Inca samples are completely separated from the local ones, (2) Late Period samples exhibit significant variability, with only one distinct cluster (shared with the Inca), and (3) Inca samples show at least three groups, two of which are shared with Early Period samples, suggesting the hypothetical presence of three different clay sources (Figure 10f).

The SUS/TL results obtained in this study show a strong correspondence with the INAA findings reported in the study by De La Fuente et al. (2015), yet each method emphasizes complementary aspects of ceramic provenance and production. Both approaches indicate that local clay sources were predominantly used during both the Late and Inca periods, but they also reveal nuances in raw material selection and standardization across these cultural phases.

De La Fuente et al. (2015) employed INAA to characterize the geochemical composition of ceramic sherds, revealing that Late period ceramics exhibited a high degree of variability in their clay sources. This variability suggests that potters during the Late period had access to – and exploited – a wide range of locally available clays, leading to distinct compositional groups. In contrast, the INAA data showed that Incaic ceramics tended to be more homogeneous, which is interpreted as evidence of a more controlled and standardized clay procurement and production process during the Inca period.

Similarly, the SUS/TL approach in the current study reinforces these interpretations but from a slightly different analytical perspective. By measuring magnetic susceptibility and thermoluminescence sensitivity, SUS/TL was able to differentiate samples at both inter-basin and, to some extent, intra-basin levels. Notably, SUS/TL results demonstrated that ceramics from the Abaucán river basin could be clearly separated from those of the Zapata River basin. This spatial discrimination supports the idea that, during the Late period, a more diverse set of clay sources was in use, while Inca ceramics – forming more tightly clustered compositional groups in the SUS/TL plots – reflect a deliberate selection of raw materials.

Moreover, while INAA provided detailed elemental fingerprints that allowed for the identification of specific clay sources and suggested multiple sources in the Late period versus a narrower selection in the Inca period, SUS/TL adds value by directly linking the physical properties of the ceramics (e.g., the concentration of iron-bearing minerals such as hematite) to their raw material origins. In essence, SUS/TL not only corroborates the variability versus standardization pattern seen with INAA but also enhances our understanding of the mineralogical contributions to the ceramic paste.

The convergence of these independent methods strengthens the overall argument for technological transmission across the Late to Inca transition. Both datasets imply that while Late period potters utilized a variety of clay sources – resulting in higher compositional diversity – the Inca period saw the adoption of a more standardized approach. This standardization could be related to administrative controls or a deliberate technological choice, possibly reflecting a process of cultural and technological transmission that maintained some elements of local practice (such as the use of grog temper) while promoting uniformity in raw material selection.

In summary, the detailed comparison shows that INAA and SUS/TL are mutually reinforcing. INAA’s elemental analysis delineates specific clay sources and their variability, while SUS/TL’s sensitivity to mineralogical and magnetic properties adds a spatial and physical context to these findings. Together, they provide a robust multidimensional picture of raw material use, underscoring the transition from diverse, locally sourced practices in the Late period to a more regulated and standardized production system during the Inca period.

6.5 Vessels, Chaînes Opératoires, and Technical Gestures

6.5.1 Common Actions for all Vessels: How to Make a Pot?

The first thing to do before shaping any type of pot is to obtain the raw materials. The primary material in pottery is clay, which is obtained from the soil by digging wells, extracting from ravines, or collecting from riverbeds (Rice, 1981; Rye, 1981). At this point, which marks the beginning of our operative chains, it is important to note that potters may not collect their clays at any time of the year or day; there are specific dates for this purpose as ethnoarchaeological data suggest (Arnold, 1993). Once we have the clay, it is processed into the paste that will be worked with. In this regard, the evidence presented here indicates a poor selection of nonplastic clay inclusions, although the presence of crushed pottery is of great interest.

Clays are usually collected in dry form, so they need to be moistened before they can be worked with. This process is slow, involving hydration in containers over several days (even weeks or months), or a somewhat faster process where the clay is hydrated before use by breaking it into small pieces or finely grinding the raw material “cakes” extracted from the field.

On the other hand, we mentioned crushed pottery (grog) as an interesting nonplastic additive. This is because it is the only one that was, for sure intentionally added. To achieve this, potters must have taken a fired piece or fragment, ground it to generate very small pieces of crushed pottery, which in turn would then have been incorporated into the clay lump used to create new vessels. These clays, along with their additives (including crushed pottery), would have been kneaded for as long as the potter deemed necessary to avoid the presence of large pores that could cause the piece to break in later stages (mainly during firing, a feature observed in the thin-section analyses), while also distributing the inclusions throughout the paste.

Once the clay paste is ready, the shaping of the desired piece begins. The potter already has a mental image of the final product and selects the techniques and tools to be used (Sennet, 2009).

6.5.2 Ceramic Forms, Technological Traces, and Technical Actions

In this section, we present and discuss the results obtained across the traceological study carried out on 73 complete vessels. Table 5 summarizes all the observations done on this sample, and Figure 11a–h highlights some of the identified and recorded traces on the vessels.

Table 5

Macrotraces identified in complete vessels (n = 73)

Morphology	TFP	Trace	DTP	Finality	N	Trace characteristic
Sanagasta bowls (n = 18).	Primary forming	Variation of concavity and convexity	Coiled	Body forming	18	In an elongated form, parallel trend, continuous arrangement, and organized structure. They are located in the internal and external body.y estructura organizada. Se localizan en el cuerpo interno y externo.
		Grooves	Stretched	Base forming	13	Elongated in shape, with a parallel trend, vertical arrangement, organized structure, and without overlapping. Located on the external base of the bowls.
		Burrs	Coiled	Body forming	18	Elongated in shape, parallel trend, horizontal arrangement, grouped distribution, disorganized structure, and without overlapping.horizontal, distribución agrupada, estructura desorganizada y sin solapamiento.
	Primary surface treatment	Surface appearance	Smoothed	Surface homogenization	18	Of parallel trend, continuous distribution, rough texture, organized structure, and located on the interior and exterior of the wall.
	Secondary surface treatment	Stripes	Painted	Decoration	18	Elongated in shape, with a smooth texture, opaque appearance, parallel trend, organized structure, and diagonal arrangement.
Sanagasta urns (n = 51)	Primary forming	Variation of concavity and convexity	Coiled	Body forming	51	Elongated in shape, parallel trend, continuous arrangement, and organized structure. They are located in the internal and external body.
		Grooves	Stretched	Base forming	39	Elongated in shape, with a parallel trend, vertical arrangement, organized structure, and without overlapping. Located on the external base of the urns.
		Burrs	Coiled	Body forming	51	Elongated in shape, with a parallel trend, horizontal arrangement, grouped distribution, disorganized structure, and without overlapping.
	Primary surface treatment	Surface appearance	Smoothed	Surface homogenization	51	Of parallel trend, continuous distribution, rough texture, organized structure, and located on the interior and exterior of the wall.
	Secondary forming	Burrs	Handle attachment	Assembly	51	Circular in shape, with a parallel trend, free arrangement, discontinuous distribution, disorganized structure, and without overlapping. Located at the point of connection between the handle and the body.
	Secondary surface treatment	Edges and projections	Relief decoration	Decoration	3	Circular in shape, with a rough texture, opaque appearance, parallel trend, free arrangement, continuous distribution, organized structure, and located on the external wall.
		Stripes	Painted	Decoration	41	Elongated in shape, with a smooth texture, opaque appearance, parallel trend, organized structure, and diagonal arrangement.
Belén urns (n = 4)	Primary forming	Variation of concavity and convexity	Coiled	Body forming	4	Elongated in shape, with a parallel trend, continuous arrangement, and organized structure. They are located in the internal and external body.
	Primary surface treatment	Surface appearance	Smoothed	Surface homogenization	4	Of parallel trend, continuous distribution, rough texture, organized structure, and located on the interior and exterior of the wall.
	Secondary forming	Burrs	Handle attachment	Assembly	2	Circular in shape, with a parallel trend, free arrangement, discontinuous distribution, disorganized structure, and without overlapping. Located at the junction between the handle and the body.
	Secondary surface treatment	Surface appearance	Burnish	Surface homogenization	2	Of parallel trend, continuous distribution, smooth and homogeneous texture, satin appearance, organized structure, and located on the exterior of the wall.
		Edges and projections	Relief decoration	Decoration	1	Circular in shape, with a rough texture, opaque appearance, parallel trend, free arrangement, continuous distribution, organized structure, and located on the external wall.
		Stripes	Painted	Decoration	4	Elongated in shape, with a smooth texture, opaque appearance, parallel trend, organized structure, and diagonal arrangement.

Figure 11

Macrotraces identified in complete vessels (n = 73) (morphological type – trace – DTP) (a) Sanagasta bowl – variation of concavity and convexity – coiled; (b) Sanagasta urn – variation of concavity and convexity – coiled; (c) detail of Sanagasta urn base – grooves – stretched; (d) detail of Sanagasta urn base and lower body – grooves – stretched – stripes – painted; (e) detail of Sanagasta urn inner body – surface appearance – smoothed; (f) detail of Sanagasta urn middle body – burrs – handle attachment; (g) Belén urn – edges and projections – relief decoration; (h) Sanagasta urn – burrs – handle attachment – variation of concavity and convexity – coiled.

6.5.3 Sanagasta Bowls (n = 18)

6.5.3.1 Primary Modeling

The Sanagasta bowls would begin with the creation of the base using the stretching technique. This technique starts with a lump of clay in a plastic state (phase III), which is stretched consistently and rotationally using the fingers until a base with a wall of approximately 3 cm is formed. This can be observed through the recording of the traces of grooves (n = 13) (Figure 11a). To create the globular body, the TFP of coiling would be implemented, as evidenced by the identification of traces of concavity and convexity variation (n = 18) and burrs (n = 18) (Figures 11a, 12b and c). Depending on the skills, preferences, and traditions of the potters, the coils can be added one at a time (sew each time a new one is added) or every few coils (two or more) (Figure 11a). This step is performed on both the inner and outer sides of the piece to join the different parts composing the vessel, providing greater strength and resistance to prevent breakage at these joints (though they remain its weaker points). This technique involves making coils of clay placed in an overlapping and concentric manner (Figure 11a). As the wall gains height, the coils are stitched together (using hands or a pointed tool) to unify and provide greater strength (Figure 12b and c).

Figure 12

Châine opèratoire sequence for Sangasta bowls (Late Period).

6.5.3.2 Primary Surface Treatment

Once the basic form (simple-profiled globular body) is created, and still in the plastic state (phase III), the next step is surface homogenization. In the case of Sanagasta bowls, the trace “surface appearance” indicates the implementation of the smoothing DTP (n = 18) (Figure 12d). Although there is no traceological evidence of the use of paddling, we consider its possible implementation after the creation of the basic form and surface homogenization.

6.5.3.3 Secondary Surface Treatment

Based on the classification of Macro Technological Processes (TFP) established by Garcia Rosselló and Calvo Trías (2013), there are no techniques corresponding to secondary modeling in Sanagasta bowls. Subsequent technical actions would have decorative purposes and would be carried out after an initial drying (phase IV) until reaching the leather-hard state (phase V). At this point, the applied DTP is painting, evidenced by the “band” trace (N = 18).

The preparation of paints and slips involves grinding clay and minerals (Figure 12e.1 and e.2), mixing them, and adding water (Figure 12e.3). The amount of water and the consistency of paints and slips depend on the preferences and needs of each potter. For polychrome painting, materials such as leather, fabric, wool, or brushes made from vegetable or animal fibers can be used (Figure 12f and g).

6.5.4 Sanagasta Urns (n = 51)

6.5.4.1 Primary Modeling

The crafting of the urns begins in the same manner as the Sanagasta-style pucos described earlier, starting from the base and applying the stretching technique evidenced by the groove’s traces observed in 39 cases (Figure 11c and d). Subsequently, the shaping of the body would be carried out through the TFP of coiling, based on the traces of concavity and convexity variation (n = 51) and burrs (n = 51) (Figure 13b–f).

Figure 13

Châine opèratoire sequence for Sanagasta infant funerary urns (Late Period).

Based on the recorded average size of the urns (40 cm in height), we interpret that the shaping of the urn body would be done gradually and in sections, with intervals of time where each part would be allowed to dry slightly to prevent collapse under the weight: the middle section would be joined to the lower section (the base left to dry) by placing the middle section on the base and connecting both parts inside and outside through stitching (Figure 13h).

With the two parts joined, the final stage of urn construction begins: the raising of the neck and lip (Figure 13i and j). Coils would again be added and sewn, reducing or expanding the opening as desired (Figure 13j). Possibly, paddling would be implemented as an auxiliary technique, especially in the narrower section of the neck, but as mentioned in the previous case, no traces were identified to confirm this hypothesis.

6.5.4.2 Primary Surface Treatment

Once the desired height and shape are achieved, surface homogenization would be performed through smoothing, a technique evidenced by the traces of “surface appearance” (n = 51) (Figures 11e, f, 13e and g).

6.5.4.3 Secondary Modeling

The next crafting stage, still with the clay in a plastic state (phase III), would involve attaching the handles (Figures 11h and 13l). In this case, the burrs observed in all the urns provide evidence of handle attachment in the middle section of the vessel, where a thickening of the section is also observed, allowing us to interpret that “patches” of clay would be crafted to cover the handle/body junction to provide greater stability. There is a possibility that the pieces have a rivet in the internal section, but to confirm this statement, a new analysis of the sample is necessary.

6.5.4.4 Secondary Surface Treatment

At this point, two different traces were identified that can be associated with two decorative techniques. On one hand, in three particular cases, edges and protrusions were recorded, which are related to the DTP of “relief decoration” (pastillage), with the particularity that this technique would be performed with the clay in a fresh state (phase III). On the other hand, once the urns are left to dry (phase IV), until the leather-hard state (phase V), they would be painted, with bands indicating this technique (n = 41), featuring geometric motifs with black paint (Figures 11f and 13o). The paint is applied with fabric, wool, leather, or a brush made of vegetable or animal fibers.

6.5.5 Belén Urns (n = 4)

Only four Belén-style urns were recorded, but based on trace analysis, we consider that the formation process would be similar to those observed for Sanagasta urns (Figure 14b–j): (1) Concavity and convexity variation and burrs are traces observed in the entire sample, associated with coiling; (2) smoothing is the technique observed through surface appearance (n = 4), a technique characteristic of primary surface treatment; and (3) the edges and protrusions trace, characteristic of the “relief decoration” technique, are observed in one Belén-style urn (Figure 11g).

Figure 14

Châine opèratoire sequence for Belén infant funerary urns (Late Period).

The differences noted are as follows: (1) an absence of the furrow trace, which suggests the bases may have been crafted using another technique; (2) we have already described the “surface appearance” associated with smoothing; however, in these cases, we also observe a grouping of the same trace but with particular characteristics (mainly due to its satin-like appearance), which is associated with the FTP of burnishing (n = 2), a technique linked to secondary surface treatment and executed while the piece is in the leather-hard state (phase V); painting is also related to the “bands” trace (n = 4), but in this case, the colors used differ, with the common presence of red on a black base.

6.5.6 Inca Aríbalos

We do not have a representative sample of this type of vessel, typical of the Inca state. However, on the basis of our participation in pottery workshops, we will hypothetically present the modeling sequence of the aríbalos and aribaloides.

6.5.6.1 Technical Actions on Wet Clay (Phase III)

The creation of aríbalos differs from the previous examples, as we hypothesize that the base would be generated toward the end of the process and not at the beginning. This is due to its convex shape, which would prevent the other part of the piece from resting on it unless placed on a flat base support. For this reason, we believe that first, the middle section is modeled, using the coiling technique described earlier for the previous morphologies, as well as the possibility of paddling. Once the shape of the upper two-thirds of the vessel (middle and upper sections) is achieved, the piece would be temporarily placed upside down, using the mouth as support, to attach the base to the rest of the body.

6.5.6.2 Technical Actions on Hard Leather-Like Clay (Phase V)

Once the general shape of the container is complete (closed vessel with a composite profile), the handles would be placed similar to urns, with the difference that the handles would be positioned vertically. In addition to the handles, an appliqué or appendage is placed at the top of the central section, centered with respect to the handles. This appendage plays a crucial role in the use of the vessel, as a cord made of natural fibers passes through the handles and this appendage during transportation. Once the vessel is formalized, it is decorated by applying paints of different colors (mainly black and red), using brushes made with natural plant or animal fibers.

6.5.7 Inca Aribaloides

These vessels would be made in the same way as described for urns, in both the Sanagasta and Belén styles. This would allow raising this section at the beginning and adding the other parts as the manufacturing process progresses. The main manufacturing technique identified is coiling, although the use of paddling to finalize the shapes of different parts of the vessel is not ruled out. The decoration technique used is slip or direct painting on the surface.

7 Discussion

Technological studies done by extensive ceramic petrography, provided new information on the pottery-making practices developed by southern Andean communities from Late and Inca periods at western Catamarca Abaucán Valley (Tinogasta, Catamarca, Argentina).

Ceramic petrography analysis allowed us to obtain seven groups of ceramics, with some similarities, among which the high presence of felsic minerals, mainly CQ, stands out in all ceramic groupings (Figure 9 and Table 3).

When we relate the ceramic styles identified in the sample, we observe relevant characteristics that should be mentioned. Late period ceramics (Sanagasta and Belén) are present in all groupings, while Incaic style ceramics are highly represented in group 6 and 2. On the other hand, when we focus on archaeological sites, we see heterogeneity in their presence, except for the Costa de Reyes No 5 site, where groups 3 and 7 consist exclusively of fragments from this site (Figure 9).

Comparing the local geological characteristics with the types of inclusions observed in thin sections, we notice a significant similarity: felsic minerals (CQ and PFs) are common in igneous, metamorphic, and sedimentary rocks that characterize the mountain ranges in the study area. In addition, plutonic rocks observed in thin sections are present in the Zapata range (granitic batholith) (Sosic, 1972; Toselli et al., 1992), the Fiambalá range (“Los Ratones” Formation) (Hongn et al., 2010; Rubiolo et al., 2003), the Copacabana range (central sector of the mountain range) (Fauqué & Caminos, 2006; Sosic, 1972), and the Narváez range (Abra Formation) (Fauqué & Caminos, 2006; Sosic, 1972). Furthermore, there is a lesser presence of volcanic rocks (andesites and vulcanites), which are observed as principal components of sedimentary clasts in the Narváez range (Vinchina/Tambería Formation and Costa de Reyes) (Fauqué & Caminos, 2006; Sosic, 1972), while sedimentary rocks characterize the Narváez range, located in the western part of the study area (Fauqué & Caminos, 2006; Sosic, 1972).

The results obtained provide indications of the pottery practices during the selection and preparation of raw materials. First, the strong correspondence between local geology and non-plastic inclusions demonstrates that ancient potters likely used readily available raw materials from the environment and near archaeological sites. On the other hand, the ceramic groups identified through cluster analysis suggest seven different choices or recipes for the transition between the late period and the Inca period (Figure 9 and Table 3). Regarding this, Sanagasta and Belén ceramics would be made using at least six different methods of preparing the material, with the voluntary incorporation of various types of non-plastic materials, ranging from sands with a high content of plutonic rocks, volcanic rocks, and/or sedimentary rocks. This is further supplemented by the practice of incorporating crushed pottery (grog) into the ceramics, a type of inclusion that is present in varying densities but becomes more prominent in group 4 (Figure 9 and Table 2).

As for ceramics classified stylistically as Inca, we observe a restriction in the choices of preparation for the ceramics, as they are concentrated in two groupings, G6 and G2. It is plausible to consider the incorporation of sands with a high content of felsic minerals and a moderate presence of micas and plutonic rocks, to which ground pottery is added as a non-plastic, but only partially.

The presence of crushed pottery (grog) is an interesting practice to highlight, a characteristic that is highly prevalent in Late Period ceramics in the regions surrounding the Abaucán Valley (see De La Fuente & Vera, 2023). We can mention, for example, the study of Sanagasta/Angualasto ceramics from the Tambería de Guandacol site, located in the western part of the province of La Rioja, where petrographic studies revealed a sample of 52 fragments (bowls, jars, and pots) with densities of ground pottery ranging from 0.66 to 14.66% (Carosio, 2017, 2018; Carosio & Iniesta, 2017). Furthermore, Belén-style ceramics investigated in both the Hualfín Valley (Belén, Catamarca) and the Bolsón Valley (north of Belén, Catamarca) revealed the presence of ground pottery in similar proportions: in the first case, the sites (El Montículo, Cerro Colorado, Loma de la Escuela, El Molino, Pueblo Viejo del Eje, and Loma de Los Antiguos) consist of ceramic fragments from bowls, pots, and jars with low and medium presence of ground pottery, except for those classified as ordinary, where the presence is higher (Iucci, 2013; Zagorodny et al., 2010); in the second case, the sites under study (La Angostura, El Durazno, and Los Viscos) contain Belén-style fragments with relatively moderate nonplastic densities (ranging from 0.30 to 30%), also with the presence of fragments classified as ordinary, where the density of this inclusion is higher (Puente, 2010; Puente, 2012, 2017). In addition, it should be noted that in the Hualfín Valley and the Bolsón Valley, the analysis samples contain, in lower densities, Santa María-style fragments with the presence of ground pottery in their pastes.

Taking this information into account, if we look back at the petrographic analysis of the southern Abaucán Valley, we can consider the existence of a common practice during the Late period, where potters used grog as a temper (among other inclusions) (Figures 5–8). Because the seven ceramic groups are not exclusively homogeneous, we interpret this as evidence of a possible continuity of practices during the Inca period. Perhaps, these groups represent some kinds of “recipes,” understanding these as a dynamic “know how,” which were in motion during Late and Inca times. Furthermore, we considered the occurrence of a “technological transfer” during the shift from the Late to Inca periods in how artisans formulated ceramic pastes and employed a distinct cultural temper, such as crushed pottery (grog) (De La Fuente & Vera, 2023; Vera et al., 2019).

Similarly, information on the provenance of pottery by SUS/TL in the area shows a local pottery production characterized by present more variability during the Late Period (probably more clay sources used by prehispanic potters) and a more controlled use of clay “sources” during Inca times (Figure 10). But overall, the same sources are in use during the two prehispanic moments. These results agree with those obtained by INAA performed in an extensive sample of pottery from the region under the study (De La Fuente et al., 2015).

When we approach the different chaîne opératoires involved in the elaboration of the vessels (bowls, infant funerary urns, and aríbalos), we see common practices in the forming techniques used by the potters. Coiling is evidenced as the main primary forming technique, whereas smoothing and (possibly) paddling – a wall thickness reduction technique – were applied systematically as secondary forming techniques (Figures 11–14). Interestingly, the three forming steps seen in the infant funerary urns call our attention to the way we visualize the whole production processes of making pottery (Figures 13 and 14). This in turn is related with the “times” of the clay, a more realistic path to comprehend the pottery chaîne opératoire (Garcia Rosselló & Calvo Trías, 2013).

The integration of thin section petrography, SUS/TL provenance analysis, and traceological investigations offers a comprehensive view of ceramic production processes and raw material choices that extends and deepens the existing body of data. Each dataset contributes unique insights that, when combined, reveal both the technical strategies and cultural continuities underlying pottery manufacture in the southern Andean communities.

Thin section petrography provides a detailed characterization of ceramic pastes, highlighting the mineralogical composition and the incorporation of nonplastic inclusions (Table 1 and Figure 9). The predominance of felsic minerals – especially CQ – and the variable presence of accessory materials such as micas and fragments of plutonic, volcanic, or sedimentary rocks echo the local geological background. This correspondence not only confirms the use of locally sourced raw materials but also suggests that potters actively selected clays based on their intrinsic mineral properties. Such findings align with previous studies that have underscored the significance of geological context in shaping paste recipes (De La Fuente et al., 2015; Sosic, 1972).

The SUS/TL analysis complements these petrographic data by offering a geochemical and magnetic susceptibility perspective. This method has successfully differentiated samples at inter-basin levels and, in some instances, at the site level. The observed patterns indicate that while Late Period ceramics exhibit considerable variability in clay source usage – reflecting a flexible, locally sourced production strategy –Inca ceramics tend to show a more restricted and controlled selection of clays. This convergence in clay sources across the two periods, despite differences in variability, suggests that technological transmission was not only a matter of adopting specific tempering techniques (such as the incorporation of grog) but also involved a sustained reliance on common raw material reservoirs.

Traceological analysis adds a further dimension by documenting the physical manifestations of manufacturing processes. The identification of consistent forming techniques – predominantly coiling for primary shaping combined with secondary treatments like smoothing and, where applicable, paddling – confirms that similar technical gestures were employed across different cultural phases (Figures 10–14). Notably, the presence of systematic traces on vessels such as infant funerary urns demonstrates that the procedural steps captured in the traceological record mirror the sequence of actions inferred from petrographic and provenance studies (Figure 10 and Table 5). This alignment underscores the role of technical know-how in maintaining continuity amid broader cultural and technological shifts.

Importantly, a subset of the ceramic samples underwent analysis by more than one technique. For instance, samples that were examined through thin section petrography were also subject to SUS/TL analysis, thereby linking textural and compositional observations with geochemical signatures. The dual analysis of these samples reinforces the interpretation that, despite variability in some manufacturing parameters, a core set of technological choices remained consistent across the Late and Inca periods. This integrative approach highlights that while each analytical method addresses different aspects of ceramic production, their combined application produces a more robust and nuanced understanding of the entire chaîne opératoire.

By juxtaposing these datasets, the current study not only validates earlier compositional studies but also illustrates the dynamic interplay between raw material selection, technological execution, and cultural practices. The observed correspondence between local geology and ceramic composition, when coupled with the detailed records of manufacturing traces and provenance indicators, paints a picture of a production system that was both adaptive and rooted in long-established traditions. This multimethod framework thus offers a template for future investigations aiming to unravel the complexities of ceramic production in other regions, contributing to broader debates on technological transfer and cultural continuity in pre-Hispanic societies.

In summary, the integration of thin section petrography, SUS/TL, and traceological analyses demonstrates that even as technological innovations and stylistic shifts occurred during the transition from the Late to Inca periods, underlying production practices remained largely consistent. This holistic approach enhances our understanding of the ceramic production process, affirming that the technical knowledge transmitted across generations was a critical component of both innovation and cultural identity in the Andean world.

8 Conclusions

Pottery-making practices during Late and Inca times in southern Abaucán Valley (Tinogasta, Catamarca, Argentina) are complex craft processes. The use of a multianalytical technological approach emphasized in the chaîne opératoire has allowed us to highlight several aspects of the organization of the whole pottery production process.

This study, through extensive ceramic petrography analysis, has provided valuable insights into the pottery-making practices of the southern Andean communities from the Late and Inca periods in the western Catamarca Abaucán Valley. The identification of seven distinct ceramic groups, characterized by a high presence of felsic minerals such as CQ, reveals a strong correspondence between local geological features and the raw materials used by ancient potters. These findings suggest that pre-Hispanic potters sourced their materials locally, selecting and preparing them according to specific technological traditions.

Our analysis highlights the complexity of ceramic production during these periods. The Late period ceramics, including Sanagasta and Belén styles, were produced using at least six different preparation methods, incorporating a variety of non-plastic materials such as plutonic, volcanic, and sedimentary rock fragments. In addition, the practice of adding crushed pottery (grog) as temper was prevalent, with variations in density across different ceramic groups. This practice appears to be a defining characteristic of Late period pottery in the region and neighboring areas, reinforcing the notion of a shared technological tradition.

In contrast, Inca ceramics demonstrate a more restricted selection of paste preparation techniques, predominantly concentrated in two specific groupings. This suggests a more standardized production process during the Inca period, with a controlled use of raw materials and specific tempering methods. However, the continuity of certain Late period practices, such as the use of grog, implies a degree of technological transfer between the two periods, reflecting both adaptation and persistence within local pottery traditions.

Further evidence from provenance studies using SUS/TL and INAA supports the interpretation of local ceramic production, with greater variability in clay sources during the Late period and a more controlled selection during the Inca period. The observed continuity in clay sources further reinforces the idea of technological transmission across these cultural transitions.

Finally, the examination of forming techniques reveals a consistent| use of coiling as the primary method, with smoothing and paddling employed systematically as secondary techniques. The detailed analysis of infant funerary urns emphasizes the complexity of the production processes and offers a deeper understanding of the temporal and procedural aspects of pottery manufacture. These findings contribute to a broader comprehension of Andean ceramic traditions and the dynamic interplay between continuity and change in pre-Hispanic pottery-making practices.