This paper briefly reports on the initial results of a new Leverhulme-sponsored four-year archaeological project (RPG 234-2014) centred on the Átures Rapids area of the Middle Orinoco River, Amazonas State, Venezuela (Fig. 1). The Cotúa Island Reflexive Archaeology Project seeks to establish the longue durée historical processes that by early colonial times culminated in the region’s reputation for being a key crossroads – where diverse ethno-linguistic groups from far-flung regions converged to trade (Oliver et al. 2014). Harnessing new evidence, it aims to elucidate how interaction between such diverse indigenous groups unfolded and the role it played in forging ethnogenesis. Through archaeological research, it seeks to gain new insights into its history and elucidate regional patterns of exchange through the study of technical and stylistic dimensions of material culture. It also aims to investigate the abundant pre-colonial rock art and its relationships to the landscape and aboriginal oral traditions. Finally, it seeks to understand how Western and Non-western archaeological knowledge is produced by engaging with the current indigenous groups as partners in (re)constructing history.

The Orinoco River is the salient physiographic feature that shaped the unfolding of history of the aboriginal groups living along its fringes. It stretches for over 2,200 km from its source in the Parima range of the Guyana uplands debauching into the Atlantic Ocean. Its basin drains an area of 880,995 km² with major tributary rivers originating in the Northern Andes and Guiana uplands, whilst the Casiquiare Canal connects the Orinoco to the Upper Río Negro. Along its course, the main channel traverses biodiverse habitats (Lasso et al. 2010); including dense tropical rain forest in the upper reaches and savannahs with gallery forests in the middle and lower reaches. The Orinoco region currently comprises over 26 indigenous groups (Gassón, 2002) representing a diverse cultural mosaic that is thought to have ancient pre-colonial roots. Several points along the Middle Orinoco have been identified as important trading loci, linking the Llanos of the Orinoco with the Guianas (Morey, 1975; Zucchi and Gassón, 2002). But among these loci, the most notorious is the Átures area. The Berrío Expedition (1584) documents indicate that the Isla de Los Adoles (or Átures Island) was the only area to qualify as a trade centre (Ojer 1960). Two reasons were given. First, the Spaniards repeatedly remarked on the bountiful availability of fish in the Átures Rapids –i.e., a reliable, year-round store of proteins. Second, upriver navigation from the delta finds its first major obstacle for navigation in the rapids, therefore forcing an obligatory detour for all canoe travellers. Thus situated, Átures (see Fig. 1) was an important crossroad for exchange and interaction among different indigenous groups. According to Spanish sources, this area was inhabited by Ádole Indians who regularly traded with Carib-speaking parties from Guiana and with various other ethnic groups from the Venezuelan-Colombian Llanos (plains). The islanders exchanged fish, crops and wood for ceramic vessels, metal weapons, beads and indigenous slaves (Del Rey Fajardo, 1966). Besides the Ádole two other groups, Yaruro and Máipure, inhabited these islands. All along the Middle Orinoco the salient characteristic is the co-presence and intersection of different cultural, ethnic and linguistic groups. A small number of archaeological studies conducted in the Átures region (Cruxent, 1950; Barse, 1989; Gassón 2002) seem to support the notion of a multi-ethnic trading centre in pre-colonial times. This assertion is based on the presence of earlier Saladoid and Barrancoid or later Arauquinoid and Valloid traditions found in the same occupation context and the occurrence of ‘hybrid’ ceramics that share decorative and tempering elements. However, the nature of the social processes (trade, exchange, migration) that may have led to such admixture or ‘hybridity’ of ceramic traditions is still poorly understood. The aboriginal trade dynamics of the 16^th to 18^th centuries, influenced by colonial powers, ought not to be assumed to be similar in pre-colonial times (Zucchi and Gassón 2002). Furthermore, these traditions are suggested to correlate with the distribution of two major linguistic families – Arawakan (Saladoid, Barrancoid) and Carib (Arauqinoid, Valloid) –, which is problematic when extended back into the pre-colonial past (Tarble and Zucchi 1984; Tarble 1985).

The first fieldwork season was conducted in September 2015 at the Culebra site, located at the confluence of the Cataniapo and the Orinoco rivers (see Fig. 1). Systematic surface collections, auger tests and three trench excavations were conducted. A total of 197 lithic artefacts and over 2,800 ceramic sherds, was retrieved. At the base of Trench C, a preceramic component was identified, consisting of crystal quartz flakes and debitage associated with several red ochre features. Although radiocarbon dates are awaited, a similar assemblage from Provinical-1 site (15 km to the north of Culebra) yielded a single uncalibrated date of 9020 ± 100 BP (Barse 1989). The top 20–25cm of Trenches A and B contained the bulk of the ceramic deposition at Culebra, mixed with appreciable amounts of cracked milky quartz. Trenches A-B lacked evidence of a preceramic occupation. The analysis of the data and artefacts is still ongoing whilst the results of charcoal samples submitted for dating are currently being processed. Tentatively, the evidence suggests that Culebra’s occupation consisted of temporary shacks where fisher-folk stayed repeatedly.

In February 2016, some scouting visits to various archaeological sites were undertaken, including Rabo de Cochino Island where an important surface collection was obtained (to be investigated in 2017). However, work focused on Picure Island, where a controlled surface collection and five test units, resulted in 10,594 ceramics, 4,084 lithics and 69 beads. The beads were shaped from exotic stones, some of which were fractured during the manufacturing process and others were unfinished. Following Roux’s (2011) technical traditions methodology, an initial classification was developed from 12,238 excavated potsherds from the Picure and Culebra sites. Based on a preliminary macroscopic analysis of forming techniques and paste preparation recipes (emphasizing type of inclusions, grain size and paste colour), six different fabric groups were identified. Culebra exhibited an additional paste recipe. These paste groups show differentiated use of inclusions/temper: sand, clay pellets, fibre and sponge spicule (or cauxi). The confirmation of such groups will be achieved through micro-petrographic and portable X-ray fluorescence (pXRF) analyses to be conducted later in 2016. At this time, it is premature to confidently determine to which styles and traditions the ceramic assemblages belong. However, diagnostic ceramics of the Barrancoid, Arauquinoid, Valloid and Cedeñoid traditions have been noted, along with others that cannot yet be assigned. To establish their relationship with the broader Orinoco region, the ceramic materials from 23 different archaeological sites in the Orinoco and the Western Llanos were reviewed. These collections are curated at the Instituto Venezolano de Investigaciones Científicas in Caracas and the Enzo Ceccarelli Museum in Puerto Ayacucho.

The corpus of engraved rock in the Orinoco region has received sporadic attention compared to painted rock art (Greer 1995). While older anthropological treatises tend to consider the subject unworthy of serious study, Im Thurn (1883) recognized the potential insights into the beliefs and practices of pre-colonial people. Cruxent (1949) provides tantalizing clues as to the richness of the Átures petroglyphs. Tavera-Acosta (1956), Sujo Volsky (1975) and Dubelaar (1986) synthesised new information, albeit in a highly abbreviated form. These earlier studies provide an excellent backdrop against which to compare the rich landscape of rock art we encountered. To situate the extensive undocumented corpus of rock art of Átures within a broader regional perspective, we initiated a systematic survey of exposed rock faces using computational photographic techniques. A total of nine locations with petroglyphs were included in the basic photographic sample (Fig. 1c). Some individual zoomorphs on Picure have a surface area greater than 5 m² (Fig. 2a), in these cases structure-from-motion photogrammetry derived from low-altitude aerial imagery has successfully reconstructed panels too large or inaccessible for single photographs. Smaller motifs of importance, for instance the depiction of a ritual flute player were subjected to data capture with reflectance transformation imaging (RTI) to capture additional surface detail (Malzbender et al. 2001). Finally, other notable rock art sites such as Cerro Pintado were also visited. Analysis is currently ongoing and a database of comparative rock art is being compiled.

Three activities were undertaken in 2016 with representatives of the Wótuha (Piaroa/Sáliva), Hiwi (Guahibo), Baré (Arawak), Baniba (Arawak), and Puinave (Macú/Caberre) groups. All participants were aboriginal school teachers who taught in indigenous settlements throughout Amazonas State. They had gathered in Puerto Ayacucho for a workshop on ‘Memory, Territory and Citizenship’, organized by the Ministry of Education. None had ever met an archaeologist before. The first activity was a class-room presentation of the aims and objectives of the project which were then roundly discussed as to how our Western aims and methods of (re)constructing ancient history were relevant to what they were learning in the workshop (Fig. 2b). The meeting was followed by a day-long visit to Picure to observe and comment on the archaeological work and the island’s rock art. Based on that field experience, the teachers’ group prepared a written report and PowerPoint presentation that we are currently evaluating. Not surprisingly, most of the participants supported (rather than contested) our Western-generated archaeological knowledge as a means to reaffirm ancestral and traditional territorial rights, but also to proudly display archaeological material evidence of their antiquity (precedence) and of their distinct cultural values, origins and traditions.