Earliest Prepared core technology in Eurasia from Nihewan (China): Implications for early human abilities and dispersals in East Asia

Significance

The alleged simple and long-lasting mode 1 techno-complex in East Asia has long been controversial. Our finding in the Cenjiawan (CJW) assemblage of a prepared core technology, standardized predetermined products, and patterned retouching tool shapes, provides compelling evidence for advanced technical abilities among early Pleistocene hominins in East Asia, significantly earlier than previously assumed. The outstanding record of refit sets in CJW prompts a challenge to the prevailing model of technological stasis in East Asia throughout the lower and middle Pleistocene. It also questions the notion that early Pleistocene sites in China should be viewed as simple core-and-flake assemblages (Oldowan-like/mode 1) due to the absence of handaxes and offer valuable insights into early human technological patterns and dispersal in East Asia.

Abstract

Organized flaking techniques to obtain predetermined stone tools have been traced back to the early Acheulean (also known as mode 2) in Africa and are seen as indicative of the emergence of advanced technical abilities and in-depth planning skills among early humans. Here, we report one of the earliest known examples of prepared core technology in the archaeological record, at the Cenjiawan (CJW) site in the Nihewan basin of China, dated 1.1 Mya. The operational schemes reconstructed from the CJW refit sets, together with shaping patterns observed in the retouched tools, suggest that Nihewan basin toolmakers had the technical abilities of mode 2 hominins, and developed different survival strategies to adapt to local raw materials and environments. This finding predates the previously earliest known prepared core technology from Eurasia by 0.3 My, and the earliest known mode 2 sites in East Asia by a similar amount of time, thus suggesting that hominins with advanced technologies may have migrated into high latitude East Asia as early as 1.1 Mya.

The emergence of the Acheulean/mode 2 techno-complex in Eastern Africa around 1.7 Mya (1–3) marked a significant milestone in human technological evolution. The production of large cutting tools (LCTs) requires sophisticated mental templates and technical abilities of an evolved early Homo, representing a remarkable advancement in human technology (4–10). Evidence from Israel and India suggests that hominins carrying an Acheulean technology may have dispersed from Africa into Eurasia as early as 1.5 Ma (11–13). However, the presence of Acheulean/mode 2 hominins in the eastern part of Eurasia remains a subject of debate. Previous evidence has been used to suggest that East Asia contained a homogeneous and long-lasting record of Oldowan-like or mode 1 technologies before the late Pleistocene, characterized by simple flaking patterns and a lack of standardization in stone tool manufacture (14–17). The alleged cultural resemblance between East Asian Pleistocene assemblages and the Oldowan stands in contrast to the widespread record of Acheulean sites in Africa, Southern Europe, and from India to the Near East. This has led to contentious assumptions such as the Movius line (18) and to models of a largely isolated biological evolution for East Asian hominins in which technological change would be negligible throughout more than a million years. Therefore, a detailed understanding of the particularities of the early archaeological record in East Asia is crucial to inform competing hypotheses on the cognitive, technological, and adaptive capabilities of early humans during their dispersal across the Old World.

Prepared core reduction lithic technologies are characterized by organized methods of flaking end products that require in-depth planning, complex mental templates, and high manual precision (19, 20). Although prepared core technologies and their associated products are well understood for the Middle Paleolithic/Middle Stone Age, recent studies indicate that this ability may have originated in the Acheulean (21–24), particularly during its later stages (1, 19, 25, 26). Thus, growing evidence suggests that relatively simple core preparation may have first appeared around 1.5 Ma in the early Acheulean (7, 27, 28), while longer reduction sequences and more systematically prepared core technologies are documented in later Acheulean sites in South Africa (Canteen Kopje, >1 Ma) (29–31) and Eurasia (Gesher Benot Ya’aqov, 0.8 Ma) (32, 33).

The Cenjiawan (CJW) lithic assemblage in China is characterized by an exceptional refit ratio and thus presents a unique opportunity to reconstruct the specific actions and decisions employed during stone tool reduction, reconstructions that are unattainable at other East Asian early Pleistocene sites. Our study shows the existence of long sequences of core preparation and sophisticated reduction and shaping sequences, thus offering critical insights into the technological abilities of early Pleistocene toolmakers. This may force a reconsideration of current perceptions of technological stasis in East Asia and provide insights into the early dispersals and adaptions of hominins in Eurasia, as well as into the debate on mode 3/ Middle Paleolithic technologies in the region.

Results

CJW Site.

The Nihewan basin, which is situated between the Inner Mongolian Plateau and the Northern China Plain (Fig. 1A), is remarkable for its significant concentration of early Pleistocene sites, probably the densest outside Africa (34–36). The CJW site (41°13′21″N, 114°40′17″E, 869.9 m a.s.l.) (Fig. 1 B and C and SI Appendix, Fig. S1A) is located on the CJW Platform in the eastern margin of the Nihewan basin, near several other important early Pleistocene Paleolithic sites. Excavations between 1986 and 2019 yielded 2,015 lithic artifacts and a large number of mammalian remains (SI Appendix, section 1) (37–43). The CJW archaeological assemblage is positioned in the lower part of the fluvio-lacustrine sequence of the Nihewan Beds (SI Appendix, Fig. S1B) (44), which consists mainly of fine sands, silts and silty clays (Fig. 1D and SI Appendix, Fig. S1C and Table S1). The CJW archaeological layer is contained within a 30 to 40-cm unit embedded in the pre-Jaramillo Matuyama reverse chron (Fig. 1D). This unit follows the Punaruu geomagnetic excursion (45), which yields a ⁴⁰Ar/³⁹Ar age of 1.105 ± 0.005 My (38, 46). Previous studies suggest minimal post-depositional disturbance of the CJW archaeological assemblage (39), allowing for a detailed assessment of hominin behavior at the site.

Fig. 1.

The CJW site. (A) Location of the Nihewan basin. (B) Distribution of relevant early Pleistocene sites on the CJW platform of the Nihewan basin. (C) Aerial view of the CJW site. (D) Lithology of the CJW profile and corresponding magnetic polarity time scale [Modified from Wang et al. (45)]. Abbreviation of sites: MJG, Majuangou; XCL, Xiaochangliang; FL, Feiliang; MDG, Madigou; DGT, Donggutuo; CJW, Cenjiawan; ML, Maliang (see references in SI Appendix, Table S7).

Prepared Core Technology.

The CJW lithic assemblage consists of cores, flakes and flake fragments, retouched tools, and pounded pieces (SI Appendix, Table S2). Raw materials are dominated by local chert (SI Appendix, Table S3) that exhibits considerable variation in color and internal flaws, resulting in a suboptimal flaking quality (SI Appendix, section 2.1). A total of 152 refit sets were identified in the CJW assemblage, consisting of 618 refitted artifacts (SI Appendix, Table S4). Then, 102 of these refit sets can be used to reconstruct reduction sequences (SI Appendix, Table S5), with 14 (13.7%) of these sets providing clear evidence for the complex preparation of cores (Fig. 2B and SI Appendix, section 3).

Fig. 2.

Operational scheme of prepared core technology and predetermined products in CJW. (A) Idealized operational scheme (explanation see detailed description in SI Appendix, section 3.1). (B) Recurrent detachment sequence of predetermined flakes at stage 4 from refit set NS65 (detailed description in SI Appendix, section 3.2). (C) Predetermined flakes produced at stage 4. (D) Targeted flakes detached at stage 5.

Through these refit sets, we have reconstructed an operational scheme of organized flaking that consists of five stages (Fig. 2A), where the detachment of predetermined products (those obtained in stages 4 and 5) was enabled by preparation of both the flaking surfaces (stages 1 and 2) and the striking platforms (stage 3). Each stage has a distinct purpose: first, the guiding ridge on the flaking surface is created (stage 1), and then, the flaking surface is flattened by small removals (stage 2). This is followed by a preparation of the striking platform (stage 3), and the detachment of a larger predetermined flake (stage 4), with the resulting flat flaking surface being used to detach subsequent relatively large and slender flakes (stage 5). The first four stages are observed in most refit sets, while stage 5 is identified in six refit sets and in many non-conjoining end products. Although less often (n = 6), recurrent detachment of predetermined products from the same core also occurs (e.g., NS65 in Fig. 2B and SI Appendix, Fig. S2 A–C). The intersection plane between the striking platform and the flaking surface is well maintained through the knapping sequence leading to the detachment of predetermined products. Once these have been removed, cores are usually exploited further using multifacial methods to maximize flake production. Given their small size, such cores resemble the product of non-hierarchical or unorganized flaking methods in their final stage (SI Appendix, Fig. S4). In the absence of refits enabling reconstruction of the reduction sequence, these small, unstructured cores would give the impression of an unorganized flaking technology such as that supposedly characterizing the early Pleistocene Nihewan assemblages.

In addition to the refit sets, we have identified numerous other products (n = 133, approximately 18% of the flakes and flake fragments) from each of the stages reconstructed in the core preparation operational scheme. While stage 1 typically yields products undistinguishable from those obtained through non-hierarchical flaking methods, stage 2 (n = 27, SI Appendix, Fig. S5 A–F) usually produces elongated flakes with guiding ridges on the dorsal surface, and stage 3 yields smaller flakes with facetted platforms (n = 11, SI Appendix, Fig. S5 G–L). Stage 4 flakes are characterized by a relatively large size and prepared dorsal surfaces (n = 42, including 3 intentionally broken flakes from refit sets; Fig. 2C and SI Appendix, Fig. S6). Stage 5 products (n = 53, which include 22 intentionally broken flakes from refit sets; Fig. 2D and SI Appendix, Fig. S7) are characterized by slender flakes with relatively large scars dominating the dorsal surfaces, which are typically flat and sub-parallel to the ventral surfaces. Close resemblance among products at each stage, coupled with the relatively large and slim predetermined products, underscore the standardization of the entire reduction sequence.

CJW toolmakers’ ability to evaluate the requirements of each individual core and their skill in employing flexible technical repertoires to achieve a standardized reduction are remarkable. While Nihewan basin raw materials are generally of poor quality for stone tool flaking (36–38), CJW knappers consistently applied prepared core schemes to the best available rocks. Flexibility observed in technical actions is conspicuous; minimal preparation is conducted when the natural shape of cores meets demands for detachment of predetermined flakes (for example, some targeted flakes lack either a prepared dorsal surface or platform preparation; SI Appendix, Fig. S6 G–I), whereas adjustments on the standardized operational scheme often occur to meet the particularities of each reduction event—ranging from cortex trimming (SI Appendix, section 3.5 and Fig. S3 E–G) to more refined preparation of flaking surfaces (SI Appendix, Fig. S8 D–F).

Manual dexterity involved in the detachment of targeted products is also significant. The CJW-prepared core refit sets contain remarkably different products in terms of shape and size, which evidence a deliberate and precise application of force in each of the reduction stages and particularly in the preparation stages versus the final targeted products. This level of manual precision is also reflected in some predetermined flakes, where percussion points are located with precision on the prepared ridges of the flake striking platforms (SI Appendix, Fig. S9).

As a result, CJW flaking techniques follow highly standardized operational schemes, where the role of flaking surfaces and striking platforms are carefully designed and evidence a conceptualization of core configuration in terms of hierarchy, resulting in predetermined products that are regular in shape and present identical technological characteristics.

Retouched Tool Patterns.

At stage 5 of the operational scheme reconstructed from the refit sets (Fig. 2A), targeted products are typically slender flakes with parallel or sub-parallel ventral and dorsal surfaces. Technological analysis of 22 refitted products detached at stage 5 indicates intentional breakage of flakes in two halves, with one or more of the resulting fragments selected as blanks for retouching (Fig. 3 A–D and SI Appendix, Fig. S7). Retouching is usually conducted on the adjacent edge of fracture planes in order to create tips with two convergent sides. Both the position of the fractures (orthogonal to the ventral or dorsal surfaces of flakes) and their conchoidal character (SI Appendix, Fig. S10) are incompatible with breakage produced during core reduction and preclude the possibility of post-depositional causes [the latter also being supported by the study of site formation processes at CJW (39)]. In short, the recurrence of this pattern indicates a systematic and specific template for the production of blanks selected for tool shaping.

Fig. 3.

Retouched tool patterns in the CJW assemblage. (A–E) Slender flakes are intentionally broken and used as blanks for retouching tipped tools. (A–D) are flakes detached at stage 5. C was first retouched, then deliberately fractured through an orthogonal motion, to which subsequent retouching of the fractured edge followed. (F) Bifacially retouched point (see SI Appendix, Fig. S12 for details of the retouching process) (G–I) Unifacially retouched points (3D models of (H and I) in SI Appendix. (J–M) Borers.

Alongside artifacts made on intentionally-broken blanks, other retouched tools show two additional shaping patterns: 1) retouching on two adjacent edges to produce points (Fig. 3 F–I), which usually involves more than ten removals and is mostly unifacial (although bifacial retouching occurs in a few cases), resulting in tools with symmetric shapes; 2) retouching on two adjacent edges to produce borers (Fig. 3 J–M).

These three retouching patterns observed in the CJW assemblage result in end products with comparable shapes and therefore evidence tool shaping standardization. Added to the predetermined core reduction reported above, this technological behavior represents a unique feature in East Asian early Pleistocene archaeological sites and supports the existence of sophisticated mental templates that are imposed throughout the stone tool reduction sequence.

Discussion

Mode 1 or Mode 2 Technologies?

The ability to produce large flakes and shape LCTs are often considered as the two key elements that set the Acheulean or mode 2 apart from the Oldowan or mode 1 (1, 4, 5, 8, 47) and involve what Isaac defined as an “extra step in stone tool manufacturing” that required the existence of mental templates (48, 49). Additionally, recent studies have highlighted the importance of a series of distinct knapping skills in Acheulean technologies that rarely occurred in the Oldowan, such as a hierarchical exploitation of small debitage cores, sophisticated operational sequences, and standardized end products (2, 7, 8, 17, 49). As a result of such technical abilities, the emergence of prepared core technologies is widely acknowledged to have occurred during Acheulean times (3, 19, 22, 50), differentiating this technology and their toolmakers from the earlier Oldowan or mode 1, evidenced by more refined knapping techniques and in-depth planning (19, 22).

The presence of a well-structured operational scheme and abundant predetermined products within the CJW assemblage strongly suggests the existence of prepared core strategies that were employed systematically. This implies that the CJW toolmakers possessed advanced technical abilities, including long-term planning capabilities, a deep understanding of flaking mechanisms, and high level of manual precision. Predetermined flakes are regular and slender, indicating standardization of the end products. Retouching tool patterns significantly alter the form of the original blanks to create standardized shapes, supporting complex mental templates among the CJW toolmakers. All of this is accompanied by an efficient exploitation of cores until exhaustion, fragmented reduction sequences, and organized management of raw materials (SI Appendix, section 4). This patterned technological behavior hints at the significant know-how transmission abilities of the CJW toolmakers. The logical continuity required to complete the successive actions involved in this prepared core technology may also be suggestive of the cumulative nature of such technological innovations.

Like the rest of the Nihewan early Pleistocene assemblages, CJW has customarily been attributed to mode 1 due to the lack of handaxes that are typical at contemporary sites in Africa, Europe, and West Asia. However, the outstanding information provided by CJW refit sets strongly suggests that, despite the absence of LCTs in the assemblage, CJW hominins possessed the technical abilities of mode 2/Acheulean hominins (SI Appendix, section 5.1). We predict that should the data provided by refit sets in CJW be available for other Nihewan assemblages (SI Appendix, section 5.3), technological similarities between the Asian sites and the Acheulean from Western Old World contexts would be reinforced further.

On the other hand, the differences between CJW (and by extension other Nihewan sites) and the typical Acheulean should not be ignored, and the absence of handaxes in East Asian sites cannot be attributed to raw material availability only (SI Appendix, section 5.2). Having established through the CJW refit sets that the assemblage does not comply with the technical simplicity attributed to mode 1 and resembles closely the technical and motor skills typical of mode 2, the question that arises is why hominins with an advanced technology equivalent to the Acheulean produced small tools to survive in the Nihewan basin.

One possible answer lies in the concept of “convergent thinking” (51). This process entails the creation of novel end products by combining available materials or knowledge (52). The high-latitude environment of the Nihewan basin, coupled with poor-quality local raw materials (SI Appendix, section 2.1), presents a vastly different context from that of Africa, Europe, and West Asia. Consequently, we propose that these mode 2 hominins with complex technical abilities were led to produce smaller tools due to the poor local raw material quality and environmental conditions (SI Appendix, section 5.2) (14, 53), which may have also required the development of new survival strategies and/or technological innovations to adapt to the local environmental conditions in this region of northern China (53, 54).

Implications for Hominin Dispersals in East Asia.

Given the evidence of early humans in Eurasia prior to the earliest Acheulean in eastern Africa (35, 55–58), it is widely accepted that the hominins that first dispersed from Africa into East and Southeast Asia were associated with mode 1 technology (59, 60). While the Acheulean outside Africa is first reported in Israel and India around 1.5 Ma (11–13), the earliest assemblages in East Asia associated with mode 2 technologies are dated to 0.8 Ma (61), while denser records of LCTs may exist after 0.35 Ma (62). In addition to its relevance for our understanding of how the Acheulean emerged in East Asia (SI Appendix, section 5.4), the CJW assemblage indicates that hominins who possessed advanced knapping abilities equivalent to mode 2 technological features occupied East Asia as early as 1.1 Ma, which is 0.3 Ma earlier than the first handaxes found in East Asia (61), and also precede the earliest prepared core technology record in Eurasia (32) by 0.3 Ma.

Conclusions

The evidence of a prepared core technology, as well as the standardized patterns observed in tool retouching, strongly suggest the existence of sophisticated technical abilities and mental templates among the CJW toolmakers in the Nihewan basin. Such technological abilities are consistent with those widely accepted for mode 2 hominins, thus forcing reconsideration of previously held views of the Nihewan basin assemblages as evidence of the technological stasis of mode 1 East Asian toolmakers throughout the early Pleistocene, which in turn was seen as supportive of the biological isolation of early human populations in the region. In light of these findings, we propose that a re-assessment is due of the early and middle Pleistocene assemblages in East Asia based on technological features rather than on the presence/absence of handaxes, in order to reconsider early human cultural and biological relations between East Asia and the rest of the Old World.

Furthermore, the CJW assemblage at 1.1 Mya now constitutes one of the world’s earliest occurrences of prepared core technologies, 0.3 Ma before the current earliest record in Eurasia (32, 33). This may suggest that mode 2 hominins dispersed into East Asia much earlier than previously thought and, given that the earliest occupation of China could be over 2 Ma (55), our results may also point at multiple migrations of early hominins into East Asia throughout the early Pleistocene.

Materials and Methods

The reduction sequence of refit sets was reconstructed through a technological reading of both refitted artifacts and their attributes during the flaking process. The lithic analysis followed standards described in SI Appendix, section 2. All archaeological remains referred to in this study are located in the Hebei Provincial Institute of Cultural Relics and Archaeology, China.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.