The book by C. J. Bae (a.k.a. C. J. Norton) is one of the rare kind, because only a few volumes devoted to the topic of eastern Asian paleoanthropology have been published before (e.g., Norton and Braun Reference Norton and Braun2010; Wu and Pourier Reference Wu and Poirier1995). It is positioned as a textbook (p. xiv). The author’s goal is “… to synthesize and evaluate the current state of the eastern Asian (including both East and Southeast Asia) paleoanthropological record and place these findings in the context of broader theoretical debates.” (p. xiv; see also pp. 211–212). The book consists of abbreviations, acknowledgments, a prologue, 7 chapters, references, and an index. The list of sources is very impressive, with ca. 870 papers and books cited.

Upon reading this volume, my impression is twofold. On the one hand, this is a useful compendium where current information about the most important hominin fossils and Paleolithic sites in eastern Asia is assembled. The author highlights the fact that paleoanthropology in this region is a very dynamic field (pp. 208–212) and that the situation is constantly changing and being updated. On the other hand, the presentation and evaluation of factual material in this book reflect shortcomings of how physical anthropology and archaeology in eastern Asia, primarily China, are practiced. A discussion of the most controversial cases, with particular focus on the chronology, is presented below.

The major problem in dating the Chinese (and to some extent Southeast Asian) hominins is that the exact stratigraphic position of human fossils in many cases is unknown; they were often found by local farmers who were digging cave deposits for guano (fertilizer), in the process of which they frequently found ancient animal bones. Scientists arrived later and they were shown the approximate location of the hominin find; therefore, the association between dated specimens and in situ position of the fossils is unclear. As in other parts of the world, the deficit of direct age determinations for hominin fossils is clear for eastern Asia; this problem is acknowledged by the author (pp. 101, 145, 182) and other scholars (e.g., Etler Reference Etler1996, 293; Keates Reference Keates2010). Here are some examples illustrating the problems of indirect dating and assumptions about the primary association between dated samples and human remains.

The chronology of the Salawusu (a.k.a. Sara-osso-gol and Xarusgol) hominins in the Ordos region of northern China (pp. 138, 168–171) is controversial, although the author accepts its age as ca. 90,000–100,000 years (pp. 138, 186). Most of these fossils, belonging to H. sapiens, were found on the surface (e.g., Wu and Poirier Reference Wu and Poirier1995, 170). The single direct AMS ¹⁴C dating of the supposedly Late Pleistocene PA 62 femur generated an age of ca. 240 years ago (see also p. 169) and the δ ¹³C signal (–13.8‰) of a millet consumer (Keates et al. Reference Keates, Hodgins, Kuzmin and Orlova2007); a similar result was obtained by other scholars (Shang et al. Reference Shang, Wei and Wu2006).

The viewpoint of some anthropologists that modern humans (i.e., Homo sapiens) have been present in China since at least 100,000 years ago (e.g., Liu et al. Reference Liu, Martinón-Torres, Cai, Xing, Tong, Pei, Sier, Wu, Edwards, Cheng, Li, Yang, Bermúdez de Castro and Wu2015), also accepted by the author (pp. 133, 137, 145–151, 159, 195), was severely compromised by the direct AMS ¹⁴C dating and DNA analysis of several teeth from the Fuyan, Yangjiapo, and Sanyou cave sites (Sun et al. Reference Sun, Wen, Lu, Zhou, Curnoe, Lu, Li, Wang, Cheng, Yi, Jia, Du, Xu, Lu, Lu, Zheng, Zhang, Sun, Wei, Han, Huang, Edwards, Jin and Li2021). It turned out that the teeth of supposedly very early H. sapiens date to only ca. 1700–9400 years agoFootnote ¹ . A heated discussion followed (see Martinón-Torres et al. Reference Martinón-Torres, Cai, Tong, Pei, Xing, Bermúdez de Castro, Wu and Liu2021). The conclusion about the much younger age of the H. sapiens fossils in Fuyan Cave was also made by Michel et al. (Reference Michel, Valladas, Shen, Wang, Zhao, Shen, Valensi and Bae2016) and Kuzmin (Reference Kuzmin2020, 133–134). Fossils of supposedly early H. sapiens from the Huanglong and Luna caves were ¹⁴C-dated by associated material (animal teeth and charcoal) to ca. 33,000–35,000 years ago and ca. 4700–15,200 years ago, respectively (Sun et al. Reference Sun, Wen, Lu, Zhou, Curnoe, Lu, Li, Wang, Cheng, Yi, Jia, Du, Xu, Lu, Lu, Zheng, Zhang, Sun, Wei, Han, Huang, Edwards, Jin and Li2021). The confidence expressed by the author about the undisturbed stratigraphy and secure chronology of the Luna Cave (p. 150) now seems to be unsubstantiated. To conclude the discussion about the chronology of early H. sapiens in China, it is plausible to accept that the oldest appearance of these populations can now only be dated to ca. 40,000 years ago (e.g., Sun et al. Reference Sun, Wen, Lu, Zhou, Curnoe, Lu, Li, Wang, Cheng, Yi, Jia, Du, Xu, Lu, Lu, Zheng, Zhang, Sun, Wei, Han, Huang, Edwards, Jin and Li2021).

As far as Korean fossils of early H. sapiens are concerned, the author (pp. 151–152) considers the Ryonggok Cave in North Korea as the oldest find, dated to ca. 44,000–72,000 years ago. This, in my opinion, is controversial, because the animal fossils and pollen indicate a relatively warm climate (Choe et al. 2021). According to the Late Pleistocene paleoclimatic data (e.g., Lisiecki and Raymo Reference Lisiecki and Raymo2005), the existence of hominins in the Ryonggok Cave should be associated with the interstadial Marine Isotope Stage (MIS) 3, dated to ca. 30,000–57,000 years ago, and not with MIS 4 (ca. 57,000–71,000 years ago) with its cold, glacial climate. In the light of uncertainties with the chronology of Ryonggok Cave (p. 151), it is plausible to suggest that the earliest H. sapiens in the Korean Peninsula come from the Gunang Cave in central South Korea, dated to ca. 43,000 years ago (Park et al. Reference Park, Kim, Lee and Woo2019). This is not discussed by the author.

In the Japanese Islands, the Early Upper Paleolithic (EUP) complexes associated with modern humans emerged at ca. 38,000 years ago (e.g., Nakazawa Reference Nakazawa2017). The author states that for the beginning of the EUP, a few sites are known (p. 173). This contradicts other sources where several dozens of the oldest EUP sites are mentioned (e.g., Izuho and Kaifu Reference Izuho, Kaifu, Kaifu, Izuho, Goebel, Sato and Ono2015).

In Table 6.1 (p. 185), the age of H. sapiens from the Laishui site near Beijing is given as “28 ka?”. The author (p. 189) refers to a paper by Etler (Reference Etler1996, 292) where no information about either the dating method or material was provided. However, in the Sackler Museum of Archaeology (Beijing University), the label of the display clearly states: “The fossils were dated by electron-spin analysis to 28,000 years ago …” (personal observation, 2008). This date was never properly published, but one can assume that bones of animals found along with the hominin remains were analyzed.

The assumption about the primary association between speleothems and hominin fossils in several caves in China, such as Chaoxian (pp. 100–102), Hualongdong (pp. 106–107), Liujiang (pp. 144–145), Huanglongdong (pp. 145–147), Zhirendong (p. 147), Fuyandong (pp. 148–149), and Lunadong (pp. 150–151), now appears to be problematic in the light of work by Sun et al. (Reference Sun, Wen, Lu, Zhou, Curnoe, Lu, Li, Wang, Cheng, Yi, Jia, Du, Xu, Lu, Lu, Zheng, Zhang, Sun, Wei, Han, Huang, Edwards, Jin and Li2021). In numerous cases, there are no flowstones that completely underlie or overlie the strata with hominin fossils, and it is not secure to assume the direct relationship between dated speleothems and hominins. The danger of accepting the straightforward association of hominin finds and geological strata was clearly demonstrated by Kaifu et al. (Reference Kaifu, Kurniawan, Yurnaldi, Setiawan, Setiyabudi, Insani, Takai, Nishioka, Takahashi, Aziz and Yoneda2022). Only direct dating of the human fossils can establish their true age (e.g., Grün and Stringer Reference Grün and Stringer2023).

Some important Chinese Paleolithic sites with hominin fossils, such as Maba and Xujiayao, assumed to be of Middle Pleistocene age, contain bones of Equus przewalskii. It is well-known that this species of horse existed in China only in the Late Pleistocene and Holocene (e.g., Qiu Reference Qiu2006), and not earlier than ca. 130,000 years ago. Therefore, these sites—especially Xujiayao, in the context of the Homo juluensis issue (see below)—appears to be younger than accepted by the author (pp. 102–104, 110).

Another hotly debated issue is the chronology of the earliest hominins and Paleolithic sites in eastern Asia. Some of fossils are surface finds (Yuanmou, p. 50); for some sites, the association between artifacts and loess/paleosol strata is dubious (Shangchen, p. 51); and some sites are found in a re-deposited context (Xihoudu, pp. 51–52; Hualongdong, pp. 106–107; Denisova, pp. 114–118). A few more words should be said about the chronology of the Shangchen site, with an assumed age of 2.12 million years ago (Zhu et al. Reference Zhu, Dennell, Huang, Wu, Qiu, Yang, Rao, Hou, Xie, Han and Ouyang2018). Similar to the case of the Nihewan Basin (see below), the finds of artifacts and sections with paleomagnetic data are from 100 m to 500 m apart. The key subsection KW is quite shallow (less than 10 m thick) (Zhu et al. Reference Zhu, Dennell, Huang, Wu, Qiu, Yang, Rao, Hou, Xie, Han and Ouyang2018, Extended Data Fig. 9), and its correlation with other, more complete sections, is dubious in the absence of undeniable markers like layer(s) of tephra (volcanic ash).

Chinese scholars, conducting paleomagnetic studies of several key sites with lithics and hominin fossils, especially in the Nihewan Basin of North China, correlated these sites with both the neighboring and distant geological sections without any cultural and/or anthropological materials. For example, the Xiaochangliang site and the Donggou section are ca. 1 km away from each other (Zhu et al. Reference Zhu, An, Potts and Hoffman2003, 344). The assumption that the strata are identical in both site and section is not secure like it would be in the case of tephra markers which are absent in the Nihewan Basin. In such a case, the age determination of the Xiaochangliang site appears unreliable. According to my personal observation of Xiaochangliang in 2008, the thickness of the deposit is not ca. 75 m as shown by Zhu et al. (Reference Zhu, An, Potts and Hoffman2003, 345 and their Fig. 2) but much less, around 6–7 m (see also Keates Reference Keates2000, 164; Yang et al. Reference Yang, Deng, Zhu and Petraglia2020, 130). It seems that what is presented by Zhu et al. (Reference Zhu, Hoffman, Potts, Deng, Pan, Guo, Shi, Guo, Yuan, Hou and Huang2001, Reference Zhu, An, Potts and Hoffman2003) is a “composite” section where different outcrops are combined without any guarantee that they are continuous. This practice has been common in paleomagnetic studies of the Nihewan Basin and North China since the 1970s (see Cheng et al. Reference Cheng, Lin, Li and Liang1978; Li & Wang, Reference Li, Wang and Liu1985). Instead of questionable correlation with another, more complete section, a horizontal trench should have been dug from the site on the slope toward the Cenjiawan Platform on the watershed where the sequence of Pleistocene strata begins with the Late Pleistocene Malan Loess, and then traced the deposits of the Xiaochangliang site directly to the Donggou reference section. Finally, the “estimated rate of sedimentation” sensu Zhu et al. (Reference Zhu, Hoffman, Potts, Deng, Pan, Guo, Shi, Guo, Yuan, Hou and Huang2001) is too rough for the exact determination of the age of Xiaochangliang. This is in sharp contrast to other studies involving paleomagnetic dating of the earliest human occupation of Europe (e.g., Gibert et al. Reference Gibert, Scott, Deino and Martin2024) where the reliable correlation of several key sections was demonstrated.

In several cases of the dating of early Chinese hominins, one can observe elements of wishful thinking. For example, at the Gongwangling locality the paleomagnetic data from a section situated 10 km away was used (Zhu et al. Reference Zhu, Dennell, Huang, Wu, Rao, Qiu, Xie, Liu, Fu, Han, Zhou, Ou Yang and Li2015). In the absence of undeniable stratigraphic markers like tephra with a known geochemical “signature” such a correlation is not trustworthy by any means; any good student of Quaternary geology knows this.

The degree of association between handaxes and tektites (dated by Ar–Ar and fission track methods to ca. 800,000 years ago) in the Bose Basin of southern China (pp. 82–87) deserves some brief comments. This issue was raised soon after the publication of the first results by Hou et al. (Reference Hou, Potts, Yuan, Guo, Deino, Wang, Clark, Xie and Huang2000) (see Koeberl et al. Reference Koeberl, Glass and Keates2000). Later on, Longbroek (Reference Longbroek2015) discussed it in more detail. It is not well-known to the international scholarly community that the handaxe-like artifacts described by Hou et al. (Reference Hou, Potts, Yuan, Guo, Deino, Wang, Clark, Xie and Huang2000) were not found in situ position, since all of them were collected on the surface (Gao Reference Gao and Yi2011, 202). In my opinion, the primary association between tektites and handaxes is still not clear.

As far as the taxonomy of the Late Pleistocene hominins is concerned, the author is a definite splitter (see also Bae and Wu Reference Bae and Wu2024). Recently, a new species Homo juluensis was proposed based on the Xujiayao and Xuchang fossils (pp. 121–123). Applying the multiregional model (a.k.a. “continuity”; e.g., Wu Reference Wu2004) to hominin evolution in eastern Asia, Bae and Wu (Reference Bae and Wu2024, 1) mentioned that “In China, this was always interpreted to mean that modern Chinese people today could trace their ancestry directly back to at least Zhoukoudian Locality 1 Homo erectus, if not to the earliest appearance of hominins in China during the Early Pleistocene (Schmalzer Reference Schmalzer2008).” Surprisingly, the Denisovans of southern Siberia are also included in H. juluensis (pp. 99, 120), despite the fact that crania of Denisovans have not been found (see Kuzmin et al. Reference Kuzmin, Slavinsky, Tsybankov and Keates2022, 326–327).

It appears to me that there is not enough data to define H. juluensis as a new species of Homo, mainly because of the fragmentary nature of the fossils and uncertainties of their chronology. For decades, hominins that are now determined as H. juluensis and H. longi (pp. 109–110, 120–123), were known as “archaic H. sapiens” (e.g., Wu and Poirier Reference Wu and Poirier1995, 114–157), and in my opinion should remain the same until more serious arguments for the definition of the new species can be found. For example, Dennell (Reference Dennell2020, 296) denied any direct association of the Xujiayao and Denisovan fossils.

In my view, the so-called “continuity” of hominin evolution in China sensu Wu (Reference Wu2004), from Homo erectus via archaic H. sapiens to true H. sapiens (pp. 96, 107, 132–134), is to a large degree driven by political motivation. In some sources, one can see statements like this: “Within the framework of the regional evolutionary model, Peking Man appears to have been one of the direct, albeit remote, ancestors of the [Chinese – Y.K.] nation.” (Liu and Chen Reference Liu and Chen2012, 15). The use of archaeology as a tool for the rise of Chinese nationalism—including the issue of the origins of modern humans—was described in detail in Schmalzer (Reference Schmalzer2008, 246–300).

Finally, by borrowing words from the review of two books on Chinese archaeology by Campbell (Reference Campbell2013, 1217), one can say that the volume by Bae (Reference Bae2024) “… should not be uncritically fed to unsuspecting undergraduates.” I sincerely hope that in coming years many of the problems outlined in this review will be at least partly solved.