Acheulean sites or at least “Acheulean” occurrences are numerous in South Asia (Gaillard, 2006). They are characterised by handaxes and cleavers comparable to those found in Africa. Although they have been known for a long time, the crucial question of their chronology remains unresolved in most cases; due to generally thin stratigraphic sequences and poor faunal preservation. However, improvements in fieldwork and progress in dating methods over the past two decades point towards a Lower Pleistocene timeframe for the onset of the South Asian Acheulean.

Interestingly, one of the main sites described one and a half centuries ago (Foote, 1868) has recently become the reference point for the early out of Africa Acheulean. The Paleolithic localities along the Attirampakkam Stream, a small tributary of the Kortallayar River, have been under reinvestigation since 1999 (Pappu, 2001) and several trenches and horizontal excavations have been opened. The deepest trenches yielded quartzite artefacts and measurements of the ²⁶Al/¹⁰Be ratio of these tools assess their burial age. The results provide an average age of 1.51 ± 0.07 Ma, corroborated by reverse paleomagnetism (Pappu et al., 2011).

At Attirampakkam, the Acheulean industry is made from local fine to coarse quartzite. Apart from the large cores from which large flakes were detached, which are missing from the site, all the stages of lithic production are present. Trench T8, for example, provided 3528 artefacts from Acheulean levels (layers 6 to 8), of which 95% are ordinary debitage products (< 10 cm), alongside a few flakes resulting from handaxe shaping. The remaining 5% are mostly large flakes (> 10 cm), half of which were shaped into handaxes (1.4% of the assemblage), or cleavers (0.6%) with a few trihedrals (n = 6), while the other half was minimally retouched or used (Pappu et al., 2011).

The rich complex of sites in the Hunsgi and Baichbal valleys allows researchers to trace technical evolution from the Early to Late Acheulean in this sector of north-western Karnataka (Paddayya, 2008) and to present hypotheses regarding the seasonal movements of Lower Paleolithic populations (Paddayya, 2014). Quartzite is not available in these valleys, except in the form of pebbles from the overlooking Trappean plateau. Silicified limestone is thus used instead, especially at Isampur in the Hunsgi Valley. A few faunal remains have been preserved and the ESR dating of three enamel samples from bovine teeth resulted in a mean age of 1.2 Ma.

Trench 1 yielded 13,043 artefacts, of which 92% are “incidental” products (flakes and debris including products >5 mm from sieving). The industry results from the exploitation of silicified limestone slabs, up to 1 m long, directly quarried at the site itself. Knappers started the reduction sequence on protruding slab corners and these opening removals on the profile then served as striking platforms for flake production, using unifacial or bifacial methods. The resulting flakes (excluding the “incidental products”) are mostly side-struck (technically short). Altogether, they average 18 cm long. The shaped tools in this assemblage comprise 48 handaxes, 15 cleavers, 18 knives, 3 discoids, 14 chopping-tools and 65 scrapers. Half of the handaxes (26/48) are made from slabs or cobbles, while the other half (22/45) are made on flakes. Cleavers and knives are on flakes, except two cleavers on slabs. These larger tools are mostly in silicified limestone while the scrapers, usually on flakes (49/65), are in chert or quartzite (Paddayya et al., 2002 and Paddayya et al., 2006).

Isampur is one of the “core activity spots” in the Hunsgi Valley beside many small sites and “non-sites” (occurrence of one or several artefacts). Acheulean-making human groups settled directly on raw material sources. Cleavers are made on flakes, whereas handaxes are often made on other opportunistic blanks.

Morgaon (Pune district, Maharashtra) is one of the rare Acheulean sites in the Deccan Trap region, as basalt is generally subject to weathering and the slightest reworking completely destroys basalt artefacts (Mishra, 1982). They can thus only be preserved in situ (Deo et al., 2007 and Mishra et al., 2009). The stratigraphic sequence is of alluvial origin suggesting that Acheulean-making populations settled near a stream. The main Acheulean level is at the surface of black fissured clay, about 2 m thick. A few more fresh and abraded artefacts occurred in sandy pebbly gravel above the clay, including two well-preserved cleavers. Below the clay lies gravel with cobbles and boulders where some laterite pebbles indicate former lateritic formations, now eroded away from the landscape. Such formations are widespread in the western part of the Deccan Traps and often yield Acheulean artefacts (Rajaguru et al., 2004). They usually lie on weathered basaltic bedrock, as at Morgaon. Lenses of tephra occur below and within the black-fissured clay. Paleomagnetic studies show that the black clay and the overlying levels have registered at least two magnetic imprints considered to be Matuyama and Bruhnes. The tephra is earlier than the Acheulean occupation, but not much earlier given its position in the stratigraphy. Dating attempts conclude that it is about 800 ka and corresponds to the Old Toba Tephra (Haslam and Petraglia, 2010 and Westaway et al., 2011).

Besides the surface collection, 180 artefacts were unearthed from a 4 × 8 m trench at the top of the black clay (Fig. 2). They were made from local basalt available in the form of “core stones” resulting from the weathering of originally prismatic basalt blocks. The assemblage is mainly characterised by large flakes, either directly struck from “core stones” (double patina), or from very large flakes, implying Kombewa reduction. A few of these large flakes were roughly shaped into cleavers. Some rare handaxes were found on the surface only.

Acheulean artefacts were collected from several places around Didwana (Nagaur district, Rajasthan) (Mishra and Rajaguru, 1986 and Mishra et al., 1982). The first discoveries occurred in calcrete quarries, especially at Singi-Talav (Didwana) and in the neighbouring village of Amarpura (3 km apart). These calcretes result from pedogenic processes and calcium carbonate precipitation within originally silty clayey sediments with some lenses of sand (Dhir, 1995), suggesting a landscape with abundant bodies of water (inherited from a drainage system disrupted by tectonic tilting in the Late Tertiary). The climate was already semi-arid with alternating dryer and slightly wetter phases. In Amarpura, the calcrete formation reaches a thickness of 12 m. One meter below the top of this formation, a sample was dated to 800 ka (Kailath et al., 2000). The Singi Talav sequence has been correlated with the middle of the Amarpura sequence on the basis of topograpy and lithology.

Acheulean populations settled near a possibly seasonal lake. The two main layers (3 and 4) have yielded Acheulean industry made from metamorphic rocks, mainly quartzite, brought from 3 km away (Fig. 2). However, a few cobbles were also used, which appear to come from 20 km away, based on the closest known source today (Gaillard et al., 2010, Gaillard et al., 1983 and Gaillard et al., 1986). The assemblages from the two Acheulean layers do not differ much, except for a decrease in the proportion of handaxes in the upper layer (3% in layer 4 as opposed to 1% in layer 3, among 891 and 401 artefacts respectively), and a concomitant increase in cores. Flakes and debris, in equal proportions, are the main component of this industry (80%). The flakes were usually struck from rather homogeneous quartzite while the large cutting tools (mostly handaxes) were struck or split from schistose quartzite, yielding slabs rather than flakes. For the spheroids, polyhedrons and hammerstones, medium to coarse-grained quartzite cobbles were preferentially used. The handaxes and rare cleavers are sometimes very roughly worked, as the original rectangular or rhombic shape of the slabs provides readymade thin, symmetrical and often pointed tools (Gaillard, 1993, Gaillard et al., 1983 and Gaillard et al., 1986).

Hominins probably established their camp at Singi Talav in the late Lower Pleistocene due to nearby water availability. Lithic raw materials were totally absent in the immediate surroundings and were collected at least 3 km from the site. There is no production or importation of large flakes, apart from two large flakes or splits, transformed into cleavers. The handaxes are made with minimal technical investment from slabs of schistous quartzite, selected for this purpose only, while ordinary debitage is on isotropic quartzite.

The Late Acheulean is known, for instance, in the Hunsgi and Baichbal valleys (Karnataka), the Renigunta region (Andhra Pradesh), Bhimbetka Shelter III-F-23, Middle Son Valley (Madhya Pradesh) and many other sites in peninsular India. In the Middle Son Valley, geological and archaeological studies began in the 1970s (Sharma and Clark, 1982, Sharma and Clark, 1983, Williams and Royce, 1983 and Williams et al., 2006) and are still in progress today (Haslam et al., 2012 and Pal et al., 2005). Four geological formations have been identified, ranging from the Middle Pleistocene to the Late Holocene. Several localities have been excavated, each yielding several tens of artefacts (Kenoyer and Pal, 1983, Mishra et al., 1983). Recent OSL dating points to ages from the end of MIS 6 and the beginning of MIS 5 (Haslam et al., 2011). Artefacts seem to be confined to the colluvium deposits accumulated at the end of MIS 6, in a dryer climate than during MIS 5. These assemblages are made from local rocks, especially quartzite, collected a few kilometres away from the sites. They are composed of a few cleavers and refined handaxes alongside abundant flakes and cores of various types, including discoid cores. Many of the flakes are retouched, mostly into scrapers. These assemblages are considered as Latest Acheulean or Middle Palaeolithic, depending on the authors.

During the late 1960s, P. Sorensen claimed that early humans were present in northern Thailand. In the Lanna area, he highlighted an “archaic” assemblage with pebbles and scrapers made of silicified wood (Sorensen, 1981 and Sorensen, 2001). This Thai Paleolithic culture was called the “Lannatian” and was controversially attributed to between 900–600 ka.

Fewer than ten pebble tools, choppers and quartzite flakes were discovered by G. Pope at the Lampang open-air site (Ban Don Mun and Ban Mae Thae, Pope et al., 1981, Pope et al., 1986 and Pope et al., 1987), above a basaltic level estimated at 700 ka by paleomagnetism (Keates, 2002). The Ban Don Mun site (Lampang) was recently re-analysed and surveyed, and assigned a minimum age of 550 ka by K/Ar dating of the basaltic level (Forestier et al., 2008 and Zeitoun et al., 2013). Pebble tools were located above this level. Recent discoveries of 25 new artefacts were added to the first series. They include numerous choppers with convex, convergent or transverse cutting edges, few chopping-tools, rare small flakes (<10 cm length), broken pebbles and only one trihedral pick.

Sao Din is another open-air site in Northeast Thailand (Na Noi Basin, see also Fig. 8.2). It presents an upper and lower alluvial terrace system comprising numerous stone artefacts. Zeitoun et al., 2008, Zeitoun et al., 2012a and Zeitoun et al., 2012b refer to more than 300 pieces in the lower part of the eroded section, 20 of which are stratigraphically located below a layer dated at 300 ka (minimum age estimated from paleomagnetism, ESR and OSL, Chenglong, pers. comm). This pebble tool assemblage shows the same technical pattern as the one observed on the Ban Don Mun artefacts, with the production of large choppers (20 cm long) with transverse, convergent or lateral cutting edges on sandstone and quartzite pebbles, rarely on quartz. These stone tools are more massive than the Lampang tools and are associated with chopping-tools, large flake-tools (scrapers, denticulates and notches), massive and heavy prismatic cores and some atypical tools with a bilateral pointed shape. The majority of the Sao Din artefacts are unifacial, although several are bifacial.

These massive unifacial pebble tools in the North of Thailand have no equivalent in other mainland Southeast Asian countries, despite the recent discoveries of stone tools on the Quaternary terraces of the Mekong River (Forestier et al., 2014).

Generally, Paleolithic stone tools from mainland Southeast Asia are dominated by choppers, i.e. unifaces, and include few handaxes (Gorman, 1972, Heekeren et al., 1967, Heider, 1958, Higham, 2014, Marwick, 2007, Matthews, 1964, Moser, 2001, Mus, 1977 and Schoocongdej, 2006).

The lithic assemblages from the Indonesian archipelago point to comparable techno-typological behaviour to the continental series, including the presence of pebble tools (Fig. 8). They also include a classic but ‘exotic and tropical’ Acheulean toolkit.

Java Island, famous for its paleoanthropological record since the discovery of ‘Pithecanthropus’ remains at the end of the 19^th century (Dubois, 1994, Koenigswald, 1936 and Koenigswald, 1940), lacked stratigraphically controlled archaeological data for Homo erectus, until recent decades, apart from isolated discoveries (see for instance Koenigswald et al., 1973, Sémah et al., 2014a, Sémah et al., 2014b and Widianto, 2006).

The “Patjitanian” industry was discovered in the southern mountains of Java by Koenigswald (1936), and defined by the presence of numerous choppers, chopping-tools, small and large flakes, and cores (including some giant cores, with volumes of more than 100 dm³). It has since become clear that handaxes existed on pebbles or flakes associated with a broad spectrum of artefacts in chert and silicified wood (from eroded Miocene series), volcanic stone (andesite) and silicified breccia.

Although the age of the Patjitanian industry is still unknown, several markers tend to attribute it to the Middle Pleistocene, in relation with the oldest flake-and-core assemblages discovered in the surrounding caves of the Gunung Sewu karstic hills (e.g. Song Terus cave, Sémah et al., 2004, dated to the second half of the Middle Pleistocene).

For a long time, the famous Sangiran dome hominid-bearing site yielded small flakes (called “Sangiran Flakes”, see von Koenisgwald and Gosh, op. cit.) made from small siliceous pebbles collected from alluvial formations. The first discoveries of “heavy-duty tools” are related to the discovery of bolas and a chopper in metamorphic rock at Ngebung (Soejono, 1982). In the north-western part of the dome, excavations conducted at the Ngebung open-air site (Sémah et al., 1992), dated at around 800 ka (Sémah and Sémah, 2012 and Sémah et al., 2011), yielded many andesitic pebbles or manuports, but also polyhedrons, spheroids and bolas with pecked tracks (shaping traces and usewear). Some artefacts were made with raw materials collected tens of km away: large fine-grained suitable andesite pebbles (usually lacking in the local lahars) or milky quartz. The toolkit includes cleavers, massive scrapers (“horsehoof” type), choppers and a few flakes. Surveys and excavations at the Sangiran dome confirmed the behavioural diversity of Homo erectus groups at the dawn of the Middle Pleistocene, most probably reflecting repetitive exchanges with the mainland owing to the drastic sea level drops that occurred during this period (Sémah and Sémah, 2012, Sémah et al., 2014a and Sémah et al., 2014b).

Other discoveries in the Sunda Islands are less detailed. It is important to mention the artefacts found in South Sumatra in the Air Tawar and Semuhon Rivers, not far from the Ogan River in the karstic region of Baturaja (Forestier, 2007 and Forestier et al., 2005). All the material from this series is massive, made in local chert and composed of handaxes with triangular and trapezoid sections (>30 cm long), and cleavers. It also includes choppers on limestone or andesitic pebbles, prismatic cores and large flake-tools (scrapers and denticulates). The quality and morphology of the raw material varies, with blocks of chert, glossy sandstone, breccia or silicified wood (Forestier, 2000, Forestier, 2007 and Simanjuntak and Forestier, 2009). Therefore, in addition to Java, Sumatra provides evidence of Lower Paleolithic occupations showing the toolkit diversity of these earliest islanders.

The evidence beyond the Wallace line is restricted to a couple of sites, the oldest one probably being the c. 1 Ma-old artefacts discovered in the Soa Basin on Flores Island (Brumm et al., 2010). We may also mention the tools collected on the terraces of the Walanae River, at Cabenge (South Sulawesi), including partial or complete triangular handaxes, unifacial tools with abrupt retouch, pebble tools (choppers and chopping-tools), and flakes and cores (Bartstra, 1976, Bartstra, 1977, Keates, 2004 and Keates and Bartstra, 2001). So far, the bifacial artefacts discovered at the site of Arubo on the island of Luzon raise the question of early human settlement in the Philippines (Dizon and Pawlik, 2010).