Relative chronology
As said earlier, for a long time and because of the whole Central Asia chronological sequence built up by our colleagues from the Neolithic down to the Greek period and the fact that the BMAC was considered as issued from the NMG V period of the Kopet Dagh area, our Russian and Uzbek colleagues dated it exclusively within the 2nd millennium, with an end around 1000 BC. The presence of a few iron artifacts -apparently from non-meteoric origin (cf. Sarianidi 1998a: 77) - confirmed these late dates.
Most of the small sites seem to be of short duration, no more than one period, and the scatter of sherds on the surface of the desert may indicate constant settlement shifts
(Hiebert 1994a: 28-29; Cattani 2008). For the largest settlements, the first architectural plan was maintained for a certain length of time with several phases of repairs and some modifications. At Gonur North, there is evidence of at least three main periods of occupation (Sarianidi 1998a: 83-85; Sarianidi and Dubova 2014a),24 followed by one where the site is used as a graveyard. The first is related to the construction of the kremlin and the buildings associated to it on its four sides, including the “Royal Necropolis”; it ended in a huge fire dated to the 20th century BC in median radiocarbon dates (see the Appendix). A second phase shows the construction of the outside wall over the hypogea of the “Royal Necropolis” and a reuse of some of the rooms for a metallurgical workshop. A third one shows the decline and reuse of the palace for ordinary houses as well as the construction of many water channels that destroyed part of the site (Sarianidi and Dubova 2012: 33-39; 2014b: 93, fig. 1). Finally, after the abandonment of this part of the site, a fourth phase shows that intrusive graves were dug within the walls. It is not yet clear, however, where Gonur South fits in this scheme.
On the same site, buildings are often scattered, and not rebuilt over several times, so that the stratigraphy is more horizontal than vertical, and their exact chronological correlation is unclear. At Gonur, for instance, the South mound dates to a later period than the North one and has been subdivided into five phases, the last one being related to the Takhirbaj 3 period or end of the Late Bronze Age (Hiebert 1994a: 113-114). Nonetheless, soundings show that below the wall of the fortress with round towers there are earlier levels almost untouched.
Many studies have attempted to establish the precise internal chronology at the major sites. A great deal of material has been published but the publications do not give all the necessary details to understand the stratigraphy. Further works should shed light on these questions. Specific chrono-typological studies have been attempted on the ceramics but they mostly concern that from graves (P’jankova 1993; Udeumuradov 1993, 2002; Luneau 2014; Teufer 2015).
All these problems have led to the constitution of separate chronological charts for Margiana and northern Bactria. In both areas, successive phases have been distinguished (in Margiana, the Kelleli, Gonur, Togolok, and Takhirbaj phases; in northern Bactria, those of Sapalli, Dzharkutan, Kuzali, Molali, and Bustan). To make the question more complex, in the last region, the succession has been contested with the Kuzali phase placed at the end, and other denominations are used (Late Bronze Age I and II) (see Kaniuth and Teufer 2001; contra Vinogradova in Vinogradova et al. 2008: 220-228) (see Chapters 16 and 13).
In general, the BMAC is divided into three main large periods, with conventional dates. The oldest period (pre-2000 BC) is mainly identified in Margiana and probably also at Sapalli and Dashly 3. The middle period (after 2000 BC) covers all the areas of the BMAC until ca. 1800-1700 BC. The last period shows signs of disaffection, abandonment, and isolation, and runs until ca. 1500 BC (see Luneau 2014; see also Chapter 18). Steppe intrusions with campsites around the BMAC settlements may have started already within the early period (see Chapter 17) but they mainly grow in importance during the middle and last periods (Cattani 2008; Salvatori and Tosi 2008: xi; Rouse and Cerasetti 2017).
This chronological and geographical distribution, however, is still debated and recent finds in southern Tajikistan show that this area was also related to the BMAC from the very beginning (see Chapters 23 and 25).
Absolute chronology
Up until 20 years ago, *4C dates were still rare, isolated, and rarely calibrated. As already mentioned, our Russian colleagues considered that the BMAC extended mainly within the 2nd millennium up to around 1000 BC. Salva tori has often dated its beginning very early (ca. 2400 BC), to make it match with Shahdad where a large amount of material similar to that of the BMAC has been discovered. With the start of international cooperation and the multiplication of analyses, the dates now admitted by all place the Oxus Civilization between 2250 and 1700 BC, while its final phase extends until ca. 1500 BC.
The constant improvement in the calibration of the dates have led us to collect once more15 most of the known dates of the major sites linked with the BMAC and to recalibrate them all according to the last standards (see the Appendix).
On the basis of the 129 radiocarbon dates made at Gonur itself, and from their median date, we can see that five of them point at an early date (first third of the 3rd millennium), seven are situated around 2500-2400 BC, five in the 2300s, 12 between 1700-1500 BC, one is an anomaly, and all the rest (99) are situated between the 2200-1700s BC. Only further excavations will eventually bring to light architecture and material related to the earlier 3rd millennium, which is for the moment lacking.
In a very broad scheme, using the available information on the provenience of the MC samples, these radiocarbon dates point out that Gonur North (including the Large Necropolis), or period I, existed mainly between the 23rd and the 19th centuries BC. It also shows that it was in part contemporaneous with Sapallitepa, the Citadel and the pre-temple levels of Dzharkutan, Tillja Bulak, Tugai, Gelot and Darnajchi, Shahr-i Sokhta IV, Hissar IIIC, and Tal-i Malyan (Kaftari period). It is contemporaneous with the end of the Indus urban phase (IIIC) and the beginning of the Late Harappan phase (IV).16 The second BMAC major period, with round towers fortresses is placed during the 18th century BC at Gonur South and the Dzharkutan Temple (VI). Consequently, a major change seems to have happened in the second half or end of the 19th century BC (see Tables 1.1,1.2, and 1.3).
Thus the changes between periods I and II are contemporaneous to those observed in the “international” sphere (decline of the Harappan Civilization, endless wars within Mesopotamia, change in the Gulf traffic from Magan to Dilmun) and we cannot rule out that they may have had an impact on the BMAC (Lyonnet 2019).
