Although this compilation has about 800 radiocarbon dates, of which some three-quarters could be calibrated in an acceptable time range, only general trends can be deduced of this compilation.
Two criteria have proved useful for interregional comparisons:
1)AMS-dates of small seeds allow very precise dating of sites. However, several sites (e.g. Abu Hureyra) show that seeds are very often displaced from upper or lower layers due to roots, animal activity, or building construction during the time of prehistoric occupation. Similar observations have been made by Danièle Martinoli for the Epipalaeolithic plant remains of Özküzini Cave in southwest Anatolia (Martionli 2004:69). It might therefore be useful to date small twigs, which are less easily displaced than seeds, but not as much affected by the old-wood effect as beams or posts.
2)For a more secure dating, shares of the same sample should be sent to different laboratories to check interlaboratory deviations.
Due to the large +/- deviations, many radiocarbon dates of the Epi-palaeoltihic have not been calibrated. The time ranges that have been suggested for the early Natufian (ca. 13400-11800 BC) and for the late Natufian cannot yet be stated more precisely. It is still very difficult to give a precise date for the end of the late Natufian on the basis of the available radiocarbon dates. In our view, it is only possible to put it in the range of 10.000 BC+/- 500 years (Benz 2000:39; Kuijt 2006; cf. Aurenche et al. 2001).
The duration of the late Natufian is uncertain for other reasons. The seed samples giving the late dates of the Epipalaeolithic layers of Abu Hureyra are clearly intrusive. Also the radiocarbon dates of the late Natufian and so-called early Khiamian of Mureybet overlap too widely and to be separated chronologically. More radiocarbon dates covering this transition are available from Tell Qaramel, but their interpretation has to await the final publication of the material discovered in the different layers of this interesting site (Mazurowski et al. 2009).
As was to be expected for methodological reasons, the more than 600 calibrated radiocarbon dates from Epipalaeolithic and early Neolithic sites show a strong decrease of dates during the Younger Dryas (Fig.1-2). Greater mobility (Munro 2004) and a decrease in the number of permanent settlements (Benz 2000:63-73) could be a reason for decreasing chances to detect sites during this climatic regression to ice age conditions. In the central southern Levant this hiatus shows up most clearly around 10400 BC to 9700 BC. Two exception are, however, the sites of Mallaha and of Hatula, which have dates falling into the Younger Dryas (Fig.2.2). But at least the dates of Hatoula should be tested again, as they are only dated by the laboratory of Gif-sur-Yvette and it could be shown that, in the case of Aswad, conventional Gif-dates were systematically about 500 years older than the other dates. The new dates from the upper layers of Mallaha (late [IB2] and final Natufian [1B1]) are also dated by Gif-sur-Yvette but with the AMS method. They can be considered reliable (Valladas, Kalteneker 2007:145).
On the middle Euphrates, on the Upper Tigris, and in the Djezirah, there are some sites that have some dates in the Younger Dryas: Mureybet, Tell Qaramel, Hallan Çemi, Qermez Dere and Nemrik 9 (Fig.1). However, as has been discussed above, some of these dates are problematic. The older radiocarbon dates of Mureybet of Niveau IA were only dated by the laboratory of Monaco, and the even older dates for this level of Louvain were discarded as aberrant (Stordeur, Evin 2008). The dates of Nemrik 9 and Tell Qaramel are also doubtful because many of them do not correlate with the stratigraphy and seem to be systematically old. The AMS-dates of M'lefaat support this suspicion. But it cannot be tested for the sites of Nemrik 9 and Tell Qaramel as all samples were dated only by the Gliwice laboratory.
A key-site for the Younger Dryas is Qermez Dere. Occupation of this site might go back to the Younger Dryas. That a decrease of dates does not correlate with a complete decrease of settlement activities is supported by the dates of Nahal Oren. Also, at Jerf el-Ahmar, the early settlement layers and the so called “Proto-Neolithic” layers of Jericho were not radiocarbon dated.
The radiocarbon dates of the far southern Levant show a completely different picture during the Younger Dryas (Fig.3). In the Negev as well as in southern Jordan, the dates of the so called “Harifian” sites Maaleh Ramon East and West, Abu Salem and Ramat Harif 3 cluster in this time frame. But also the Natufian dates of the site of Beidha as well as the “old- wood” dates of Wadi Faynan 16 date into this same time horizon.
One of the central problems of the early pre-pottery sites of the Near East is the distinction between the Khiamian and the presumably later pre-pottery Neolithic phases, the so-called Mureybetian and Sultanian (Ronen, Lechevallier 1999; Kujit 1997).
The radiocarbon dates demonstrate that neither in the northern nor in the southern Levant is a chronological distinction between the Khiamian and the late Natufian possible on the basis of radiocarbon dates alone.The Khiamian dates of Hatula and of Mureybet overlap with the dates of the late Natufian, although some dates of the Khiamian may be too old (Evin, Stordeur 2008). Its is thus still difficutl to decide whether the Khiamian started before or afater 9800 BC (s. Kuijt, Goodale 2006:44; Aurenche et al. 2001:1194).
The date for the commencement of the construction of the tower of Jericho remains an open question. As was just mentioned, Jericho’s proto-Neolithic layers have not yet been dated. Moreover, it is not clear how fast the occupation layers accumulated. Thus, the tower could be either older than, or contemporary with the first polyfunctional compartmented communal buildings of the Euphrates region. Anyhow, it was certainly used until Jericho Stage VI (PPNA) of Jericho and abandoned at the latest during Stage IX (MPPNB).
The site of Hatula was abandoned around 9300 BC. Consequently the layers attributed to the Khiamian and the Sultanian, and therefore also all burials in this site, must be older than the earliest large special buildings of Mureybet and Jerf el Ahmar, providing that the dates of Gif-sur-Yvette of Hatula are not too old. Therefore it is still uncertain whether in the Southern Levant so-called Khiamian sites were contemporary with the Sultanian and just reflect economical, social, and cultural differences or in fact represent a chronologically distinct phase.
Around 9200 BC +/- 100 years many new sites were occupied, including Djadé, Göbekli Tepe, and Netiv Hagdud. The occupation of Wadi Faynan 16 dates back at least to this phase and other sites on the middle Euphrates can be attributed to it, specifically Abr’3 (Yatah 2004; 2005) and Körtik Tepe (Prov. Diyarbakır) at the confluence of Batman Çay and the Tigris (Özkaya, San 2007).
According the earlier publication of de Contenson, the early settlement of Aswad should be dated to the same time frame. However, on the basis of the new excavation the chrono-cultural attribution of the early layers of Aswad has been revised and a pure PPNB settlement now postulated (Stordeur 2003a).
Whether these sites were continuously occupied can only to be determined by detailed stratigraphic and biological analyses. According to the radiocarbon dates, some of the sites were occupied during the Natufian and the PPNA, e.g. Mureybet, Tell Qaramel, Jerf el Ahmar, Qermez Dere, M’lefaat and Nemrik 9 (The hiatus of the dates in Mureybet between Phases IIA and IIIA is only an artefact resulting from the lack of dates for Phase IIB.)
Despite these unsolved questions, it is likely that the PPNA can be separated into two main chronological phases in both regions: an early phase starting around 9800/9700 BC and lasting until 9300/9200 BC, and a later phase starting around 9300/9200 BC. In the southern as well as in the northern Levant, the elaborate special buildings, the artistically decorated architectural elements, and the decorated stone bowls, shafteners and stone pebbles, document a vigorous social dynamic only from about 9300 BC onwards (Stordeur 2003b; Benz 2006; Aurenche 2007; Köksal-Schmidt, Schmidt 2007).These trends should be compared to the development of the flint typology and industry. For example, the cultural differences between ZAD 2 and the adjacent site of Dhra’ could be due to this chronological difference.
The early and middle PPNB
The next distinct chronological phase starts around 8700/8600 BC. This is the beginning of the PPNB as defined by Hours et al. (2004). A revision upward to 9200 BC (Aurenche et al. 2001:1195) seems doubtful because it would overlap almost completely with the second phase of the PPNA.
In the northern region Abbès and Stordeur (2002) also defined, on the basis of typo-technological developments in the flint industry, a ‘phase de transition’ lasting from 8800 BC to 8600 BC. Archaeological remains for this transitional phase were found at Jerf el Ahmar and Dj’adé. The dates of M’lefaat and Qermez Dere terminate around 8800 BC.
For the southern Levant, the definition of the early PPNB is disputed (Sayej 2004). There are very few dates for the very early PPNB, except for some from Motza and possibly Aswad, Horvat Galil and SefunimCave. The late dates of ZAD 2 also fall into this time span (Khalaily et al. 2007).
Between 8700 BC and 8600 BC the data show a clear break associated with the steep drop of the calibration curve. However, it is not yet clear whether this chronological break is the result of a true decline in settlement activities or merely a methodological problem (s. above). Some of the sites – Dja’dé, Mureybet, Nemrik 9, Aswad, Jericho and ZAD 2 – have a relatively continuous sequence of radiocarbon dates, whereas others like Göbekli Tepe were abandoned around 8600 BC. New settlements from this early phase of the PPNB are Çayönü, Nevalı Çori in the northern Levant, and Horvat Galil, NachchariniCave, Sefunim Terrace and Wadi Shueib in the central and southern Levant.
The next subphase of the PPNB can be discerned between 8300/8200BC and 7700/7500 BC. This sub-phase can be correlated with the beginning of the middle PPNB. It is especially obvious in the north and east, where the settlements of Dja’dé, Ain el-Kerkh and Nemrik 9 were abandoned. It also comprises the round- and grillplan building phase of Çayönü and the earliest settlement phase at Nevalı Çori. The oldest occupation phase of'Ain Ghazal starts at about 8400 BC, but its main occupation falls into middle PPNB and lasts until the late PPNB/PPNC.
Sefunim Terrace was abandoned 8300/8200 BC and Jericho itself might have been temporarily unoccupied. But new settlements appeared at Wadi Jilat 7 in eastern Jordan, and Nahal Hemar in the southern JordanValley, southwest of the Dead Sea.
Between about 7700 BC and 7500 BC many old settlements were abandoned and new sites occupied. This transition began first in the north around 7800 BC with the occupation of Abu Hureyra, Halula, Akarçay Tepe and of Ganj Darehin the east, and the central Levant followed slightly later, 7700-7500 BC, when the sites of Aswad, NachchariniCave, Yiftah’el and Wadi Shueib seem to have been abandoned. A gap in the radiocarbon dates during this time range is also documented for 'Ain Ghazaland Nahal Hemar. The middle PPNB thus ended at different times in the two regions, around 7800 BC in the north and about 7700-7500 BC in the south.
In the far southern Levant, there is only one date from the time span between 8800 BC and 8500 BC – which, however, the excavators attributed to the abandonment of the site (Finlayson, Mithen 2007). Only around 8400 BC are settlement activities again documented, the earliest at Wadi Faynan 16, and only slightly later at BeidhaA1, Shkârat Msaied, and Ghuwayr 1. Around 7600 the middle PPNB ended in the southernmost Levant, and from 7500 BC onwards, the late PPNB settlements of Basta and Ba’ja were occupied.
The late PPNB
As was mentioned, it is not possible, because of the steep decline of the calibration curve during that period, to confidently fix the beginning of the late PPNB to precisely 7500 BC. It is more probable that cultural change had already begun some 200-300 years earlier. Several of the sites – e.g. Çayönü, 'Ain Ghazal, Jericho, and Nahal Hemar – were occupied until the late PPNB. Many of the other sites were, however, abandoned. At the same time, there are numerous sites that were occupied for the first time during the late PPNB, including Halula, Tell el Sinn, Ramad, Es Sifiya, Khirbet Hammam, Azraq, Dhuweila, Ain Jammam, Ain Abu Nukhayla, Basta, and Ba’ja.
The end of the late PPNB is also difficult to fix. There are several reasons for this. For one thing, though many of the sites were abandoned at around 7000 BC, others seem to have been abandoned only 300 years later. Another problem is determining the demarcation between PPNC and the pottery Neolithic. However, a detailed discussion of this is beyond the scope of the present work.
On the basis of the above discussed radiocarbon dates, the following phases can be discerned:
PPNA (late Proto-Neolithic): Phase 1: 9800/9700 BC- Phase 2: 9300/9200-8800/8700 BC
Transitional phase: 8800-8600 BC
Early PPNB:8600-8300/8200 BC
Late PPNB:7800/7500-6900? BC
Fig. 4 demonstrates that three of the changes of the radiocarbon dates correlate with a steep slope and a following plateau in the calibration curve.
1.) Both the early and late PPNA correspond to plateaus in the calibration curve and the break around 9300/9200 BC could be theoretically caused by the dating method. However, the many cultural differences between the early and late PPNA testify to a significant cultural change during the latter phase.
2.) The second break corresponding to a steep slope in the calibration curve, is at around 8300/8200 BC. The next phase, the middle PPNB, synchronizes well with a plateau in the calibration curve.
3.) The final break, probably caused by the steep course of the calibration curve, is at about 7600/7500 BC.
Steep declines in the calibration curve correspond to climatic changes to cooler regimes and result in fewer observed radiocarbon dates for any given cultural phase. But it must be kept in mind that these ‘cultural’ breaks may not only be caused by methodology, but can in fact also reflect cultural and economic adjustments to climatic changes.
However, Fig. 4also shows three phases for which there are no methodological or obvious climatic explanations:
1. the so-called phase de transition between 8800 BC and 8600 BC,
2. the transition to the early PPNB at about 8600 BC,
3. the possibly earlier start of the late PPNB at around 7800/7700 BC.
Two tasks of future research are to (1) interpret these latter transitions, and (2) to determine whether the first three breaks (early to late PPNA, early to middle PPNB, middle to late PPNB) actually do reflect cultural changes or whether they are artefacts of the methodology of radiocarbon dating and changes of climate indicating spurious breaks in the occupation of the sites.
Although the radiocarbon dates of the Natufian layers of Beidha have higher deviations than +/- 150 years, they were calibrated. If they had not been used, a gap in the dates would have been appeared where there in fact is none.
Mithen and Finlayson (2007) discarded these dates as too old.
The radiocarbon dates of Tell Qaramel suggest that this site will be key site for dating the transition from the Younger Dryas to the Holocene. However, on the basis of the preliminary publications, it has not yet been possible to compare the material culture of this site with its chronology (Mazurowski 2004, Mazurowski et al. 2009).
The lower layers of Jerf el Ahmar have not been radiocarbon dated. The occupation of the site may have started before 9500 BC.
The dates of Göbekli Tepe only date the use of the site. The stone pillars were set up before these dates. An extraordinary exception to this observation and a very early illustration of this change would be the findings at Tell Qaramel (Mazurowski 2004, 2003, 2002, 1999; Mazurowski Yartah 2001; Mazurowski, Jammous 2000 ), assuming the radiocarbon dates of this site are correct.  The typological attribution of the site of Iraq ed Dubb is interesting, as its lithic material differs from that of Dhra’ but parallels the one of Netiv Hagdud, although, on the basis of the radiocarbon dates, Iraq ed Dubb should be contemporary to Dhra’. However, the differentiation of the layers and the flints at Iraq ed Dubb is difficult because of the mixing of older material from the lower Natufian layers (Kuijt, Goodale 2006).  There are, however, some dates that may document a continuous occupation of Abu Hureyra.
The radiocarbon dates of Halula range between 7750 BC and 7300 BC. Although it is traditionally attributed to the middle PPNB (e.g. Willcox et al. 2009), according to the above mentioned earlier start of the late PPNB in the north, it might also be attributed to the late PPNB.
Neolithic Heritage Trail
Basta (u.c.) Joint Archaeological Project
PIGPA Project (u.c.) Palaeoenvironmental Investiga-
tions in the Greater Petra Area
PPND Plattform for the publication of Neolithic Radiocarbon Dates
'Ain Rahub Project (u.c.)
Fig. 1 Compilation of sums of calibrated radiocarbon dates of the northern Levant, middle Euphrates (blue), southeastern Turkey (green), and the Djezirah and Zagros (orange-red-violet).
Fig. 2 Compilation of the sums of calibrated radiocarbon dates of Epipalaeolithic and pre-pottery Neolithic sites (2.1) and of sites exclusively dating to the early Holocene (2.2): Central Levant-Damascus-to Galilee-Region (blue); Mediterranean Coast and Mount Carmel region (green); Jordan Valley and Highlands of Amman, Dead Sea (red); Judean Mountains (orange); Eastern Jordan Desert (black).
Fig.3 Compilation of the sums of calibrated radiocarbon dates of sites south of the Dead Sea: Negev (red); Wadi Faynan to Wadi Rum (green); discarded dates/dates with too high standard deviations (white).
Fig. 4 Succession of chronological phases according to calibrated radiocarbon dates.