NEAR EASTERN ARCHAEOLOGY 74:2 (2011) 105 The Iron Age Chronology Debate: Is the Gap Narrowing? Another Viewpoint Amihai Mazar The Hebrew University of Jerusalem Introduction The following is a response to an article in this journal by Israel Finkelstein and Eli Piasetzky (2011) summarizing their current opinion on the Iron Age chronology debate, which is an abridged presentation of recent more detailed studies by these authors (Finkelstein and Piasetzky 2009, 2010). The title of their article includes the question: “Is the gap narrowing?” I will try to explain that the gap indeed is narrowing, yet the authors went only halfway and there are still significant points of dis- agreement (see already Mazar 2005, 2008; Mazar and Bronk Ramsey 2008, 2010; see table 2 on p. 111 for a presentation of the Modified Conventional Chronology presented below). Methodological Comments Much of the current Iron Age debate is based on hundreds of 14 C dates measured during the last decade at various sites in Israel and Jordan (Gilboa and Sharon 2003; Mazar et al. 2005; Sharon et al. 2007; Levy et al. 2008; Garfinkel and Ganor 2009, 35–48). Most of the results of these studies were evaluated by using Bayesian models constructed with the help of OxCal soft- ware developed by Christopher Bronk Ramsey of Oxford Uni- versity. But it has to be asked whether Bayesian models, where hundreds of dates from various sites are mixed together, are the proper tool for such investigation. The answer is complex and involves both archaeological and statistical queries. Bronk Ramsey himself doubted if the Bayesian models are sensitive enough when so many samples from various sites are being investigated and when there are suspected gaps in the sequence of available dates (such as the early Iron IIA phase; see Mazar and Bronk Ramsey 2008, 178). In the future, new statistical models may emerge to help resolve this quandary. 1 For the time being, I prefer to assess the radiocarbon dates from each individual site independently, using as much as possible high- quality dates from laboratories that have successfully withstood the procedure of interlaboratory comparisons conducted dur- ing the past decade (Boaretto et al. 2005). We also observe that, when several repetitious measurements from the same sample are made, we often get a wider range of dates (Frese and Levy 2010). We must be aware of these and other problems related to radiocarbon dating when dealing with a large number of dates and when the debate is sometimes over only a few dozen years. Yet, in spite of the skepticism of some scholars as to the valid- ity of radiometric dating in such a subtle debate, some of the results are extremely significant. The Beginning of the Iron Age Finkelstein and Piasetzky (2011, 52, fig. 2) differentiate between Late Bronze III—the time of the Egyptian Twenti- eth-Dynasty presence in Canaan until circa 1130 b.c.e.—and Early Iron I, which they date to circa 1130–1050 b.c.e. In their opinion, the initial Philistine settlement occurred only after 1130 b.c.e. This issue is highly controversial. While several scholars accept this division, others, including myself, adhere to the traditional terminology/chronology (Stern 1993, 1529), according to which most of the twelfth century b.c.e. (the time of the Twentieth-Dynasty presence in Canaan) is defined as Iron IA and is characterized by the initial Philistine settlement (monochrome ware), the beginning of settlement in the hill country (Giloh assemblage), and the continuity of Canaanite and Egyptian centers at certain lowland sites such as Megiddo VIIA, Lachish VI, and Beth-Shean VI. Radiocarbon dates can- not resolve this issue due to the shape of the calibration curve at this time (Manning 2006–2007). 2 Finkelstein’s low chronology of the Philistine culture creates many unresolved problems, one of which is a presumed total occupation gap in the southern coastal plain sites where Philistine monochrome pottery is lacking, as well as creating an unexplained time gap between the presence of this pottery in Canaan and very similar pottery in Cyprus (Mazar 2007; Mazar and Bronk Ramsey 2008; Mazar 2009, 23–25; Sherratt 2005). The Transition between Iron I and Iron II The conventional date for the Iron Age IB in the second half of the twelfth and the eleventh centuries is supported by many 14 C dates from various sites (Sharon et al. 2007; Mazar et al. 2005; Mazar and Bronk Ramsey 2008). However, the conventional date of circa 1000 b.c.e. for the transition from Iron I to Iron II was challenged by the Low Chronology system, wherein Finkel- stein’s initial suggestion to move all tenth-century b.c.e. contexts to the ninth century, and thus change the entire archaeological profile of the tenth century b.c.e., was later supported by Sha- ron, Gilboa, Jull, and Boaretto (2007), based on the results of their Iron Age Dating Project as estimated by Bayesian models. One of Finkelstein’s basic claims was that Shoshenq I destroyed Megiddo VIA (the last Canaanite city at Megiddo) as well as other sites that mark the end of the Iron Age I (Finkelstein 2002). This conclusion, which I refuted on the basis of archaeo-