Chapter 2: MESOLITHIC 1 (Pages 37-42)

Pre-History and Archaeology Glossary

Excerpts and Definitions and Addendums:

Mesolithic 1 throughout the Levant followed the Aurignacian or Levantine Upper Palaeolithic. The Aurignacian was divided into three phases by Neuville called Upper Palaeolithic III, IV and V, the last being transitional between the Aurignacian and the Mesolithic. Garrod believed that the Aurignacian of the Levant did not sufficiently resemble the Aurignacian of Europe to justify using the name so she introduced the terms Lower and Upper Antelian to describe stages III and IV of Neuville. She also invented another term Athlitian to describe the material from Mugharet Wad which roughly corresponded with Neuville's transitional phase V (See Page 41 in *1 Below). Copeland has recently introduced new terms to describe the material from Lebanon and Syria using Ksar Akil as the type-site (See Page 342 in *2 Below). She uses Levantine Aurignacian phase C for Ksar Akil Levels 8 to 6 which corresponds very approximately to Neuville's stage V. This phase is marked by very varied assemblages at most sites on which it occurs, emphasizing its transitional nature.

It seems fairly certain that Mesolithic 1 developed directly from the Aurignacian. There are indications that a gradual change took place in the stone industries. Microliths and retouched bladelets are found for the first time in the Aurignacian deposits of El Khiam E, Kebara D (See Page 177 in *3 Below) and Ksar Akil (See Page 343 in *2 Below). Two sites at least, Ksar Akil and Yabrud III seem to have fairly complete sequences that span the transition from Aurignacian to Mesolithic 1. Perrot has also claimed a degree of continuity from level E to D at El Khiam while at Kebara itself no break was detected in the stratigraphy between D and C (See Page 186 in *3 Below) although the Athlitian was missing here. On the other hand there was a break in the sequence between the Aurignacian and the Kebaran at Nahal Oren (See Pages 37ff in *3.5 Below) while at Mugharet Wad there was no Kebaran occupation at all after the Aurignacian of layer C. Such breaks may partly be a reflection of the way of life of the inhabitants of these sites. It is probable that such groups used sites intermittently over several years and then left them unoccupied for long periods.

The stone industries from Athlitian sites are so diverse typologically that Bar-Yosef has questioned even the existence of this phase (See Page 26 ibid). Others believe that a number of sites can be included in this stage despite the differences between them. Such sites would be Mugharet Wad C, El Khiam E, the Nahal Oren terrace and Ksar Akil 6. Changes of some kind were taking place at this time which manifested themselves in the stone industries first of Athlitian sites and then of Mesolithic 1. The latter with its large proportion of microliths is very different from the Aurignacian.

Several factors brought about these changes. One of the most crucial was the abrupt deterioration in climate and environment that set in about 20,000 or 18,000 BC when the last cold phase of the Pleistocene began. We have seen that the Levant became relatively arid and the forest vegetation retreated to be replaced by steppe. The inception of this cool dry phase just preceded the beginning of Mesolithic 1 on present evidence. The hunter-gatherers of the Aurignacian would have had to modify substantially their way of life, particularly their settlement pattern, to take account of changing conditions. This shift to a new adaptation may be responsible for the confused nature of Athlitian deposits. The crystallization of this new pattern resulted in Mesolithic 1; not only was the settlement pattern modified but a new stone industry of distinctive character was developed.

The inhabitants of the small Mesolithic 1 sites found in the Levant left behind little more than their chipped stone tools. Any definition of the Kebaran thus rests upon a description of the characteristic features of this chipped stone industry (See Page 168ff ibid). The industry was essentially one of small tools made on bladelets struck off single-platform cores. Much of the waste also consisted of bladelets. There were rarely more than a few blades on any site in contrast with the Aurignacian. The most numerous and typical tools were non-geometric microliths, particularly microgravettes, backed bladelets and backed truncated bladelets. The only other tools found with any frequency were end-scrapers and burins. A few bone tools and some of ground stone have been found on Mesolithic 1 sites but little else.

Before considering the distribution and nature of Mesolithic 1 sites I will present the little information we have about the chronology of this stage. Mesolithic sites are far less numerous than those of the Neolithic and only a few have been dug recently. Of these not all have yielded material suitable for carbon 14 dating so that there is a dearth of chronological data. Those dates which are available suffice only to fix Mesolithic 1 approximately in time and to establish its duration. Following standard practice all C-14 dates I shall quote in this thesis will be calculated on the Libby half-life. In presenting the determinations in this form I recognise that they will sometimes differ by as much as several centuries from the absolute age of the samples which they date. At present it is not possible to estimate except in the most general way the absolute age of dated samples since the whole Mesolithic and much of the Neolithic lie beyond the range of the calibration curves published so far. I discuss this question more fully in the Appendix but it should be pointed out here that uncorrected dates still provide a sound sequence of relative chronology. Carbon 14 is the only method of near absolute dating that has been used for archaeological sites and contemporary environmental phenomena during the Mesolithic and Neolithic in the Levant so that chronological data from these two sources may be directly compared to build up as complete a picture as possible of the evolution of human society in the Levant during these stages.

In order to determine when Mesolithic 1 began it is necessary to consider dates from sites occupied late in the Aurignacian and also the few Kebaran sites in Palestine from which C-14 determinations are avilable. There is one date of 26,890 ± 380 B.C. GrN-2195 (See Page 173:5:1963 in *4 Below) from level 8 at Ksar Akil. This is supported by new, as yet unpublished, dates for slightly later material from the site (Tixier quoted in Page 343 in *2 Below) but these dates are believed to be rather early for such a late stage in the Aurignacian. The other dates are all from Ein Aqev, an Aurignacian site in the Negev. The samples for these determinations were collected in stratigraphical order. From bottom to top they are: 18,080 ± 1200 B.C. SMU-5; 15,440, ± 560 B.C. SMU-8; 15,940 ± 600 B.C. SMU-6 (See Pages 378-379:16:1974 in *4 Below); 15,560 ± 290 B.C. I-5495 and 14,950 ± 250 B.C. I-5494 (See Page 295:15:1973 ibid). Again these determinations are almost consistent internally.

The Ein Aqev dates pose a problem because most of then overlap with the earliest dates now available from Kebaran sites. At Ein Avdat only 3.5 kilometres from Ein Aqev there is a site (D5) classed as Geometric Kebaran A which has given the following dates: 16,890 ± 680 B.C. SMU-7 (See Page 379:16:1974 ibid); 13,870 ± 1730 B.C. Tx-1121 (See Page 484:14:1972 ibid) and 11,220 ± 230 B.C. I-5497 (See Page 296:15:1973 ibid). This is not a very satisfactory series: SMU-7 is considered suspect by the excavator of the site because it seems so much earlier than the other dates and Tx-1121 has such a large standard error that it could really be much earlier or later. It remains possible however that Ein Avdat does have scme occupation deposits dating from early in Mesolithic 1.

Two sites in northern Palestine, Nahal Oren and Rakafet, have Kebaran levels that are dated very early. From Rakafet there is a single determination of 16,960 ± 330 B.C. I-6865 (See Page 96 in *5 Below). This is supported by one date from Nahal Oren of 16,300 ± 320 B.C. UCLA-1776C from layer IX (See Page 77 ibid) near the beginning of the Kebaran sequence on the site. There are also later dates in the sequence of 14,930 ± 34O B.C. UCLA-1776B and 13,850 +/- 300 B.C. UCLA-I776A (ibid). This series of dates is consistent and the standard errors are small. The layers dated by I-6865 and UCLA-1776C are not however thought to be at the very beginning of the Kebaran sequence in this region (See Page 96 ibid).

The Nahal Oren dates in particular suggest that Mesolithic 1 began before 17,OOO B.C. and possibly as early as 18,000 B.C. in Palestine and perhaps Lebanon. The Ksar Akil dates would not preclude the possibility that it began earlier still. This estimate would include the Ein Avdat dates but does not resolve the problem of those from Ein Aqev. It is possible of course that Mesolithic 1 began somewhat later in the Negev. It should be borne in mind however that the flint industries of the Aurignacian and Mesolithic 1 in the Negev do not resemble very closely those from further north in the Levant. When the assemblages from Ein Avdat (See Page 20 in *6 Below) and Ein Aqev (See Page 4 in *7 Below) are compared they do not appear to be very different. It may be that both represent a range of activities carried on by similar groups over a long period of time.

The change from Mesolithic 1 to Mesolithic 2 can be dated more closely though still not very exactly. This is because the dates available from Mesolithic sites are mostly derived from early phases in the sequence so one has to rely on dates from early Mesolithic 2 sites to arrive at an estimation. The latest date for a Mesolithic 1 site anywhere in the Levant is 12,150 +/- 500 B.C. Mc-411 (See Page 337:15:1973 in *4 Below) from Ksar Akil. The earliest date from a Mesolithic 2 site is 11,140 ± 200 B.C. I-5496 (See Page 295:15:1973 ibid) from Rosh Horesha in the Negev. This determination may be inaccurate as it does not agree with two others taken from similarly stratified samples at the same site; those determinations are 8,930 +/- 280 B.C. SMU-10 and 8,540 ± 430 B.C. SMU-9 (See Page 379:16:1974 ibid). There are several determinations from three other sites that fall in the 10th millennium. These are 9,200 ± 400 B.C. for Kebara B, two dates from Mugharet el Wad B2 of 9,970 ± 660 B.C. and 9,525 ± 650 B.C. (See Page 291 in *8 Below) and a date from Jericho of 9,216 ± 107 B.C. P-376 (See Page 84:5:1963 in *4 Below), although two other samples from the Natufian deposits here have given later dates of 7,850 ± 240 B.C. F-72 and 7,900 ± 240 B.C. F-69 (See Page 8 in *9 Below). None of these dates is entirely satisfactory: the Kebara and Mugharet Wad determinations were all made recently on bones from skeletons excavated many years ago (See Page 99 in *8 Below) and the Jericho samples were processed when carbon 14 dating was still quite new. Nevertheless this series of dates forms a consistent pattern which suggests that the transition from Mesolithic 1 to Mesolithic 2 in Palestine at least took place about 10,000 B.C. or possibly a little before. This estimate can only be approximate for the moment since the evidence is so uncertain; it ignores the possibility that a version of the Kebaran continued on beside the earliest Natufian (See Page 173 in *3.5 Below) ...