Chapter 5: Neolithic 3 Byblos (Pages 329-339)

Pre-History and Archaeology Glossary

Excerpts and Definitions and Addendums:

Byblos, now the little town of Jubail, lies about 30 kilometres northeast of Beirut on the Lebanese coast. To the east of the site there is a narrow though fertile coastal plain and then the Mountains of Lebanon which rise steeply behind. The ancient settlement is situated on a promontory just to the south of a tiny inlet. This inlet is the old port of Byblos which is still used as a fishing harbour. On the south side of the prommontory there is another cove at the mouth of a little valley which ran across the site before it became filled with occupation debris. This cove was probably a subsidiary landing place in ancient times; it may have been used more by the Neolithic inhabitants than the inlet to the north which lay further away down a steep slope. Beyond this cove is an open sandy beach which would have been a good landing place in fine weather.

It is important to remember that the sea probably only reached its present height during the 6th millennium as we saw in Chapter 1 so that it was not until Neolithic 3 that Byblos could have served as a port. By the same reasoning it would only have been during this period that the bay of Minet Beidha to the west of Ras Shamra would have assumed its present configuration and provided a convenient harbour for the large settlement a little way inland. We know that fish were eaten in considerable quantities at Byblos during this period since their remains comprised 7% of the bones identified from the site and they were still probably consumed at Ras Shamra as fish vertebrae had been recovered in the Neolithic 2 levels. It is likely that by now some of these fish were caught from boats at sea which could have conveniently been launched from the new harbours. We have no evidence that maritime trade had commenced along the Levant coast in Neolithic 3 though since we know that small quantities of many materials were being exchanged between sites it would not be surprising if a little of this traffic was conducted by sea. In the absence of positive evidence, however, we must conclude that the potential of the harbours at Byblos and Ras Shamra was not realised until much later.

The ancient site of Byblos has been excavated almost continuously by Danand for over 40 years. At least 1.5 hectares has now been cleared to bedrock so that a greater area of the prehistoric settlement at the base has been exposed than on any other site in the Levant. A considerable amount of information has thus been recovered about the structures of the successive Neolithic settlements and the artifacts used by their inhabitants. The original topography of the site has also been determined fairly precisely.

The promontory at Byblos once consisted of two hills, one higher than the other, separated by a little valley in which lay a spring of good water. The higher of the two hills was on the west side of the valley. The first substantial settlement at Byblos was established on the seaward slope of this hill and later spread south into the valley. This phase of settlement has been designated Neolithique Ancien by Dunand. The principal attractions of the site have always been the spring, the protected landings for boats and the fertility of the immediate hinterland; presumably these factors also induced the inhabitants of the Neolithique Ancien site to settle here. We may note in passing that the site was occupied briefly at an earlier period. A small deposit was found which contained no pottery but which yielded a tanged and notched arrowhead as well as several microliths, one of them backed with Helwan retouch. This suggests that a group may have lived here for a while in one of the two earlier Neolithic stages or Mesolithic 2.

The debris of the Neolithique Ancien settlement was spread over about 1.2 hectares, an area that would have been more extensive originally since part of the site has washed away on the seaward side. Much of the deposit was composed of occupation soil, building remains being concentrated on an area of only 5000 square metres. Dunand estimated that about 20 houses were occupied at one time and that the settlement was no more than a small village. These houses had been robbed for building materials so that in some instances little of the original dwelling remained. It seems probable to me that there were more houses in the village than Dunand suggests and that some of these were completely destroyed by robbing and disturbance after they were abandoned. Thus the likely area of intensive occupation would have been greater than the 5000 square metres proposed by Dunand.

The houses were rectangular with a single room to which other more lightly built structures were sometimes added. These houses varied in size but several were about 5 metres long and 4 metres wide. Their walls were built of stone but none was found standing more than 1 metre high. The entrance was in the long side of a room. The floors were usually made of hard white lime plaster laid on a bed of pebbles which sometimes curved a little way up the walls. The surface of the plaster was then polished.

Dunand did not think that the stone walls of many of the houses ever stood much higher than 1 metre. This led him to suggest that the houses were framed and roofed with poles covered by mats or skins. He believed that the wall posts must have been set up outside the low stone walls as no trace of roof supports was found within the buildings; we may note in passing that he does not report that post-holes were found along the outsides of the walls either. The stone walls and plaster floors of the Neolithique Ancien houses were stoutly and carefully built so it is probable that their superstructures were completed in a similar fashion. We have already seen that the walls of these buildings were damaged by robbing, an observation supported by the absence of collapsed walling within them. The walls may have been built high enough to support the roof but subsequently were reduced by robbing to no more than stubs. If the walls had originally carried the roofs that would explain why no trace of wall or roof supports was found inside or outside the houses.

Several of these houses had a little platform built inside against one of the short walls. These may have been hearths but Dunand preferred to think of them as associated with a domestic cult. In one house a mortar had been set in the floor while in others a depression was found on the floor where mortars or querns had probably stood. Scattered on the floors were grinding tools and other domestic artifacts. Outside the houses stubs of walls and other stones were used as benches.

The houses were aligned either north-south or east-west. This happened to be the most convenient way to build them because of the slope of the ground but it also meant that the doorways of many of the houses faced the sun. The houses were rebuilt on the same spot on about the same alignment usually three times but in one area as much as six or seven. Each house stood alone and was separated from its reighbours by large open spaces. A little paving was found in these open spaces and also heav stone grinding tools left in place, bedrock mortars and hearths.

The dead were buried in the settlement between the buildings. The corpses were laid in a crouched position on their left sides in shallow graves; the bodies of infants were buried in jars. Two groups of adult burials were noted, one in which the bodies were placed in single graves with a few artifacts and a second in which the corpses were laid on a bed of stones with more grave goods; the former was more common. The accompanying artifacts consisted of flint tools, polished stone axes, pottery and occasional ornaments. 33 burials were found altogether, not many when one considers the size of the settlement and the length of time it was occupied. Dunand noted that some graves may have been destroyed by building activity in this and later phases and also remarked that not all the dead may have been buried within the village anyway.

The inhabitants of the Neolithic settlements at Byblos obtained their flint from nodules on the floors of neighbouring wadis. In the Wadi Deir Banat which reaches the sea just south of Byblos beds of excellent flint exposed in either bank were apparently quarried for raw material. Not many cores were found on the site which implies that blades were usually prepared where the flint was quarried. Such cores as were recovered were pyramidal or double-ended in type. 18 pieces of obsidian were found in the Neolithique Ancien settlement three of which have been analysed spectrographically. Two came from the 1 e-f source, probably Acigol near Kayseri and one from Ciftlik (See Pages 63 and 65 in *1 Below).

Most of the flint tools found at Byblos were of four classes: sickle blades, arrowheads, burins and axes with chisels. The sickle blades were all made on segmented blades and hafted together to form composite sickles. They were retouched straight or obliquely at one or both ends and some were abruptly retouched along the back. The cutting edge was denticulated but sometimes both long sides were retouched in this way. A great many truncated blades were found in one spot in the Neolithique Ancien settlement which were in the process of being made into segmented sickle blades. The place where they were found was thus a working floor for the manufacture of these tools.

The sickles of which these segmented denticulated blades make up the cutting edge are poor tools with which to cut cereals although they could have been used to scrape ears of wheat off the stalk into baskets. It has also been suggested that they may have been used to cut down plants with tough siliceous stems such as reeds. The problem is difficult to resolve since there is evidence for both cereal agriculture and the use of reeds at Byblos and at other sites on which these tools have been found. As I have explained elsewhere (See Page 43 in *2 Below), we do not know if these sickles were used for harvesting cereals or some other plant and so cannot definitely associate them with agricultural activities as Cauvin has done.

Most of the arrowheads were of three types, all of which were tanged. Two were the Amuq points 1 and 2 and the third was what Cauvin has called the Byblos Point. Byblos points had a tang defined by pronounced shoulders or occasionally a pair of notches. They were retouched by pressure-flaking around the tang, tip and part or all of the shaft. A few were from 10 to 17.8 centimetres in length but most were between 5 and 10 centimetres long. Several leaf-shaped arrowheads retouched by pressure-flaking were also found. Some very large oval retouched blades retouched in a similar manner were classed as daggers by Cauvin.

The three principal types of burins found at Byblos were burins on lateral preparation, burins on truncation and dihedral burins. Some single or multiple blow burins on a break or natural surface were also recovered as well as a few of other types.

The Byblos axes were usually made of flint but there were some examples in limestone, granite and basalt. The commonest type was trapezoidal or rectangular in shape, flaked all over and with abruptly retouched sides. The cutting edge of most of these had been flaked almost straight and then polished. A group was of much the same shape but with vertically flaked sides; these were also thicker in cross-section and heavier. The third group was composed of almond-shaped axes, again usually with a polished cutting edge. Almost all of these tools were probably used as axes but a few with asymmetric cross-sections may have been used as adzes.

Among the other heavy flint tools were a number of chisels made in a similar fashion to the axes. They were quite narrow and plano-convex or biconvex in lengthwise cross-section. They were flaked all over and had a narrow cutting edge which was usually polished. Both these and the axes were probably used to cut and shape timber. The coastal plain and the mountains behind the site had a dense cover of forest at the time the Neolithic settlement was occupied and these tools indicate that much use was made of this raw material.

Several small greenstone axes and chisels were also found in the Neolithique Ancien settlement. These were made of steatite and two kinds of hard stone, amphibrolite and gabbro. The steatite axes could hardly have been used as tools but the others when hafted would have been most serviceable artifacts for finishing wooden objects which had already been partly shaped by the axes and chisels.

The remainder of the Neolithique Ancien assemblages at Byblos consisted of a few borers and scrapers with a number of notched and denticulated pieces. Most of the borers had broad points which were more suitable for piercing materials such as leather rather than making deep holes. The scrapers included both end-scrapers on blades and side-scrapers on flakes.

Other kinds of stone tools were common at Byblos. The querns were saddle-shaped and made of limestone or more rarely of sandstone or basalt. The rubbers with which the grain was ground were of the approximate plano-convex type. Mortars and pestles were made of the same rocks as the querns and rubbers.

Stone dishes were used in the Neolithique Ancien settlement but not in great numbers. They were circular or elliptical in shape with a shallow depression in the centre. The bases were rounded in the simplest ones, flat or hollow in tht others. One retained traces of pigment in the bowl. Several coarse porous stone vessels had been coated with lime plaster to make them watertight.

The Neolithique Ancien pottery was simple in technique and shape. Thick sherds of a buff coarse ware were found but no vessels of this fabric could be reconstructed. The more common ware had an even fabric with a little straw, limestone and quartz filler. Vessels made of this fabric were carefully smoothed on the outside and fired until the surface was buff in colour and the core buff or pink. The surface of the vessels was then lightly burnished or in a few cases highly burnished to an even shiny finish. None of these vessels, however, resembled the dark grey burnished or dark polished ware of sites in the North Syrian group although the technique of manufacture was much the same at Byblos as further north.

The shapes of these Byblos vessels were of the simplest shapes. The two principal classes of vessels were globular jars and hemispherical bowls. Some of the jars were hole-mouths while others had everted rims; a few had collared necks. These jars were mostly quite large vessels but some with the same shapes were no more than 13 or 15 centimetres high. Several of the larger vessels had ledge handles or sets of two or four pierced lugs for suspension.

The bowls also varied markedly in size. Some with handles are best described as cups while others were quite deep. A few of these bowls were carinated. The fabric of these vessels seems to have been quite porous for a number of the bowls were lined with white plaster which was then polished, presumably to give them a waterproof lining. In addition to the bowls there were a few open dishes with ring bases and also two rough spoons. Almost all the jars and bowls had rounded bases but a few had flat bottoms or even ring bases like the dishes.

Many of these vessels were decorated with impressed or incised designs. The most characteristic patterns were made with the back of a cardium shell which was pressed into the clay all over the surface of many of the globular jars. Other patterns were made by scraping with the edge of the shell over the surface of the pots. A few sherds of cord-impressed ware were found but this type of decoration was not much favoured.

Many of the bowls and a few of the jars were decorated with rows of horizontal lines often just below the rim. Sometimes the surface was divided by groups of vertical lines and the spaces between filled with stab marks. Another common design was rows of pendant triangles or loops incised below the rim lines and then filled with stabs. Rows of oblique lines or herringbone patterns were also incised around the pots. These incised patterns were sometimes filled with a white paste. A few sherds were found which had been decorated in relief with applied clay pieces but this technique was rare at Byblos.

The inhabitants of the Neolithique Ancien settlement also made a few white plaster vessels. These were all open bowls on ring bases. The very large plaster jars found on some Neolithic 2 sites were not at Byblos in Neolithic 3.

A large number of bone tools was preserved at Byblos. 51 borers made from sheep or goat metapodials were found. Almost all were the same type with a stout handle and fine point. There were 18 spatulae mostly made of sheep or goat bones. Among the more unusual tools were six fishhooks and a needle. There were also eight bone hafts and objects of adornment such as beads and amulets. A curious object was a sheep spatula with a row of incisions along one edge which Dunand interpreted as a musical instrument. It could perhaps have been used in this way but more probably served as a counting or recording device.

Among the other tools was a series of baked clay spindle whorls with a biconical cross-section. Several stone discs pierced through the centre may have served the same purpose. Another curious group of objects consisted of several discs and rectangles made of sherds which had a cut around the edge. Dunand thought a cord might have been wrapped around the edge and pressed into the surface of new pots to produce cord-impressed patterns; however they were used, they and the cord-impressed ware were only found in the Neolithique Ancien settlement.

A distinctive class of artifacts at Byblos was the stamp seals made for the most part in baked clay although a few were cut from stone. Dunand divided these into two groups, a series of baked clay pintaderas and an assortment of other seals. The pintaderas were oval in shape with a curved sealing surface and a knob on the back to be gripped by a thumb and forefinger. Simple patterns of chevrons, concentric ovals and straight or oblique lines were incised all over the sealing surface. The other seals made of baked clay and stone had a flat surface with a similar range of simple incised patterns.

The two other main classes of artifacts were objects of adornment and figurines. The beads were discoid or cylindrical and made of greenstone, steatite and carnelian as well as dentalium shells. Rectangular shell pendants were also found and a number of carved bone amulets. More unusual were two tiny squatting human figures carved in a crystalline greenstone and pierced for suspension. Several small grooved stone and baked clay objects were also recovered which may have been nose ornaments, labrets or buttons.

One group of figurines consisted of long pebbles with a few lines incised at one end to indicate human features. There were fifteen of these of which five were found in one area and six in another so most of them were made or used in two restricted locations. The other figurines were made of baked clay but there were very few of these. Several were recognisably human and some others were four-legged animals but of what species could not be determined. One of the human figurines was a stylized type found on other sites. It was a female with arms placed against the front of the torso; the head was pointed and the eyes were marked with incised blobs of clay shaped like coffee beans.

It is possible to estimate the duration of the Neolithique Ancien phase at Byblos from two C-14 determinations made on charcoal samples from different layers. The first sample which came from a level in the middle of the Neolithique Ancien sequence was collected and processed in 1957. The Groningen laboratory at first obtained a date of 7000 +/- 80 B.P. or 5050 ± 80 B.C. for this sample. They sent the result to Dunand who published it as 5043 +/- 80 B.C. For many years it was thought that this was the true figure from which it was estimated that Neolithique Ancien Byblos was first occupied in the second half of the 6th millennium. This determination has since been corrected by the laboratory which has now published it as 5410 ± 70 BC GrN-1544 (See Page 50:14:1972 in *3 Below). This figure is nearly four centuries earlier thus considerably lengthening the Neolithique Ancien phase of occupation. It now appears that Neolithique Ancien Byblos must have been occupied in the first half of the 6th millennium, perhaps about 5600 or 5700 B.C.

The second sample was collected in 1955 from a level late in the Neolithique Ancien sequence. When processed in 1957 this gave a date of 4600 ± 200 B.C. W-627 (See Page 183:2:1960 ibid) which was published by Dunand as 4593 ± 200 B.C. The first point to note about this determination is that the standard error is quite high. Secondly, the determination was made long ago and has not since been corrected. Samples processed more recently from Neolithic sites in the Levant have tended to give earlier dates than those obtained when the technique was being developed. I think, therefore, that the Neolithique Ancien phase at Byblos ended about 4800 or even 5000 B.C. rather than 4500 B.C. as this determination would suggest.

The chipped stone industry of Neolithique Ancien Byblos is noticeably more developed than that of the Neolithic 2 sites of Tell aux Scies, Saaideh and even Tell Labweh in Lebanon and other Neolithic 2 sites in the West Syrian group. It would appear that several centuries elapsed between the end of Neolithic 2 about 6000 B.C. and the foundation of Neolithique Ancien Byblos. This lends weight to the suggestion based on the evidence of the C-14 determination GrN-1544 that the settlement of Neolithique Ancien Byblos was founded about 5600 or 5700 B.C. rather than at the beginning of Neolithic 3. It then appears to have been occupied throughout Neolithic 3 until the end of the 6th or early in the 5th millennium BC .....