En Flintplads i Qatar


  • Hans Jørgen Madsen


qatar, flint sites, flintpladser, umm al ma, salwa, reconnaisance, rekonnosering


A Flint Site in Qatar.

The Danish Archaeological Expedition to Qatar under Professor P. V. Glob's leadership, concentrated, in its 1960 campaign, on a thorough reconnaissance of Qatar's west coast, from Umm al Ma in the north to Salwa, near the Saudi-Arabian border, in the south. The terrain is dominated by a ridge of limestone cliffs 30-100 m. high, dividing the narrow coastal strip from the great expanses of desert which make up the interior of the peninsula. The two expedition members were the sculptor Holger Kapel and the author of this article. We were able to cover nearly the whole of the 60 km.-long area of investigation from our base in Dukhan, where the Qatar Petroleum Company and the police were superlative in their efforts to make our work easier, so that we could devote all our energy to the primary task of archaeological survey.

To the previously known sites 1) were added: over 100 burial mounds, 2 probable palaeolithic workshops, a number of finds characterised by large round discoid scrapers (skiveskrabere), several sites with varying inventory and finally, 3 finds, one large and two small, where a very pronounced technique of surface chipping, which will be described below, was evident. A total of 29 sites with implements was located in the 1960 campaign, besides a number of sites with only flint waste.

The large site with surface-chipping technique mentioned, proved to be the largest Stone Age find of the Arabian Peninsula. It lies on blown sand in a horse-shoe shaped valley, which opens onto the sea 5 km. south of Dukhan (fig. 1). Here and there the sand has conglomerated, the flint being embedded in a very porous sandstone. The site covers about 2½ acres and is 150 m. from the shore. Excavation at various points convinced us of the expediency of surface collection, the biggest concentration of artifacts occurring on the desert surface, incidentally enabling us to keep within our terms of reference. An area 8 X 1 m. was divided into an upper layer (from the surface to a depth of 2 cm.) and a lower layer (from 2-10 cm.). The latter contained only 2 implements and 347 chips, although it was four times as thick as the upper layer which contained 5 implements and 688 chips. Thus a relative frequency of 10:1 for implements was shown and the same tendency observed in other spaced trial diggings justifies the view that surface collection yields at least a usable statistical material. The high figures for swarf in this connection are due to the counting of even the smallest chips which the technique used resulted in. In an area of 130 sq. m., 121 finished tools or fragments of blanks were collected, giving an idea of the wealth of material.

Finds were almost exclusively of worked flint. Only a pearl, a piece of a spindle-whorl and some sherds fell outside this category, and the sherds are hardly contemporary, as similar pottery is in use locally at the present day. We have, however, good reason to suppose that the flint material constitutes an integral find, characterised as it is by an extremely uniform technique. Surface-chipping (fladehugning) dominates completely. It is clearly the core and the thin surface-chipped lamina (flintblad), not the blade (flække) or flake (skive), which is the basis for the whole inventory and not a single piece showed any trace of polishing. Whitish, honey-coloured, reddish and greyish-black flint, of varying quality, was used. Different-coloured flint is quite common in western Qatar, so there is no reason to suppose that it was imported.

The following artefact types were noted, with in parentheses: first the number of pieces found, and then the length, breadth and thickness of the largest and smallest entire samples, in mm.


Tanged arrow-heads (95 entire and 128 fragments, sizes 59/24/4-22/11/4) are the most characteristic artefacts. They are usually beautifully surface-chipped, but, sometimes thin chips (spåner) or blades have been used and merely shaped by edge-chipping.

The surface-chipped points, fig. 2, are the most numerous (175) and are nearly all chipped on both sides. The cross-section of the head is lens-shaped; the tang relatively long and stout, often ending in a point and lens-shaped or rhomboid in section. The barbs vary considerably in relative size. Normally they measure a quarter of the blade edge, but are in some cases reduced to mere excrescences, though these may be unfinished pieces. The angle of the extreme tip varies also, both slender and broad points occurring, the latter being the most common. Edges can be quite straight, but are most often slightly convex. They are usually finely serrated and one specimen has large notches along the edge, in common with several of the African tanged arrow-heads.

The edge-chipped points, fig. 3, (46), resemble the surface-chipped points in general dimensions but are of much poorer quality. The barbs are asymmetrical outgrowths and the tang is often reduced to a small tooth. As only edge-chipping is involved, usually from both sides but sometimes only from one, the arrow follows the shape of the original chip, giving rise to many irregular shapes. These edge-chipped points are usually thinner (2-3 mm.) than those with surface-chipping (c. 5 mm.). Sometimes the edge is serrated. One specimen clearly stems from a regular blade, being the only one with a sharp ridge from tip to tang.

Two tanged points are quite atypical, lacking any suggestion of barbs. They are fashioned from irregular chips and only retouched at the tang and tip.

Unfinished tanged arrow-heads, fig. 4 (6 entire, sizes 49/32/5--22/17/3). These represent the intermediate stage between the blanks as in fig. 6 and finished tanged-points fig. 2. They resemble the former but have been taken a stage further by beginning a tang. The procedure was as follows. First by careful surface-chipping, a beautiful, regular lamina with sharp lateral edges and a point at one end was produced. Then from the opposite, thick, rounded end, the blade was hollowed out by chipping obliquely towards the axis. When this had been done from both edges, a ridge was left standing in the middle, the tang.

Leaf-shaped arrow-heads, fig. 5, (3 entire, 1 fragment, sizes 76/25/6-45/1/6). Elegant, slender weapons, beautifully surface-chipped on both sides. One has an even, convex edge and round, thickened base and its greatest width is just below the middle. Two others have a pointed base, their greatest width near the base and the upper, longest part of the edge slightly S-turned. The fourth specimen is best regarded as an intermediate form.

Sturdy points, (48 fragments). Unfortunately no entire specimens occurred, but at least some of them may have been used as spear-heads and there is a possibility that some are fragments of sharp lateral knife-blades etc. All are surface-chipped on both sides.

Blanks, early stage, (108 entire, 78 fragments, sizes 70/41/25-29/21/4). These are thick, roughly chipped, oblong pieces, usually with a point. It is not always possible to determine what the final product would have been, but in the majority of cases the size indicates arrow-heads.

Blanks, late stage, fig. 6, (70 entire, sizes 66/28/7-28/21/6). No fragments are found in this category, as they cannot be placed here with certainty. It comprises pieces much thinner than the preceding group, the vast majority covered with surface-chipping. It is not possible to distinguish between semi-fabricata to leaf-shaped arrow-heads and to tanged points at this stage, as the latter can be as slender as the former. Several are beautifully fashioned with special trimming at the point and there is nothing to prevent their having been used as arrow-heads in that condition, but when one recalls the delicate execution of the "unfinished tanged arrow-heads" there is no pressing reason to regard them as finished weapons.

Fragments of small points (255). A large collection of various finished and half-finished broken small points. Further differentiation impossible.


Lateral knife-blades, fig. 7, (11, sizes 75/43/8-34/22/6). These pieces vary considerably in size and are not unlike certain spear-head forms, but have presumably been used as lateral knifeblades. The basic shape is triangular with the long edge fashioned for cutting by retouche on both sides, while the whole implement is covered in surface-chipping. In two cases there is double retouche all the way round, but the longest edge has there received special treatment so the knife-function is a reasonable assumption. Similar types are known among the Eskimos 2) and according to verbal information received from Jørgen Melgaard, of the National Museum of Denmark's Ethnography Dept., this triangular form gives just the right distribution of pressure when hafted as in fig. 8.

Diverse surface-chipped knives (9, sizes 70/34/9-37/16/4). This category includes pieces of rather variable shape. Three have a tang and double edge, the remainder are merely surface-chipped with the edge prepared as a cutting-edge.


Discoid scrapers (12, sizes 67/48/18-35/30/9). This is a somewhat variable type, several having merely a scanty retouche along part of the edge, whereas four are good, regular pieces of medium size with steep edges.

Hollow scrapers, fig. 9, (27, sizes 80/28/8-33/20/4) are here fashioned from blades or oblong chips which by retouche on one side are equipped with a steep, concave scraper edge. Usually these sturdy flints are well suited for work in wood or bone. In four instances, two edges have been fashioned on one scraper.

Flakes, worked on both sides, fig. 10, (164, sizes 58/52/20-25/23/7). All are coarsely chipped on both sides, with an uneven edge all the way round, and some still have remains of the original flint crust. They are sometimes plano-convex in cross-section but there is no hard and fast rule. It cannot be determined whether these are finished or unfinished tools, but Inspector Meldgaard has kindly supplied the information that this type, mounted as a scraper, is used among the Eskimos. It is held in both hands and in Meldgaard's experience is quite adequate for skin dressing, but when wood or bone is involved, a special scraper edge is required. Several of our specimens could have served as skin-scrapers.

Planes, fig. 11, (7, sizes 81/46/14-52/41/8). These resemble adzes, but not entirely. All are roughly the same size, broadest at the working edge and narrowing evenly towards the other, rounded, or in one case pointed, end. The edge is convex and very steep and in several cases appears to have been re-sharpened by chipping on one side only. Meldgaard knows this tool from former 3) and existing Eskimo cultures. At the present day it is made of iron and in wood working is used with a combined chopping and shaving movement. Fig. 12 shows the mode of hafting, and fig. 13 an alternative method from Alaska.

Flake scrapers, (26, sizes 110/63/24---23/23/6), i. e. chips with a scraper edge.

Other artefacts

Borers, fig. 14, (7, sizes 63/29/11-29/19/5). A rather irregular group. The majority are sturdy and all except one surface-chipped. Three show clear propeller retouche.

Saws, fig. 15, (3, sizes 38/23/6-26/12/2). Irregular chips with fine serration. Two are double edged, while the third has saw retouche down one side only.

Axes, fig. 16, (2, sizes 65/40/9-62/40/19). A crudely worked flint, fig. 16 a, is a symmetrical, double-sided core axe, reminding one of European mesolithic. The other, thin, piece, fig. 16 b, is completely covered with surface-chipping. lts greatest breadth is at the edge and the convex sides have probably tapered to a point, now broken off. The edge has fine retouche on one side, whereas on the other side, the middle three-quarters of the edge have been struck off in one blow. It should be borne in mind that as the extreme end is probably broken off, there is the possibility that this is a damaged "sturdy point".

Thin, surface-chipped specimens, (39, sizes 50/32/9-32/20/5). This group is hardly uniform as regards function. It consists of thin, oblong or triangular pieces, surface-chipped on both sides, without a point, which can have been used as knife-blades as well as scrapers.

Unidentified flint objects, (404). This category includes atypical and fragmented implements, of which some are in such a condition that they cannot be placed and others are so characterless and rare that a description is hardly warranted.

Flint swarf. About 10,000 pieces of swarf and irregular cores were collected on the site, which was, however, by no means emptied. Nearly all of them were chips and the relatively few blades were small and poor.

Hammer-stones, (22, sizes 100/76/44---45/34/25). Dispersed over the site were stones with clear abrasion-marks. Of the 22 carried away, 7 were of fairly good flint and the remaining 15 of reddish impure flint or softer rock. There had been a clear attempt in 16 cases to achieve discoid shape.

Spindle-whorl, fig. 17, (1 fragment, 66/36/9), apparently made from local sandstone. It is strongly wind eroded, flat on one side and slightly raised on the other, with a cylindrical perforation.

Pearl, (1, size 29/7). This is elongated with a perforation at each end, fig. 18. The type is known from Karim Shahir (Kurdistan) where it is dated to 10,000-8,000 B. C. 4). An examination by Ulrik Møhl and Jørgen Knudsen, of the Zoological Museum, Copenhagen, revealed that our specimen was made from the columella of one of the larger sea-snails, perhaps murex, a mollusc one finds today on the coasts of Qatar.

Sherds (32). These red, greenish or yellow undecorated sherds contribute nothing to dating, several being similar to the pottery in use today.

The sparse finds from the Arabian Peninsula contribute little to interpretation of the find at Dukhan. The profusion of arrow-heads, scrapers and knives indicate hunting activity with bow and arrow as principal weapon, although the spear could occasionally be used. When the quarry had been caught, a set of different knives was available for cutting it up, and scrapers for cleaning hide and bone. No definite burins for working bone were found.

In Rub-al-Khali, the south-western Arabian Desert, the mineralised core of a gazelle horn was found at a site with 19 tanged points 5) similar to those at Dukhan which can perhaps also have been used in gazelle hunting.

The proximity of the site to the sea makes fishing a not unreasonable supposition and there is fishing there at the present day.

There were absolutely no indications of agriculture, neither in the form of querns, sickle­blades nor silica polishing on flint. This might have been expected, since silica-polished sickle-blades have been found on Bahrain at two similar, though much smaller sites 6). The spindle-whorl may indicate sheep rearing or flax cultivation, although a later adulteration cannot be excluded.

Unfortunately, no organic materials were preserved, but containers of wood and leather, as is practicable for hunters and nomads, have probably been used, as the amount of pottery found does not correspond to the size of the find, even disregarding the fact that some of it can be dated to our own period.

Two fire-shattered flints were strangely enough the only signs of fire, although we had been on the look-out for these. This points to a rather short habitation.

All things considered one gets the impression that the Dukhan people have been nomadic hunters with perhaps some sheep rearing or flax cultivation, but one must not overlook the possibility that cultivators from a nearby, unknown culture group settled here for a short time to hunt.

Any attempt to place the find culturally or chronologically is hazardous, as the embedding of the flints in sandstone gives no basis for dating, as established by the head of Q. P. C.'s geological laboratory at Dukhan, D. M. Morton, who kindly examined a sample.

The most typical implements of the site, the surface-chipped tanged arrow-head and the leaf-shaped arrow-head, are found in numerous different cultures without any definite chronological horizon. They are known in the east from Susa 7) and Woolley mentions them at Ur in definitely early strata and in great quantity in layers from the 3rd dynasty and even later 8). In Egypt they occur in neolithic and pre-dynasty times, as well as in the Old Kingdom and perhaps later 9). The Fayum find, dated by C14 to 4,200 ± 250 B.C. and 4,400± 180 B.C. 10), shows similarities with Dukhan 11), but it is clear that Fayum contained an agricultural element with sickle-blades and polished axes. Finally it should be remarked that there is also a difference in the type of arrow-head, a point where farming and hunting cultures can be compared, as the heart-shaped arrow, which is so characteristic of the early phase at Fayum, is entirely absent from the Arabian Peninsula. If an Egyptian influence is involved, it must start later, when Fayum was producing tanged points.

By and large the discovery accords with the extensive complex of North African neolithic cultures, which cover the Magreb, Sahara, Egypt and perhaps as far as Somaliland and Abyssinia and which, moreover, leaves its mark in the Spanish neolithic. The lack of parallels in the east can be due to insufficient information on the Stone Age cultures of the great West-Asiatic regions.

Hans Jørgen Madsen.





Madsen, H. J. (1961). En Flintplads i Qatar. Kuml, 11(11), 185–201. Hentet fra https://tidsskrift.dk/kuml/article/view/103351