My research on Arabian Prehistory

When I began my PhD on the Middle Palaeolithic of Arabia in 2009, not a single Pleistocene site in Arabia had been chronometrically dated. Since then, much has been learned about the prehistory of the region, and I have contributed to this for multiple time periods and regions, working with Michael Petraglia and other colleagues (link, link). Debate has focussed on the dispersal of Homo sapiens out of Africa and into Arabia, but I am also interested in behavioural changes within the peninsula. Aside from its geographical location at the interface of Africa and Eurasia, Arabia makes a remarkable natural laboratory. It seems that populations were periodically able to enter the region (probably from the north, from the Sinai/Negev area) during periods of improved climate, but they then contracted to the south as climate deteriorated. Cut off and isolated, these populations developed all kinds of distinctive material culture aspects, and provide fascinating case studies on human-environment interactions.

While there have been claims for very early (Oldowan) occupations in Arabia, the evidence is so far rather unclear. Much clearer is evidence from the Acheulean period. This is characterised by the production of large cutting tools, particularly handaxes. We have found many of these in Arabia (link, link, link, link). These sites tend to be associated with either outcrops of particular stone, and therefore represent workshop-like sites, or are associated with the remains of ancient lakes.

The Acheulean sites of Arabia provide important behavioural information, as they show that these hominins were living in relatively open and arid environments. Crucial to our work in Arabia has been the application of chronometric dating methods, particularly optically stimulated luminescence  (OSL) dating to produce scientific estimates for the age of sites. This has produced some interesting results. At Khall Amayshan, we found what is currently the oldest dated human occupation of the area, at around 400 ka (link). At the same site, a subsequent phase of lake formation and renewed human presence produced a large assemblage of beautiful small handaxes, dating to just over 300 thousand years ago. Dating sites of this age is tricky, and we also have evidence from butchery marks and a few lithics from the site of Ti’s al Ghadah which seem to be a broadly similar age to these sites and is best known for its animal fossils (link, link, link).

Another very interesting set of Acheulean sites that I have had the pleasure to investigate are found near Dawadmi in central Arabia. This is a fascinating area with a very flat and rather barren landscape, periodically dissected by ridge-like ‘dykes’ of hard rocks such as andesite and rhyolite which were used by prehistoric hominin to make stone tools. The area is also geographically interesting as it sits at the interface of some major ancient river systems. While some areas of Arabia were used by hominins at repeated points of prehistory, this area seems to have only faint traces of post-Acheulean prehistory. It therefore represents a fascinating ‘Acheulean landscape’. Our surveys of the area found dense lithic assemblages near the raw material sources (link, link). We returned to an excavation which had been conducted by Norman Whalen and colleagues several decades ago, and for which no absolute dates were available. With further excavations and the application of luminescence dating, we were able to date this particular site to around 200 thousand years ago, and even younger in the upper layers (link). This came as a surprise, as the handaxes are rather large and crude, so we had guessed that they were much older. At this age, the Lower Palaeolithic in central Arabia overlaps with the Middle Palaeolithic in northern Arabia, and hints at complex behavioural and demographic dynamics.

As well as stone tools, animal fossils are found at some of these Lower Palaeolithic sites, as well as in association with younger sites discussed below. I have helped in the excavation and analysis of some of these animal fossils. Particular highlight have included excavating elephant tusks and other fossils at Ti’s al Ghadah, and hippopotamus fossils at multiple sites. At Al Wusta, for instance, we were able to directly date a hippo tooth to around 90 thousand years ago (link). In this palaeontology research I have worked closely with several colleagues including Dr Mathew Stewart who was my postdoctoral researcher. These animal fossils give us a window into the environments of ancient Arabia, with species such as hippos indicating good water access in terms of small lakes. In biogeographical terms, there are also indications that Arabia shows closer faunal connections with Africa than the Levantine woodland area does (link), again providing important context for human dispersals.

A major area of research for me has been the Middle Palaeolithic of Arabia. This is again associated with raw material sources and palaeolakes, and over the past few years I have been involved in the discovery of many Middle Palaeolithic sites, their excavation, and dating. I have worked in areas from Dhofar in southern Arabia (ref), Mundafan in the western part of the ‘Empty Quarter’ desert (link), the Nejd of central Arabia (link), and particularly northern Arabia (link, link, link). The Nefud Desert has proved to be a particularly rich area for research, both because of a rich Middle Palaeolithic record and because the frequent presence of lake sediments allows sites to be chronometrically dated. Mundafan is the first dated Pleistocene site in the Empty Quarter (Rub’ al Khali) which is the largest sand desert in the world. Our work there showed that humans were there around 80-100 thousand years ago, and they were making similar stone tools to Homo sapiens in the Levant and northeast Africa around that time.

In northern Arabia, we have excavated multiple sites in the Jubbah basin. These mostly date to MIS 5 (125-75 ka), but after a lot of digging we also found a small assemblage to around 200 ka (link). Through work in areas like Jubbah we have also explored the role of raw material sources and variation in how hominins made and used their stone tools (link). Over the years we have begun to explore the wider region. At the site of ALM-3 we found a younger Middle Palaeolithic assemblage dating to around 55 thousand years ago (link). The technology of this is rather different to the older Middle Palaeolithic sites in the area, so could it represent a different population, or even species of hominin?

In the western Nefud, I was part of the team that has identified multiple Middle Palaeolithic sites. At KAM-4, for instance, we found three Middle Palaeolithic assemblages of different ages. At Al Wusta we found not only an interesting lithic assemblage, animal fossils including hippos, but something that we have spent a a lot of time looking for…a hominin fossil! In fact, this fossil – the middle finger bone (phalanx), probably of the middle finger – is the only pre-Holocene hominin fossil known from Arabia. I led a paper on this fossil and site, and fortunately it turned out that this particular bone is usefully diagnostic. The shape of the fossil indicated that it belonged to Homo sapiens, rather than another species such as Neanderthals (link). We were also able to directly date the fossil, and also date animal bones from the site, which all indicate that it dates to around 85-90 thousand years ago. This finding showed that the ‘early’ (i.e. MIS 5) migrations of our species out of Africa were not merely limited to the Levantine area, but extended deep into Arabia (and with recent findings in eastern Asia, maybe much further!).

How did these early humans get to Arabia? To explore this question I have worked with various colleagues. Aside from stone tool comparative studies, a key part of this involves exploring the spatio-temporal distribution of archaeological sites and how this correlates with environmental conditions. For instance, we demonstrated how fluvial corridors linking the southern Levant and northern Arabia existed at key points of the past (link). Working with Sam Nicholson and colleagues, we explored the relationship between sea level and regional climate in terms of the hypothesised Red Sea crossing in the Bab al Mandeb area. We found that this route was unlikely – as it already seemed to me based on the archaeological record of Arabia – because local climatic amelioration lagged behind sea level rise (link).

I have also been involved in research on the late prehistory of Arabia. As much as I love stone tools – and it has been a pleasure to work on Neolithic stone tool assemblages from both southern and northern Arabia (e.g. link), it is also enjoyable to be able to work with more diverse kinds of evidence for this more recent period. It is also nice to work in a period when radiocarbon dating can be used, allowing clearer correlations between humans and climate than for earlier periods (e.g. link). Working with colleagues such as Maria Guagnin I have been involved in studies of rock art in northern Arabia, where ancient humans engraved sandstone cliffs and boulders and left a fascinating cultural legacy (link, link, link). Our work highlighted the significance of areas such as Shuwaymis with its spectacular rock art, and they subsequently became UNESCO sites.

The Holocene also sees a proliferation of various stone structures, a topic that I have become increasingly fascinated by. While we often think of Arabia as a land of sand dunes, much of the surface of the area actually consists of rock. From sandstone ‘jebels’ emerging from between sand dunes, to the remains of ancient lava flows which often weather into vast masses of blocks of rock of a very convenient size for building, there is a lot of rock around, which ancient societies took full use of. And this was perhaps catalysed by a relative paucity of organic materials such as wood, and likewise the character of erosion and deflation in the area means a highly visible record. The end result is that across Arabia, there are thousands and thousands of stone structures. They cover at least the last ten thousand years, and range from dwellings, to ritual structures, to giant hunting traps.

As with most else in Arabian archaeology, little specific was known about prehistoric stone structures in Arabia until the recent past. I have been involved in studies on stone structures which have helped improve our understanding. The first form refer to ‘mustatils’ (the word means rectangle in Arabic). While previous work had mentioned the existence of these large (upto 600 m long) structures, very little was known of them. My own interest came from the fact that to get to Palaeolithic sites such as Al Wusta in the Nefud Desert, we drove through an area containing these structures. We conducted the first on-the-ground surveys of them, and were able to obtain a radiocarbon date from one, showing that it was around 7000 years old (link). These structures were enigmatic; there was very little associated archaeology around them, and little initial clear reason on why Neolithic people in the area had built hundreds of these structures. It seemed that they were ritual structures built by pastoralist societies, in the context of the later part of the Holocene humid period. It has been great to see other teams in the region follow this foundation and a variety of excavations and surveys are currently ongoing. My PhD student, Amy Hatton, is currently conducting spatial analysis of mustatils and their landscape positioning.

The other kind of stone structure that I have worked on are known as ‘desert kites’. These are vast structures consisting of converging walls to an enclosed area. These are believed to be giant hunting traps, into which herds of animal like gazelle could be funnelled and killed. These kites are best known from areas like Jordan, but were also known from northern Arabia, Syria, etc. In fact, similar although somewhat simpler, structures are known from areas as diverse as the Nile Valley, South Africa, and Yemen. We therefore have another example of a situation where different hypotheses can be proposed to explain similarities in material culture, i.e. convergent evolution or some kind of dispersal or diffusion. Some researchers begin with very restrictive definitions of desert kites, based on the most complex forms, and say that other forms are merely ‘kite-like’. In general when it comes to comparing similarities in material culture, I prefer to start in broad terms and I think that we need to focus on quantitative comparisons rather than arguing about narrow definitions.

In a study with my postdoc Christ Carleton, we used remote sensing techniques  to explore areas of Arabia for kites (link). We found lots of new kites; particularly in the Harrat Nawasif area of central Arabia where they were not reported before. We also found multiple new kites in northern Arabia. What was particularly interesting for me is that our study allowed us begin to think through spatial variation in the desert kites. Kites we identified in northwest Arabia have a similar morphology to those previously known from the Sinai/Negev area. Those we identified in the Ha’il area were similar to those from eastern Jordan. The central Arabian kites provided a link between the complex kites of northern Arabia and the simpler forms in Yemen. This patterning did not look random to me, but it looked as though there was a gradient from north to south, in which the kites became morphologically simpler. It could also be that some forms are younger than others; the Negev/Sinai form, for instance, might be younger than the others.

I continue to work on the prehistory of Arabia, with its remarkable record of human societies across time. Not only interesting in their own right, these societies are key to understanding some of the major debates in prehistoric archaeology; how do we understand and find evidence for migrations? How did early humans react to environmental change? How do we distinguish material culture similarities representing dispersal, diffusion, and independent (convergent) evolution?