Lawrence of Arabia’s War

4 mins read

The Great Arab Revolt Project

With the CWA-backed Great Arab Revolt Project at an end after ten years’ work on the deserts of southern Jordan, we asked Co-director Neil Faulkner – also Editor of our sister magazine Military History Monthly – for some concluding thoughts

Bristol University’s Nick Saunders, in characteristic thinking pose, surveys the terrain on site in Jordan.
Bristol University’s Nick Saunders, in characteristic thinking pose, surveys the terrain on site in Jordan.


In 2006, here in the pages of CWA, we launched our Great Arab Revolt Project (GARP). With the project newly completed, the results have been game-changing – despite the fact that, in many ways, we have been pursuing the archaeology of the invisible.

For many years I accepted Christopher Hawkes’ famous ‘ladder of inference’ without question. The basic idea, if you recall, is that archaeological evidence allows much to be said about technology and technique, less about economic life more generally, still less about social organisation, and not much at all about beliefs, values, and ideas. The ladder has attracted growing criticism. Many archaeologists now reject it outright. I am one of them. Not the least reason is my experience of nine seasons in the desert.

A loophole in the breastwork wall of a Late Ottoman redoubt on the Batn Al-Ghoul escarpment. The redoubt overlooks the Hijaz Railway. Every sector of the line was under observation from an Ottoman post in 1917-1918.
A loophole in the breastwork wall of a Late Ottoman redoubt on the Batn Al-Ghoul escarpment. The redoubt overlooks the Hijaz Railway. Every sector of the line was under observation from an Ottoman post in 1917-1918.

GARP – a Bristol University project that I co-directed with Nick Saunders and David Thorpe – was run from the outset as a multi-disciplinary investigation of a multi-dimensional conflict. We put a team of about 30 archaeologists and volunteers into the field for two weeks each autumn, working mainly along a 120km stretch of the former Hijaz Railway between Maan and Mudawwara (the latter now on the Jordanian–Saudi border).

The archaeology comprised fortified railway stations, hilltop redoubts, blockhouses, campsites, and scatters of expended munitions, discarded military equipment, and the detritus of everyday army life: evidence of the Ottoman counterinsurgency effort between 1916 and 1918. Occasionally, too, we caught a glimpse of the shadowy enemy, the Bedouin guerrillas and British demolition experts, weapon specialists, and liaison officers who supported them.

Holistic approach

Using archive sources, satellite imagery, and ground reconnaissance – mainly the work of GARP landscape archaeologist John Winterburn – we plotted the sites spread across the desert wilderness. Some we then investigated in detail, clearing wind-blown sand from breastworks and tent-rings, digging out trenches and machine-gun posts, drawing and photographing loopholed blockhouses. Thus, bit by bit, we built a picture of ‘Lawrence of Arabia’s war’.

We never made a virtue of separating observation from interpretation. We debated the meaning of what we were seeing as we went along, and working hypotheses became the basis for new lines of investigation. Archaeology is not a linear process: it is a holistic one involving a feedback loop where material (the imprints in the landscape), method (the way we sample and recover data), and meaning (the interpretation) – I call them ‘the three Ms’ – constantly interact. Or they should do.

The more I think about it, the more convinced I am that Hawkes’ ladder of inference does not begin to describe what we did. I want to illustrate this by offering some ‘big picture’ conclusions. Let me start by picking up my somewhat tongue-in-cheek opening reference to ‘the archaeology of the invisible’. What I have in mind is the razor-sharp contrast between the archaeological imprint of the Ottoman occupation and that of the Bedouin insurgency.

Asymmetrical evidence

The Ottoman Army was a highly visible presence in the landscape, and that is reflected in the material evidence that survives – the blockhouses and redoubts, the trenches and breastworks, the campsites, and, of course, the railway itself. The occupation forces were essentially static – they lacked the resources to mount an active counter-insurgency – remaining pinned to the 1,300km-long Hijaz Railway, the all-important supply line running from Damascus to Medina: thus the strong archaeological traces.

Tooth Hill Camp was eventually located mainly because of the curious shape of the hill that gave the place its (wartime) name. Fieldwalking aimed at total recovery of discarded artefacts, yielding a rich assemblage.
Tooth Hill Camp was eventually located mainly because of the curious shape of the hill that gave the place its (wartime) name. Fieldwalking aimed at total recovery of discarded artefacts, yielding a rich assemblage.

Lawrence contrasted this with the nature of the Arab forces: ‘Armies were like plants, immobile, firm-rooted, nourished through long stems to the head. We might be a vapour, blowing where we listed.’ And what does ‘a vapour’ look like archaeologically? Not much. So little, in fact, that for several years we found no trace of the Arab forces at all, save an occasional .303 bullet near an Ottoman trench, representing incoming fire. And then we stumbled on two exceptional sites that proved the rule – that an asymmetrical war had generated a visible/invisible contrast in the archaeological record.

The first was Wuheida, the site of the main base of Prince Feisal’s Arab Northern Army for several months during 1918. Along about two kilometres of wadi, we discovered a dozen or so separate complexes of stone alignments, representing enclosures, paths, and tents. The identity of the occupants was confirmed by a sizeable assemblage of diagnostic metalwork.

We interpreted this as the distinctive imprint of an Arab tribal army, grouped in separate contingents, with liminal space between each, reflecting the in-group/out-group anthropology of the desert, with its kin-based allegiances and bloodfeud rivalries. Only later was it possible to confirm this, when a contemporary photograph turned up depicting exactly what we had surmised.

The second ‘exceptional’ site was Tooth Hill Camp, where British special forces, sometimes in company with Lawrence, had camped on a number of occasions prior to attacks on the railway. In this case, the site was defined by campfires, each a little cluster of rocks around a fire-pit, associated with general spreads of discarded ration-tins, broken rum jars and bottles, spent cartridges, and occasional vehicle parts.

Around the former campfires were broken-up biscuit boxes and ration tins.
Around the former campfires were broken-up biscuit boxes and ration tins.

Invisible archaeology

With Tooth Hill Camp, here was another kind of ‘archaeology of the invisible’, this time representing rapid desert movement and only the most transient occupation, mere overnight stops. Tooth Hill reminded me of another class of evidence that we had recovered accidentally in the early days of the project.

Working near a ruined railway station at Wadi Rutm, the metal-detectorists had discovered an extensive spread of medieval and early modern coins and trinkets. A few kilometres to the north, at Batn Al-Ghoul, where the railway passes through a shallow gorge, they picked up a line of such material. Could there really be any doubt that we had ‘seen’ the camel caravans – both in movement and in camp – that had preceded the railway, following the same route, giving us a deep-time perspective on trade and pilgrimage along this desert routeway?

The railway, built between 1900 and 1908, was the direct replacement of the camel caravans. We may have known this from historical records, but we could equally well have divined it from the archaeology alone.

Images: Ali Baldry/Nick Saunders/John Winterburn/Great Arab Revolt Project

This is an extract from a feature published in CWA 78. Read on in the magazine or click here to subscribe.

Leave a Reply

Your email address will not be published.