The Desiccation Theory Revisited

Leone Caetani in Egypt (1888)

In the beginnings of the 20th century new findings in the field of natural sciences made it possible for historians and other scholars to try and implement these findings into their respective areas of expertise. Thus two versed orientalists in the most classical sense of the term Hugo Winckler and Leone Caetani, devised independently from each other a theory stating that the main motive force behind the course of historical events in Arabia and the Middle East region as a whole was continuous desiccation and aridification. For them, Arabia was the cradle and primordial home of Semitic peopleswhich was fertile at the dawn of history, but through the millenia became more and more barren, with this steady process driving her children, the Assyrians, Arameans, Canaanites and Arabs away from the Arabian Peninsula in respective waves of migration. To Caetani, all societal and economic changes accompanying the rise and expansion of Islam were mere manifestations of greater shifts in the natural world.

An obvious question suggests itself at this point. Of what benefit would it be to put one’s mind to this obsolete theory which was shortly after its appearance subjected to severe critique concerning its logical inconsistency? H. Lammens, a sharp-witted Belgian scholar, summed up some of the arguments in his publication on Hejaz before the rise of Islam (Le Berceau…). He pointed out how hard it is to imagine that despite its millenia long agony, the dying land was still able to produce one migration wave after another(124). Lammens also remarked (134) that the element of the camel was so tightly interwoven into the pre-Islamic society that it had to have been introduced and domesticated a long time before first textual evidence (« textual » doesn’t stand here for « literary » or « written »), thus reflecting stable dry climate. At the time, Lammens couldn’t have had any factual evidence for this statement but his his brilliance foreshadowed what we now know which is that the domestication of the camel started at the turn of 4th and 3rd millenia B.C. and completed between the 12th and 14th centuries B.C. (Hoyland 90). Whereas the Winckler-Caetani theory, as it is sometimes named, places the beginning of the desiccation period somewhere between 10000 and 6000 B.C, i.e. at the end of the glacial period, according to current knowledge, but accounts even as recent as some from the time of Tiglatpilesar I (1114-1076 B.C.) reveal that the area of Magan (identified roughly with present-day Oman) was able to sustain animals as large as rhinoceri.

To answer the aforementioned question one can say that even if the theory was hopelessly wrong, its main idea, i.e. the interaction of natural and historical processes in Arabia, deserves to be revived in light of the great progress during the last two decades in the field of paleoclimatology. There is a handful of recent studies more or less concerning the Arabian Peninsula and I think that by not going through it, we would deprive ourselves of quite a colourful and vivid picture of the climate change in Arabia’s history before the emergence of Islam.

Historical climate fluctuations as portrayed by today’s science

Most of the sources available agree that there was a significantly humid period in Arabia before the LGM (Last Glacial Maximum), dated at approximately 18000 B.C.. which can be confirmed by wadi terraces found in Wādī aḏ-Ḏajd, Liwāʾ (both UAE) and Mundafan (Saudi Arabia) areas, together with other findings of alluvial fans and sediments (Parker, Goudie 462-3).

Conversely, the LGM brought extreme aridity because, simply put, the more water is bound in ice, the less of it is available for vegetation in the warmer areas of our planet. Arabia witnessed these severe conditions, characterised by enormous formations of linear dunes in Rubʿ al-Ḫālī and significantly lower sea levels, throughout the LGM until it loosened its grasp by 12000 B.C. The coastline was 120 meters under its current level and the Arabian Sea reached no farther than the Strait of Hormuz. Thus, it is probable bordering on certainty that the Tigris-Euphrates system would have been prolonged equivalently (Parker, Goudie 464).

With fading influence of the LGM came re-flooding of the Gulf (begun in 10500 B.C. and finished 6 000 years ago) and general improvement of climatic conditions. Many different and unrelated sources reflect the return of a more humid climate roughly between 9000 and 3000 B.C. , although quite substantial regional fluctuations can be found.

The fading of the LGM’s influence itself, however, would not have been sufficient for a climate supportive of life, so there was another element involved. The ITCZ (Intertropical Convergence Zone) plays a crucial role in the formation of monsoon system in the Indian Ocean, thereby greatly influencing the distribution of precipitation. It can move northward and southward and it did so in the concerned period, shifting the monsoon radius north of its position in both the LGM and our times. Chronological delimitation of the ITCZ’s journey to the north and then back could be somewhat tricky because its movement is relatively slow and it took 1,800 years for it to reach the northernmost point of the shift. In the case of this movement’s beginning we can refer to the renewed formation of alluvial sediments in Fallāğ al-Muʿallā (UAE) between 9100 and 8500 B.C. (Parker, Goudie 463), the rise of sea-floor sediments in the Hormuz area in 9500 B.C. (Sirocko, et al. 480) and stalactite analysis done in Huta cavern in Oman (Fleitmann, D. 1738).

As far as the duration and exact ending of these humid conditions is concerned, we cannot be as precise. Two possible borderlines can be found – in the 2nd millennium B.C. and the two centuries that overlapped at the turn of the era. Both have been recorded in the sediments on the Pakistani coast, whereas the decline of rainfall begins with 2nd milennium B.C. and reaches its minimum during the so called Roman Warm in the 1st century B.C. / A.D. (Rad 51).

Many other sources provide us with analogous evidence, as a huge influx of eolian sand and dust deposits dated between 2025-1620 B.C. has been identified in the Strait of Hormuz area, and also at Hili 8 locality in al-ʿAjn (UAE), where wells were considerably deepened between 2500 B.C. and 1800 B.C. and the settlement ceased to exist completely around 1500 B.C. (Jorgensen, Takriti 41-42). A line of connection can also be drawn to the collapse of the Akkadian empire in 22nd century B.C. as the Tell Leilan locality in today’s Syria was covered with a layer of sand over one meter thick (Sirocko 379).

We can suppose that the humid period was shorter in northern and inland areas of Arabia. Water amassed in the natural underground reservoir of al-Liwāʾ was determined to be over 5,700 years old and it has not been replenished ever since. The youngest Rubʿ al-Ḫālī lake floors are dated to 4000 B.C. (Jorgensen, Takriti 46). Drying-up of al-ʾAwāfī area lakes (UAE) occurred between 4000 and 2000 B.C. and the an-Nafūd lakes situated farther north were dry even 2,000 years before that (Parker, Goudie 464,467).

Every piece of evidence we have gathered here speaks in favor of a stable dry climate in Arabia since the 4th millennium B.C. It also corresponds with the aforementioned domestication of camels and other sources tell us (Tosi 475,477) that the culture system of Arabian oases was fully developed by 2000 B.C.

That this was not an isolated local process and that it had something to do with greater climate changes can be ascertained by data from other parts of the ITCZ’s region of influence. Thus there is an apparent humid period which ended in the 2nd milennium B.C. recorded in the Abhe lake in Ethiopia, and a similar period ending 2,000 years sooner as evidenced by the inland position of Tibet‘s Lake Sumxi. The same applies to the Thar Desert in India (Gasse, Van Campo 441,445).

Climate change and historical events

The collapse of the Akkadian Empire can be clearly connected to the end of the humid period. Not only by the sand covering of Tell Leilan locality but also by identification of the sand and dust particles brought to the Strait of Hormuz by the Šamāl wind (Sirocko 380) in the concerned period. It is hard to imagine the full extent of the phenomenon we are speaking about because in only a single year the Šamāl wind is capable of carrying more than 100,000,000 tons of sand and dust material.

Another striking coincidence can be found in the temporal congruency of the Roman Warm (more of a Roman Drought in the case of Arabia) and the northward migration of the South Arabian tribes – notwithstanding the fact that this could have been partly caused by the deflection of circumlittoral navigation from Ptolemaic Egypt to India which had used South Arabian ports, and thus provided trade opportunities and port tolls, until direct navigation was discovered, probably by Eudoxus of Cyzicus around – 120-110 B.C. (Hourani).

The question of the South Arabian tribes migration is more complex and was described here rather simply. But I think that a more detailed and elaborate study of the relationship between climate change and historical events could prove both fruitful and interesting. So far only one thing can be stated with certainty: so far, no evidence has been found that would support any sort of climatic origins to the rise of Islam and itsexpansion.

Premysl Kubat
Institute of Near Eastern and African Studies, Charles University in Prague
Doctorant associé à l’Ifpo


Glennie, K.W. a A.K. Singhvi. „Event stratigraphy, paleoenvironment and chronology of SE Arabian deserts.“ Quaternary Science Reviews 21 (2002): 853-869.

Hourani, G.F. Arab Seafaring. Princeton: PUP, 1995.

Hoyland, R. G. Arabia and the Arabs. London: Routledge, 2001. 90.

Jorgensen, D.G. a W.Y. Takriti. „A hydrologic and archeologic study of climate change in Al Ain, United Arab Emirates.“ Global and Planetary Change 35 (2002): 37-49.

Lammens, H. Le Berceau de l’Islam. Romae: Sumptibus Pontificii Instituti Biblici, 1914.

Parker, A.G. a A. Goudie. „Geomorphological and Paleoenvironmental Investigations in The Southeastern Arabian Gulf Region.“Geomorphology 101 (2008): 458-470.

Rad, U. von, et al. „A 5000-yr Record of Climate Change in Varved Sediments from theOxygen Minimum Zone off Pakistan, Northeastern Arabian Sea.“ Quaternary Research 51.1 (1999): 39-53.

Sirocko, F., et al. „Climate change and the collapse of the Akkadian empire: Evidence from the deep sea.“ Geology 28.4 (2000): 379-382.

Tosi, M. „The Emerging Picture of Prehistoric Arabia.“ Annual Review of Anthropology 15 (1986): 461-490

(photo : NASA/GSFC/METI/ERSDAC/JAROS, and U.S./Japan ASTER Science Team)