- In 1953, Werner Caskel produced a theory which he called “the Bedouinization of Arabia”. In this, he maintained that around AD 100 Arabia was peaceful, dominated by settled states, with some non-tribal nomads who were simply “shepherds near the cities”. He contrasted this with Arabia in the sixth and seventh centuries in which he claimed “the Bedouin form of society and ideo logy prevailed”. The evidence he provided for this false dichotomy con sisted of errors, misunderstandings and argumenta ex silentio, as was pointed out at the time, but his theory has nevertheless been widely accepted in the years which followed. In 1959, it was taken up and adapted by Walter Dostal who tried to explain the “Bedouinization” by producing a novel definition of the Bedouin as “camel-herders accustomed to fighting as rider warriors” and said that “Voll beduinen” were only those who used the shadād or so-called “North-Arabian” camel saddle which, he imagined, gave them a secure seat from which to fight. In fact, however, there is no evidence at all that nomads in Arabia have ever fought from camel-back if they could possibly get off to fight on foot or on horseback. Nevertheless, in 1975, Richard Bulliet adopted Dostal’s idea and took it further by claiming that the use of the shadād made camel-riders an almost invincible force and this produced “an alteration in the balance of political power in favour of the nomads”. This gave apparent support to Caskel’s idea that the North Arabian Bedouin were able to sweep to military and political domination of the sedentaries by the sixth century AD. The present examination of these theories shows that there is no basis to them and, in doing so, argues that what is known of nomadic life in Arabia between AD 100 and 500 suggests continuity both in its structures and in its relations with the sedentaries.
- Caskel and “the Bedouinization of Arabia”
Some sixty years ago, Werner Caskel produced a theory which he called “the Bedouinization of Arabia”. For a man who had devoted so much of his career to the study of pre-Islamic and early Islamic Arabia, it was based on a shock ingly superficial, and often incorrect, view of a very limited amount of evidence and on a great many argumenta ex silentio. This was pointed out at the time by Martin Noth, among others (Caskel 1953a, 25–27), but despite this, it was adopted unquestioningly, and developed enthusiastically, by later writers such as Dostal, Bulliet, Knauf, Högeman, etc., and is still with us. The most recent example of it I have noticed, dates from 2003 (Retsö 2003, 276–277, 582–583). For Caskel, “the social situation in northern Arabia” in the sixth/seventh centuries AD was one “in which the Bedouin form of society and ideology pre vailed” (Caskel 1954a, 36). For him, this form of society was the “tribe”, which he defined purely in terms of nomadic life, even though he considered it was found “also with the settled Arabs”. However, his theory is undermined by his apparent ignorance, and/or misunderstanding, of some of the basic facts of nomadic life in north ern and central Arabia.⁶ Thus, for instance, he says correctly that the Bedouin of Arabia spend the spring in the desert pastures, and the summer near permanent water, but then claims that they spend the autumn and winter in oases.⁷
- In fact, of course, the Bedouin of north Arabia move to the inner desert with the first rains in October, and spend the winter, and the season of lesser rains (known as as-smāk, February to April), and the early summer (ṣayf) there. They will only move to per manent water sources in the desert in the dry season of qayẓ – roughly between June and October – after which the cycle begins again. They do not normally live for any part of the year in the oases. This is a pattern which is also described in the Safaitic inscriptions carved by nomads in southern Syria, north-eastern Jordan, and northern Saudi Arabia, from the first century BC to the fourth century AD (Macdonald 1992a). Similarly, Caskel misinterprets other aspects of Bedouin life and society. For instance, he sees the Benī al-ʿamm – the only group in Bedouin society in which each member is totally responsible for each of the others – solely in terms of “blood vengeance”, whereas, of course, the responsibilities affect all aspects of life. He also claims that the responsibility inherent in the Benī al-ʿamm “cuts across their socio-political organization, the tribal system”,⁹ whereas in fact, it forms the basis both of the ideology and the practical applications of the Bedouin tribal system. At the heart of Caskel’s thesis are four assumptions: (a) that the term “Bedouin” should be restricted to the nomadic groups distinguished by “tribal organization and its ideological superstructure,”¹⁰ (b) that the “tribe” as a social organization is restricted to the Bedouin, and therefore to nomads or their settled relatives;¹¹ (c) that “tribal organization did not arise in Arabia proper” though he thinks it did exist among the authors of the Safaitic inscriptions;¹² and (d) that by the sixth century AD, all Arabian oases, except (for some unexplained reason) those in the north-west, were peopled by ex-nomadic Bedouin.
- Caskel’s “pre-Bedouinized” Arabia
- On this basis, he developed a theory in which he envisaged a “pre-Bedouinized” or “non-Bedouin” Arabia at around AD 100 (Caskel 1953b, 28, 30; 1954a 36, 38). This he characterized as predomi nantly sedentary, peaceful and trading; an Arabia of states linked by trade routes, in which the nomads were simply “shepherds near the cities” (Caskel 1953b, 30; 1954a: 39) with a very limited range of migration and little or no ability to threaten the settled population. North of Yemen, Arabia was never primarily a sedentary area, nor, until the twentieth century, could it be dominated by the sedentaries to any real extent. The nature of the region inevitably gives to nomads an independence of exter nal authority which can never be achieved by sedentaries, hence, of course, the periodic attempts of the sedentary governments to settle the nomads (Donner 1981, 251–267). On Muhammad’s attempts to settle the Bedouin see ibid., 79–82; on government policies on settlement during the conquests see 265–267. In these circumstances we have to be careful to define what we mean by such terms as “kingdom”, “principality”, “city-state” when used of an area where seden tary occupation is limited to oases separated from each other by large areas of desert which are home to nomadic groups who do not need to recognize external authority. Moreover, Caskel’s characterization of the nomads in around AD 100 as non-tribal, and therefore “non-Bedouin” is based on a misunderstanding of the epigraphic evidence, particularly of what he calls the “thamûdenic” – i.e. “Tha mudic” – inscriptions.
“Thamudic” is a misnomer invented by modern scholars for texts in a number of different scripts which, for various reasons, have yet to be classified. The name was borrowed from that of the ancient tribe of Thamūd but does not imply that the members of this tribe necessarily used one of the scripts. “Thamudic” is a sort of “pending file” and it is nonsensical to refer to “the Thamudic script”, since, by definition, there are several different scripts lumped to together under this rubric (see Macdonald 2009a, text III, 33–35). Caskel, however, followed van den Branden’s discredited view (1950, 17–23) that the various alphabets lumped together by modern scholars under the “hold-all” label “Thamudic” were in fact a single script at various stages of development, in which “the oldest [texts] … come from Taimâ, the younger ones from Tebûk” (Caskel 1954a: 39). Winnett had already demonstrated in 1937 that at least five different scripts could be roughly identified in the “Thamudic” pending file and showed that there are marked differences in the sorts of inscription found in each of them, as well as considerable difficul ties in arranging them chronologically (Winnett 1937, 53–54). By failing to distinguish between the various scripts lumped together as “Thamudic”, Caskel’s argument becomes confused. For instance, one of the scripts which at that time was included under this rubric (Winnett’s “Tha mudic A”) has now been recognized as the script used by the settled inhabitants of the oasis of Taymāʾ and yet Caskel lumps it with the other “Thamudic” scripts and assumes it was used by nomads (1954a, 39).
-
- Tribal affiliations: Contrary to Caskel’s claim, “Thamudic” inscriptions do contain tribal and clan names, as shown by Al-Rūsān’s work.
- Misinterpretation of evidence: Caskel draws conclusions about social structure from limited information in the inscriptions, which is not justified given the nature of the texts.
- Assumptions about social groups: Caskel assumes that terms like ʾl in Safaitic inscriptions mean “tribe” in the same sense as later Bedouin societies, which is not necessarily accurate.
- Ex silentio arguments: Caskel makes unsupported claims about the “peaceful milieu” of Thamudic society based on what the inscriptions don’t mention, which is risky given the limited nature of the texts.
- Unfounded assumptions about cultural origins: Caskel suggests that writing, drawing, and religiosity in these societies are “remnants of ancient civilizations” without evidence.
- Misinterpretation of artistic and religious elements: Caskel claims connections between Thamudic rock drawings and Minaean reliefs, and assumes deities were borrowed from city dwellers, all without supporting evidence.
- Contradictory reasoning: Caskel emphasizes similarities between Thamudic inscriptions and later Arabic poetry while denying that Thamudic authors were Bedouin.
- Caravan theory: Caskel’s belief that the inscriptions were made by caravan members rather than nomads contradicts the wide distribution of these inscriptions across desert regions.
- Lack of evidence: Caskel provides little concrete evidence for his claim that “Bedouin form of society and ideology prevailed” throughout Arabia during this period.
- Unsupported assertions: Caskel asserts that most oases were controlled by nomadic or settled tribes, but doesn’t demonstrate how this proves “Bedouinization.”
- Assumption of tribal uniformity: Caskel assumes that the tribal systems of settled and nomadic groups were identical, without providing evidence.
- Genealogical fallacy: Caskel seems to base his claim on the existence of genealogical relationships between settled and Bedouin tribes, as systematized in the Umayyad period.
- Lack of explanation: Caskel doesn’t explain how the supposed adoption of Bedouin society and ideology by sedentary populations occurred.
- Overgeneralization: Caskel appears to assume that any genealogically-based society in northern Arabia must be “Bedouin,” ignoring that such social structures are not unique to Bedouin societies.
- Creating artificial conditions: To support his theory of change in “pan-Arabian” society, Caskel downplays the existence of tribal structures before the 6th century AD.
- Caskel’s Process of “Bedouinization”
- Implausibility of the proposed process: The idea that urban traders could easily transition to nomadic life is highly improbable due to:
- Practical difficulties in acquiring and managing herds
- Competition for resources with existing nomads
- The vast difference in knowledge and skills required for nomadic life
- Contradictory claims: Caskel inconsistently attributes “Bedouinization” to both ex-urban traders becoming nomads and ex-nomads settling in oases.
- Lack of depopulation: Contrary to Caskel’s assumption, oases in Arabia were not depopulated in Late Antiquity.
- Misunderstanding of oasis economics: Oases could survive without long-distance trade due to their primary resources and local trade with nomads.
- Unsupported distinctions: Caskel’s claim that pre-4th century AD nomads were very different from later “Bedouin” is not based on reliable evidence.
- Overgeneralization: Caskel assumes that the presence of “tribal” organization in both nomadic and sedentary groups proves “Bedouinization,” without considering other explanations for this social structure.