Frontiers in North Africa

(a) The Tunisian border

As far as Tunisia was concerned, the precolonial situation involved a delimitation between the Beylik of Constantine and the Beylik of Tunis. This, as was the case with the border march between Tunis and Tripolitania, was an area of indeterminate political authority which was linked to communal loyalties to separate centres of power in Tunis or Constantine. Insofar as a 'border' could be defined, this followed the delimitation created by the expansion of the Beylik of Tunis between 1609 and 1683 (Pitcher 1971; 143). This roughly followed the modern border line from the Mediterranean coast down to the border west of the Chott el-Djerid. It then swung eastwards, to parallel, first, the southern edge of the Chott before running into the Jabal Nafusa escarpment. The line was not impermeable, however, and the Tunisian Beylik considered that its authority extended towards the Saharan oases south of the Chott (viz Brown 1974; 132).

With the arrival of the French Protectorate in 1881, detailed delimitation and demarcation was undertaken between the Algerian administered areas of Constantine and Tunisia. This culminated in a definitive demarcation from Tabarka, on the Mediterranean coast, to Chott Rharsa by 1900. In 1901, an agreement between the Bey and the Resident General delimited the border between Chott Rharsa and Bir Romane, an agreement that was formalised on June 28, 1902 by the Bey (Brownlie 1979; 92).

The remaining section of the border was to show a sharp divergence from the precolonial situation, for a large element of the Sahara was included in Tunisia. This was largely the result of the fact that this delimitation involved the Algerian TDS administration and that it had to coincide with a delimitation undertaken with the Ottoman empire in Tripoli in order to establish a tripoint with Tripolitania close to Ghadames. According to the 1910 Ottoman-Tunisian agreement, the tripoint was to be located 15 km south of Ghadames, at boundary marker 223 and a subsequent demarcation in 1911 established this to be so. French administrative practice in Algeria, however, set the tripoint at boundary marker 220, to the north of Ghadames in 1911 and again in 1923. This delimitation was followed by metropolitan French practice after 1929 and, when a definitive delimitation and demarcation of the border was agreed between Algeria and Tunisia after independence in 1970, it was this boundary marker that was chosen as the tripoint.

The Tunisian boundary with Algeria, therefore, represents a compromise between two different principles. The section to the north of Bir Romane has been delimited on the basis of a French interpretation of customary precolonial usage. The result was that the flexible and moveable boundary based on demographic and communaI allegiance of the Ottoman period was replaced by an impermeabIe territorial demarcation which did not allow for traditional transhumant populationmovement and corresponded to European preconceptions of rigid territorially based administration. The section further south shows all the characteristics of a boundary created for administrative convenience in a region of sparse population. In fact, the failure of the border to correspond to population location in these regions led to constant problems with the Cha'amba nomads whose traditional movement patterns regularly transgressed it (Martel 1965; 1-109).