The sources for the study of medieval Libya:
the corpus diversity and the problematics linked to its exploitation
Hafed Abdouli (Sfax University)
Session II — 15 December 2021
Our colleague Hafed Abdouli (Sfax University, Tunisia) gave a talk in Arabic — المدونة المصدرية لتاريخ ليبيا خلال العصر الوسيط: تنوع المخزون وإشكاليات/محاذير الاستعمال — in two parts. Firstly, he reviewed in details the richness and variety of the corpus available to work on medieval Libya. Secondly, he deeply and critically revised the Islamic conquest of Tripolitania as a case-study. This innovative approach opens the way to a fruitful rereading of the accepted traditional narrative.
I. The diversity of the corpus for Tripolitania
The written sources, i.e. the discourses, could be gathered under three questions: When? Who? Where? Respectively for historiographies, biographies and geographies.
Historiographies are mainly centered on events basically organized chronologically. They comprise subgenres like the conquests (futūḥ), histories (aḫbār and tawārīḫ), chronicles (ḥawliyyāt). For the present purpose, we could distinguish between different areas. They are listed in the LibMed Zotero database (https://www.zotero.org/groups/2901756/libmed). Primo, Eastern texts, among them Ibn ʿAbd al-Ḥakam’s Futuḥ Miṣr will be extensively discussed. Secundo, Western texts (Ifrīqiya and the Maghreb), where we will refer to al-Raqīq’s lost Taʾrīḫ. Tertio, Shia writings, e.g. al-Qāḍī l-Nuʿmān. Quarto, the genealogies (ansāb).
Biographical dictionaries (tabaqāt and tarāǧim) are chiefly centered on actors. We could divide them between Western ṭabaqāt — ʿulamāʾ Ifrīqiya and ibāḍiyya, e.g. Abū l-ʿArab, al-Mālikī and al-Qāḍī ʿIyāḍ or al-Darǧīnī and al-Šammāḫī — and Eastern tarāǧim.
Geographical works are finally centered on places. They encompass different shapes, be they itineraries (masālik wa-mamālik) — see the editions by de Goeje and their subsequent revisions, adding that al-Bakrī and al-Idrīsī play a part in our reflections —, encyclopaedias (maʿāǧim ǧuġrāfiyya) — Yāqūt being a major one — and the numerous travelogues (raḥalāt): al-ʿAbdarī, al-Tiǧānī/al-Tuǧǧānī (see A. El Bahi’s convincing recent article on this orthography), al-ʿAyyāšī, etc.
The archeological sources, i.e. the field or material evidences, allow a decisive “cross-checking” with the texts. As will be noticed, LibMed has strongly invested in the discipline with conferences planned this year and next year.
The extant reports and studies started under the Italian colonization, to be developped by other surveys and explorations after WWII. The moutains of Tripolitania received attention at the end of the 40s, to be lately updated by Muftāḥ al-Ḥaddād (Tarhuna University, Libya). The predesertic zone was searched via British cooperation. As for the coastal area, mainly in the vicinity of Leptis Magna, joint missions with Italian archeologists began at the end of the 20th century and stopped with the war.
Given the nature of the corpus considered above, some problem(atic)s arise and result from the following aspects. These points will be illustrated in the subsequent case-study.
There is a huge temporal gap that separates the events of the first centuries from their documentation. Ibn ʿAbd al-Ḥakam (ca 182-257/798 or 9-891) exemplifies this issue, especially as we will see for the referential question. The meaning of words (technical expressions, toponyms or patronyms) can be understood in a way that misses the original context. The same would apply for relations between groups and authorities.
The selection of the first sites to be excavated and the methodology employed during these early phases is not devoid of biases and their reports have to be reassessed.
The elitist approach of the men who wrote down the texts we are relying on, focusing mainly on the active groups, politically and ideologically speaking, tend to leave aside the rest of the population. The Riḥla of al-Tiǧānī/al-Tuǧǧānī betrays this aspect blatantly.
The contradictions found in the narratives should make us aware of the urgent need to handle the texts we use with care and try to examine the lemmata and decide for the most reliable.
The mistakes made by the authors who tried to localize some toponyms is more frequent than thought, even when it touches well-known sites (cf. Leptis Magna in al-Bakrī’s Masālik).
II. A case-study: the transition from Antique Tripolitania to medieval Tripoli/Ṭarābulus
The book of Isabella Sjöström, Tripolitania in Transition: Late Roman to Early Islamic Settlement (1993) was a milestone to study the evolution of the region of Tripoli from Antiquity to the early Islamic period, and more generally the issue of how the cities were tranformed in the course of this passage.
Interrogations remain, chiefly the transfer of the capital from Leptis Magna to Oea/Tripoli. Was there a break (qaṭīʿa) as is commonly accepted? This would be explained by the decline of Leptis Magna and the rise of Oea at the end of Antiquity, or was even simply caused by the Islamic expansion. The first step requires returning to the sources.
ʿAmr b. al-ʿĀṣ conquered Tripoli for the caliph ʿUmar between 21 and 23 AH/642 and 644 CE. Four sources have to considered: Ibn al-Ḫayyāṭ who quotes Ibn Lahīʿa (d. 194); Ibn ʿAbd al-Ḥakam who quotes al-Layṯ b. Saʿd (d. 179); al-Balāḏurī who quotes Bakr b. Ḥayṯam; Ibn ʿAbd al-Ḥakam who quotes ʿUṯmān b. Ṣāliḥ quoting Ibn Lahīʿa, the most detailed version so far.
But Tripoli returned to the Byzantine domination until a new and final conquest. The consultation of an Armenian Geography by Ananias of Shirak corroborates this situation. In the meantime, Barqa (and Cyrenaica) had remained controlled by the Muslims led locally by ʿUqba b. Nāfiʿ.
ʿAbd Allāh b. Abī Sarḥ is credited for the capture of Tripoli. The mention of the city walls (sūr pl. aswār) on the landside is a key to understand the problem of toponyms. Leptis or Sabrata had such fortifications. What about Tripoli/Oea? Nothing has been found — whether in Procopius or through excavations — to sustain any. Does it mean that we have to look for another Tripoli? Yes. A hint to this could be that Leo Africanus, drawing on a lost part of al-Raqīq, mentionned an “old Tripoli”.
It must have been Leptis Magna. It was called Tripoli at the time of the conquest. It had a wall on the landside, as was shown on the map. Leptis was the only city to be taken twice in the Arabic sources, in 22 AH/642 (under ʿUmar) and in 43 AH/663 CE (under Muʿāwiya), to expel once for all the Byzantines from Tripolitania.
The toponymic confusion is explained by the discrepancy between the events and their written relation. Ibn ʿAbd al-Ḥakam was not aware that Ṭarābulus meant in illo tempore Leptis Magna. How did Oea become Tripoli then?
It was known as Nubāra, even if Ibn ʿAbd al-Ḥakam and his editor, Charles Torrey, had a problem with that. The latter had to changed the text and equate Nubāra with Sabrata. His manuscripts — see mss 1686 and 1687 kept at the BnF in Paris — are however clear: after reading Tripoli, you have the parenthetical phrase “and its name was Nubāra”, where the Lisān al-ʿarab defines al-Anbār as a seaboard. The local memory can also be helpful: al-Mālikī (5th/11th c.) calls Leptis the old sūq, understanding that it had been the main place before the capital was transferred to Oea.
This must have happened under the rule of the governor ʿAbd al-Raḥmān b. Ḥabīb around 131-2 AH/749-50 CE, after the repression of an Ibāḍī Berber revolt. Only then did Oea become the regional capital and was called Tripoli.