Maridi Arabic was a possible Arabic pidgin apparently spoken in the upper Nile valley around 1000 CE. If legitimate, it would be the oldest record of a pidgin. It is known from just fifty words in an 11th-century text.
In 1068, the Spanish geographer Abū ʿUbayd al-Bakrī published the words of a traveler from Aswan, who had complained to the Caliph in Egypt that in the town of "Maridi", black Africans had "mutilated" the Arabic language.
It appears that the text may be a caricature of an actual pidgin, or at least of partially acquired Arabic. It shares several features with Juba Arabic and Nubi Creole: a complete lack of inflection, including articles, possessive markers, and agreement; a single negative marker; ʔmny for the third-person dual pronoun, and perhaps the use of jaʿal 'make' rather than colloquial qaʿad 'stay' for the progressive.
It is not known where "Maridi" was. It has been concluded that it was in Mauritania, with the single non-Arabic word in the text (below), kyk (people?), hypothesized to be from Songhai-koi 'people who' plus the Arabic plural -k. However, there are also indications that it was located in Upper Egypt or northern Sudan, as several of its features are found today only in that region. Specifically, Classical q has become ʿ (ʿayn), fwq "on" has become fwʾ, and j has become d. The last is a sound correspondence found today only in Upper Egypt.