156 EARLY MEDIEVAL SPAINtaken to wearing a crown. However, before he could be removed
from office by the caliph, he was murdered in 715/6 by some of his
own followers, incensed at being required to perform acts of obei-
sance before him.
Put as baldly as this, such an outline narrative seems reasonably
credible, and Miisa has often been compared with those such as
Cortes, who achieved great conquests only to be suspected and ill-
treated by the rulers in whose name they had acted. However, it must
be said that a considerable number of discrepancies exist between
the various Arab accounts of these events, and some of the more
fanciful and clearly legendary accretions have been omitted or ration-
alised here. Some of the details, such as the respective sizes of the
armies at the battle on the Guadalete, are also patently unreliable.^24
The earliest form of these stories relating to the events of the con-
quest of both Ifriqiya and of Al-Andalus may be found in the work of
a ninth century Egyptian writer called Ibn Abd al-I:Jakam (d. 870).
Analysis of this has shown that his historical narratives constitute a
body of legal traditions, intended to provide justification for contem-
porary arguments and opinions.^25 In other words, their relationship
to the realities of early eighth century events is by no means assured.
It was this Egyptian school of historiography that provided the first
inspiration for the beginning of historical writing in al-Andalus itself
in the late ninth and early tenth centuries, and it is thus not surpris-
ing to find the same corpus of stories forming the core of the nar-
ratives of the conquest contained in the extant Spanish Arab texts.^26
That none of these predate the eleventh century in their present
forms may help explain the proliferation of variants in the stories, but
ultimately they all represent no more than elaborations on the kernel
of mid-ninth-century Egyptian historico-juristic traditions. These lat-
ter, as previously stated, were devised to provide justifications for
contemporary religious and legal doctrines, not as antiquarian records
of events dating to a century and a half earlier.
There is thus a prima facie case for distrusting the value of much or
some of this material. If, though, it were impossible to produce an
alternative account of these events that could be shown to be both
closer in time and place and equally as internally coherent and cred-
ible, it would be necessary to leave the Arab narratives as the prime
sources of our information on the conquest, flawed as their testimony
might be suspected of being. As previously mentioned, however, there
does exist an alternative source, the anonymous Latin work, probably