`zebra'

[Rick Mc Callister]

I'm wondering if the word was PERCEIVED by Spanish speakers as an Arabic
word, thus the initial vowel as dropped as an "article" or whether indeed
the word may have passed from Romance to Arabic to Spanish as some words did

[snip]

>But this pass via (Andalusian) Arabic is unnecessary if what we want to
>account for is the form <zebro> OSp /dzebro/. In *eciferum <c> would
>palatalize/affricate before a front vowel > [ts] and would lenite to
>[dz] intervocalically, as would /f/ > [v]. The second <e> would be lost
>by syncope in a post-stressed non-final open syllable. <-um> > /-o/ is
>entirely regular, as is short <i> > /e/. That gives us /dzevro/. Now we
>sometimes find, without it being exactly regular, /br/ or /rb/ for
>expected */vr/ or */rv/. A case in point would be <Iberum> > Ebro
>(intervocalic <b> normally lenited to /v/ in OSp). The aphaeresis of
><e-> is the most irregular feature, but not all that surprising, and not
>assisted by the Andalusian Arabic story.
>	The above is precisely why Romance philologists proposed
>*eciferus. The <equi-> > *eci- remains ad hoc.

[snip]