As surmised in SED I No. 204, the unusual Anlaut ˀa- may be due to contamination with the reflexes of PS *ˀanp- > *ˀapp- “nose”, which came to designate “face” throughout Aramaic.
PS
*piˀ-at- - border, side
(EDA P 519, SED I No. 204)
*-V- in the protoform should probably be replaced by *-i- supported by Eth. and Soq.; u- in Akk., Mnd. and Hbr. (if related) might be explained by the influence of *p-, and -a- in Syr., by the influence of -ˀ-.