I've lately been researching the origin of the Libyco-Berber script, and I ran into this article by Salem Chaker and Slima Hachi: http://www.centrederechercheberbere.fr/tl_files/doc-pdf/chak-hach.pdf
There is a very strange discussion in this article on the words 'to write' and 'writing', which are very commonly attested in a wide variety of Berber languages.
The authors draw a connection with another verb which means 'to open', and considers them derived from the same root R(W/H). On the surface, this might seem reasonable, cf. Kb. aru'to write' and Tuared ar'to open', but even at the time of writing (2000) this is evidently wrong.
The only consonant that these two roots have in common is the letter r, which it also has in common with the verb 'to open' and ' lung', etc.
The verb 'to write' can be reconstructed as *ăʔrəβ, cf. Ghd. orəβ, Awj. arəv, Kb. aru, Zenatic ari, and probably Zng. iʔriʔ, īriʔ 'to dictate'. Dictionaries and wordlists of all of these languages were available in 2000 (except for Zenaga), and multiple articles and books have been written about the Proto-Berber consonat *β which is just completely absent in the word 'to open', which we must reconstruct as *arəʔ, or *ăhrəʔ, cf. Ghd. ar Awj. yar Tuareg ar
Even if one does not believe that Proto-Berber as a reconstrutable entity, one can still conclude that these roots must be different, as both verbs are found in Ghadames and Awjila, where one clearly has different root consonants from the other verb.
A connection between these two verbs is therefore completely absent.