halachic paradox revisited: ksav al gabei ksav

I wrote about halachic paradoxes a few months ago, but since my son came across this sugya in his learning, I wanted to revisit the topic and mention a case which came up in a comment to the earlier post (thanks AddeRabbi). (My son attends a very nice Shabbos chaburah in Esvan d’Oraysa by a talmid chacham who told the boys who attend that when he was a teenager he had a notebook where he jotted down the many paradoxes raised by the Minchas Chinuch. Not a bad idea.)

Gittin 19 raises the question of “ksav al gabei ksav” – is one traces over letters which have already been written, is that tracing considered an act of writing (or erasing the existing letters) or not? Among the issues this impacts is 1) whether a get written shelo lishma can be traced over to create a new ksav done lishma; 3) whether tracing over existing letters is an act of chilul Shabbos. The Minchas Chinuch (in meleches koseiv) asks what the din would be if someone traced over a get on Shabbos to make it lishma? If ksav al gabei ksav is considered writing, then the writer, in his attempt to create a valid get, has been mechalel shabbos. However, if the writer is a mechalel shabbos, then he is not an acceptable sofer to write gitin and the get is pasul. If the get is pasul, then the second ksav has not accomplished anything, and it is not an act of writing. But if it’s not an act of writing, the sofer is not a mechalel shabbos, back to square 1 and the circle repeats.