Shavuos 35b - Erasing the Name of Hashem

There is a machlokes in the braisa whether the prefix and suffix of the names of Hashem are included in the issur to erase the name of Hashem. We pasken like the Acheirim who hold that a prefix is mutar to erase, but a suffix is assur to erase because the name of Hashem has already been written so it spreads the kedusha onto the suffix as well. This would imply that if a sofer was supposed to write l'hashem and forgot the prefix of lamed, and added it after writing the name of Hashem, it would be assur to erase the lamed since it is being added on to a preexisting name of Hashem so it should be treated like a suffix.

R. Akiva Eiger (teshuvos 70) records a machlokes ha'poskim whether one can be in violation of erasing the name of Hashem by adding a letter. For example, can a daled be added to the yhvh, in between the final to letter so that it would now spell yehuda. He is medayek from a le'vush (y.d. 272) that this would be an issur of erasing the name of Hashem. However, R. Akiva Eiger points out that in Teshuvos Avodas HaGershuni (95) there was a situation where the sofer should have written elokim and instead wrote elokeichem, and he permits turning the kaf into a mem, so that it would spell elokim – mem, and it would then be permitted to erase the final mem since it is no longer part of the name of Hashem. Clearly, the Avodas Ha'Gershuni doesn't consider writing to be erasing the name of Hashem, because by turning the kaf into a mem, he would essentially be erasing the suffix of “chem”, which is tantamount to erasing the name of Hashem. Following the method of trying to reduce the machlokes, R. Akiva Eiger suggests that any erasing through writing would only be an issur d'rabonon because it is similar to mechika but doesn't truly qualify as erasing. Based on this, in a situation where the sofer should have written – ויאמר יעקב אל יוסף א-ל ש-די, and instead wroteויאמר יעקב אל ש-די, he would be allowed to add the words יוסף אל, because even if the original “el” was the name of Hashem it would be an issur d'rabonon, and it is only a safeik if the original “el” was meant to be the name of Hashem.

R. Akiva Eiger equates adding a letter that would change the entire name to be chol, to adding a letter that would cause the suffix to no longer be part of the name of Hashem. But, perhaps there is a difference. When adding a letter would entirely erase the name of Hashem such as a daled to yhvh, it could be an issur d'oraysa, but by simply chaning the kaf of elokeichem to a mem thereby causing the final mem to be extra, would not be d'oraysa only d'rabonon. Furthermore, perhaps the actual erasing of a suffix would only be d'robonon. If this is true, it could be that we can be more meikel with erasing through writing so that the levush and avodas hagershuni wouldn't be arguing. The levush held it was assur to add a daled since the actual erasing would be d'oraysa, but the avodas hagershuni allows the changing of the kaf to a mem since the actual erasing may only be d'rabonon.