Maintaining and Repairing an Invalid Sefer Torah

Last week's issue discussed what one does if a mistake is discovered in a sefer Torah during k'riat haTorah. This week's issue will discuss what must be done with a sefer Torah that contains mistakes. What actions must be taken to ensure that the sefer Torah is not used for k'riat haTorah? Furthermore, at what point do we suspect that the discovered mistakes are not the exception to the rule but rather the norm? What actions must be taken in such a situation to restore the sefer Torah's validity?


Sefer She'aino Mugah

The Gemara, Ketubot 19b, states that one may maintain possession of a book that was not edited properly (sefer she'aino mugah) for thirty days. After thirty days one must dispose of it. What type of book does the Gemara refer to? Rashi, ad loc. s.v. Sefer, explains that the Gemara refers to any of the books of Tanach. Rambam, Hilchot Sefer Torah 7:12, rules that a sefer Torah that was not edited properly may not be maintained for more than thirty days. The implication is that Rambam is of the opinion that the Gemara refers only to a sefer Torah, and not to the other books of Tanach.

R. Avraham I. HaKohen Kook, Da'at Kohen no. 174, explains that the point of contention between Rashi and Rambam revolves around the nature of the concern for maintaining possession of an unedited book. Rashi is of the opinion that one may become misinformed by learning from a book that contains errors. Therefore, this concern applies to all of the books of Tanach. Rambam is of the opinion that the concern is that one may read the book publicly for the mitzvah of k'riat haTorah. Therefore, the concern only applies to a sefer Torah and not to the other books of Tanach. [R. Kook is of the opinion that Rambam does not allow an invalid sefer Torah for k'riat haTorah. See last week's issue.]

R. Kook notes that there is a practical difference between Rashi's opinion and Rambam's opinion. According to Rambam, since the only concern is that someone may use the Torah for k'riat haTorah, one does not have to dispose of the sefer Torah. It is sufficient if a symbol is placed on the sefer Torah in a way that everyone knows that it should not be used. However, according to Rashi, placing a symbol on the unedited book is insufficient. A person will not be deterred from learning from this book because he will assume that the mistake is not in the section which he is learning from. Therefore, one must dispose of the book to ensure that nobody learns from this book.

R. Kook writes that according to Rambam, one can understand the practice that when a sefer Torah is invalid, a gartel (belt) is placed on the outside, indicating that the sefer Torah is invalid. According to Rashi, one can question what the justification is to maintain possession of an invalid sefer Torah even when a gartel is placed upon it. R. Kook explains that according to Rashi only books which are normally used to learn from are included in this prohibition. However, nowadays, nobody learns from an actual sefer Torah. Therefore, one is permitted to maintain possession of a sefer Torah that contains errors.


Repairing an Error

The Gemara, Menachot 29b, states that if a sefer Torah is ridden with errors, one may not correct the errors. Rather one must bury the sefer Torah. The Gemara defines a sefer Torah ridden with errors as a sefer Torah with four errors on every page. If there is one page with less than four errors on it, and majority of the letters of the sefer Torah are free of error, one may correct the entire Sefer Torah.

Although one may correct a sefer Torah that is not ridden with error, if three mistakes are discovered, the process is more complex. Rashba, Teshuvot HaRashba 7:287, rules that if a sefer Torah is found to contain three errors, it is not sufficient to correct those errors. Rather one must check the entire sefer Torah to ensure that there are no other errors. Shulchan Aruch, Yoreh Deah 279:3, codifies Rashba's ruling.

R. Yonah Landsofer, Bnei Yonah 279:3, lists two leniencies regarding Rashba's ruling. First, one does not have to recheck the entire sefer Torah unless all three mistakes are known to exist simultaneously. If a mistake is discovered and corrected, and then two more mistakes are discovered, one is not required to recheck the entire sefer Torah. Eliah Rabbah 143:10, disagrees and contends that if three total mistakes are discovered, even if they are not known to exist simultaneously, one must recheck the entire sefer Torah.

Second, R. Landsofer quotes an opinion that Rashba's ruling only applies where the identity of the original sofer is unknown. If however, the sefer Torah was originally checked by an established expert, one does not have to recheck the sefer Torah upon finding three mistakes. Mikdash Me'at, ad loc., claims that R. Landsofer personally disagrees with this leniency, and only quotes this opinion in order to reject it.

R. Ovadia Yosef, Yechave Da'at 5:58, rules that while each leniency may not be sufficient to allow use of the sefer Torah without rechecking it first, one may be lenient by combining both leniencies. Therefore, if the sefer Torah was originally checked by an established expert, and all three errors were not known to exist simultaneously, one is not required to recheck the sefer Torah.

It would seem that there is another ground for leniency regarding Rashba's ruling. The errors that are found must be divided into two categories. There are errors which are assumed to have occurred at the time the sefer Torah was written. There are also errors that are assumed to have occurred due to the fading or smudging of the ink. If three errors of the first category are discovered, according to Rashba, the entire sefer Torah must be checked for errors. This is because the three errors indicate that the sefer Torah was never written and checked properly. Similarly, if three errors of the second category are discovered, the entire sefer Torah must be checked. This is because the three errors indicate a strong likelihood that the ink faded or smudged in other places. However, if one error is discovered from the first category, and two from the second, or vice versa, there should be no obligation to check the entire sefer Torah. There is no clear indication that the sefer Torah was not written properly, There is also no clear indication that ink faded or smudged in other places. Therefore, there exists no indicator that the sefer Torah must be rechecked.


Checking With a Computer

Soferim have developed computer programs that can check sifrei Torah for errors. Both R. Yitzchak Weiss, Minchat Yitzchak 10:89, and R. Eliezer Waldenberg, Tzitz Eliezer, 18:57, conclude that a computer cannot replace a human in fulfilling the requirement that a sefer Torah must be checked. They both note that the computer does not have the ability to detect certain errors, and therefore cannot be relied upon.

Both R. Weiss and R. Waldenberg address using a computer to check a sefer Torah before its initiation. Perhaps one can argue that even though the computer does not provide the degree of accuracy necessary to check a sefer Torah initially, there may be instances where the computer can obviate the need to recheck the sefer Torah after discovering three errors. If the computer can show that from a statistical standpoint, the three errors are an anomaly, the three errors no longer indicate that the rest of the sefer Torah was written improperly. For example, if three errors are discovered, and after checking the sefer Torah with the computer, the computer only finds those three errors, one can assume that the rest of the sefer Torah is valid. Although the computer can neglect to discover an error, if the three errors really indicate that there are other errors, the computer should have found at least one other error. Since the indicator itself is suspect, one should revert to the original assumption that the sefer Torah was written properly.