Bechoros 29a: The Sages permitted Illa of Yavneh to charge a fee for checking a firstborn animal for blemishes (in the absence of the Beis Hamikdash, a firstborn is permitted to eat only if it has a blemish). He charged four issar for a sheep or goat and six issar for a cow, regardless of whether he declared it blemished or unblemished.

Shulchan Aruch Yoreh Deah 18:18: The shochet, who inspects the animal for defects after slaughter, must charge the same fee whether the animal turns out to be kosher or treif, lest he come to rule leniently in order to make more money. For this reason, it is customary in some places that no butcher slaughters and checks his own animals, but rather he has it done by the community shochet.

בכורות כט. כאילא ביבנה שהתירו לו חכמים להיות נוטל ארבע איסרות לבהמה דקה וששה לגסה בין תם ובין בעל מום.

שו”ע יו”ד י”ח סי”ח הטבח צריך שיטול שכר מן הטרפות כמו מן הכשרות, הגה שלא יבא להקל להכשיר כדי לקבל שכרו מן הכשרות ולכן נהגו בקצת מקומות שאין אדם שוחט ובודק לעצמו אלא אותן הממונים מן הקהל.

Once, the kosher bakery owners of Antwerp met together and decided to stop giving the keys of their bakeries to the mashgichim. [Presumably, they didn’t want to have to wait for the mashgiach to come and open the bakery in the morning.] The Rav Hamachshir, Rabbi Chaim Kreiswirth, did not back down. Taking several of the heads of the kehillah with him, he went around on Shabbos to all the shuls and announced that until further notice, he was removing his hashgacha from all the bakeries, with no exceptions. He added that if anyone wanted bread, he permitted them to eat “pas paltar” – kosher bread baked by a non-Jewish baker.

The following Friday, an oven was set up in the courtyard of the shul, and the Rav himself, together with the heads of the kehillah, baked challah for Shabbos, for those who wished to be strict and not eat pas paltar on Shabbos.

One of the bakery owners was a member of a prominent Chassidishe kehillah, and an influential member of that Chassidus announced in shul, “So-and-so is totally reliable and I will continue to eat his products, with or without hashgacha!” And to show he meant what he said, the next morning he brought cakes and other baked goods from that bakery to shul.

Not long afterward, this man contracted a stomach illness and died. After that, no one dared defy Rav Kreiswirth’s authority in Antwerp. 

Source: Mayim Chaim, page 144

[The above quotation from Yoreh Deah 18 explains why we need hashgacha in general, rather than relying on a business owner who is a shomer Torah. But this story deals with a different aspect of hashgacha: how to handle a store where the labor is done by non-Jews. Does the mashgiach have to be present all the time, or can he make occasional, unexpected visits (יוצא ונכנס)? Rav Kreiswirth held that in a bakery, the mashgiach must be present all the time, and the only way to enforce that is by giving him the only key to the bakery.

Why wasn’t he willing to rely on occasional visits? After all, the Mishnah (Avodah Zarah 61a) says that יוצא ונכנס works, and it’s brought in Shulchan Aruch. The answer may be that occasional visits make workers afraid to do something that takes a while, such as unloading non-kosher ingredients and using them. But in a Jewish-owned bakery, which needs to be pas yisroel (the heter of pas paltar only applies when the owner is a non-Jew, as the Shach 112:7 says), lighting the oven takes only a second. If the workers were to open the bakery in the morning and find the pilot light out, they might relight it, and when the mashgiach arrives half an hour later, he would never know that anything had gone wrong.

This point is made by R’ Moshe in his teshuva requiring a mashgiach temidi on fish (Igros Moshe Yoreh Deah 3:8). These fish were sold without the scales, so the customer would have no way of knowing if they were kosher fish. He writes that if the skinning were done by hand, the mashgiach could walk in and out, and that would be enough to deter the workers from skinning a non-kosher fish lest the mashgiach walk in and catch him in the middle. But the skinning is done very fast by a machine, and it would be easy to get it done in that moment when the mashgiach’s back is turned. Therefore the mashgiach must check every single fish.]

Bechoros 28b: If a rabbi judged a case and erred, exonerating the party who owed money, or forcing the party who did not owe to pay; or he declared the clean unclean, or the unclean clean – what was done was done, and the rabbi must pay for the loss out of his pocket. But if he was an expert judge, he is exempt. Rabbi Tarfon once ruled that a cow that had its uterus removed was treif, and as a result the owner fed it to the dogs. Later they asked the Sages at Yavneh, who permitted it, because Todus the doctor said that all cows and pigs exported from Alexandria of Egypt are sterilized by having their uteruses removed. Rabbi Tarfon said to himself, Tarfon, you will lose your donkey. Rabbi Akiva said to him: No, you are an expert judge and and therefore you need not pay. 

Gemara: Why did Rabbi Akiva need to say that Rabbi Tarfon was an expert? Even if he not been, since his mistake was in an explicit Mishnah (Chullin 54a), his psak is immaterial, he never really forbade the meat, and he should not need to pay for the fact that the asker threw out the meat (because it is grama)! – Rabbi Akiva was giving two reasons: First of all, you are exempt from paying because the mistake was in an explicit Mishnah. Secondly, even if you had erred in logic (in a halachic point that is not explicit in the Mishnah), and you ruled incorrectly, you would still be exempt from paying because you are an expert.

בכורות כח ע”ב: דן את הדין, זיכה את החייב וחייב את הזכאי, טימא את הטהור וטיהר את הטמא ־ מה שעשה עשוי, וישלם מביתו, ואם היה מומחה לב״ד ־ פטור מלשלם… ומעשה בפרה שניטלה האם שלה והאכילה רבי טרפון לכלביםֹ ובא מעשה לפני חכמים ביבנה והתירוה שאמר תודוס הרופא: אין פרה וחזירה יוצאה מאלכסנדריא של מצרים שאין חותכין האם שלה בשביל שלא תלד. אמר רבי טרפון: הלכה חמורך טרפוןִ אמר לו ר׳ עקיבא: ר׳ טרפון, אתה מומחה לב״ד ־ וכל המומחה לב״ד פטור מלשלם. גמרא. ותיפוק ליה דטעה בדבר משנה, וטעה בדבר משנה חוזרִ חדא ועוד קאמר, חדא ־ דטעה בדבר משנה חוזר, ועוד א״נ בשיקול הדעת טעיתה ־ מומחה לב״ד אתה, וכל המומחה לב״ד פטור מלשלם.

There was once a rav who was mesader a get based on his assessment that the husband was generally mentally sound; his insanity was confined to one particular area of life. But in his older years, the rav learned the sugya about insanity in Chagigah 3b more carefully, re-examined the case, and decided that the husband had been mentally unfit to give a get. By this time, the woman had remarried and raised a whole family. According to the rav’s current conclusion, her children were now mamzeirim. The rav asked Rabbi Yitzchok Zilberstein whether he would be punished by Hashem for causing these children to be mamzeirim.

Rav Zilberstein responded by comparing this to various cases in Shas:

1. We find that Tannaim and Amoraim changed their halachic rulings. For example, in Niddah 14a: “All of Rabbi Chiya’s life, he used to rule that this case was unclean, but in his old age he would rule it clean.” He was allowed to reverse his earlier psak, even though his later line of reasoning was more lenient, because the Gemara (Shabbos 152a) says that the older a talmid chacham gets, the greater his wisdom. Similarly, the Levushei Mordechai (Orach Chaim 96) was asked about a talmid chacham who used to be strict about a certain halachic question, but in his old age decided to follow the lenient opinions. The Levushei Mordechai says that he has the right to do so, because the older a talmid chacham gets, the greater his wisdom.

• We also find that Rabbi Tarfon once ruled that some meat was treif and then changed his mind, and he was held responsible: he had to give his donkey to pay for the loss of the meat, which had already been fed to the dogs. (And Rav Elyashiv commented that Rabbi Tarfon mentioned his donkey, although he was wealthy (Nedarim 62a) and could certainly have paid in cash, to show that he was not considered like a regular debtor, who has bankruptcy rights to keep his work animal; he is a mazik, who must pay no matter what.) Doesn’t this show that the rav in our story would be held responsible for causing the children to be mamzeirim?  

Rav Zilberstein says no. Rabbi Tarfon’s case was different because he erred about an explicit Mishnah (טועה בדבר משנה). But if the rav studies and makes a mistake in logic (שיקול הדעת), his original psak is not reversed (i.e. even if the meat were still in existence, it would be fed to the dogs as per the original psak). But the rav still has to pay for the loss. Here too, this rav’s psak remains intact that the woman was permitted, and the children are not mamzeirim.

Source: Chashukei Chemed, Niddah, p. 149

[This piece in the sefer Chashukei Chemed is cryptically written, and as the author says in the introduction, “With my brief notes I was unable to cover all the wellsprings of the Torah’s endless wisdom, and I have only come to stimulate the heart of the learner; may the wise hear and become wiser.” Let’s do our job and break this question down, and then see what his sources prove.

It seems that both the asker and Rav Zilberstein started with the assumption that the earlier psak is reversed and the children are mamzeirim. The question was only whether the rav would be blamed in heaven.

Source 1) is a proof to this assumption: Rabbi Chiya changed his mind and ruled leniently, and his lenient ruling became the final halacha because he became wiser with age. Here too, the rav became wiser with age, his later psak is binding and the children are mamzeirim.

However, there is a big difference. Rabbi Chiya changed his mind about a certain commonly-asked shailah (a woman who wiped herself with a cloth that might have been bloody already, and then found blood on it). It does not say that he went back and contacted women he had declared impure, and told them that they were actually pure. He merely began to rule leniently on cases brought to him from then on. Here, on the other hand, we want to know whether the actual get ruled on by the rav is now retroactively invalid.

Source 2) comes closer: it seems to prove that the actual meat pronounced treif by Rabbi Tarfon is retroactively kosher. To this, the response is that in Rabbi Tarfon’s case, the mistake concerned an explicit Mishnah. But, the Gemara says, when the mistake was in logic (which the Gemara in Sanhedrin defines as ruling like one valid opinion when the other one is the widely accepted opinion), the psak is not reversed. The rav pays for forbidding the meat, but the meat remains forbidden. In our case, the get is still a good get and the children are not mamzeirim. And here, the rav would not pay, because there is no damage to pay for. Furthermore, in the case ofשוטה בדבר אחד (insanity in one area only) there is no clear-cut halachic consensus; indeed, this was the subject of a major controversy in the 1700’s (the Klever Get). And no two cases are alike; the psak would depend on the exact symptoms. This is all the more reason to say that the rav’s first psak stands.

However, we can ask: one of the cases in our Mishnah in Bechoros is טיהר את הטמא – “he declared the unclean clean.” On this, the Mishnah concludes: “What was done was done, and the rav must pay from his pocket.” This is certainly not talking about a rav who erred in an explicit Mishnah, because in that case his psak would be reversed; we don’t say “what was done was done.” Rather, it is talking about an error in shikul hadaas. (See as well Sanhedrin 33a which says that the Mishnah in Bechoros is talking about shikul hadaas.) Therefore, “what was done was done” – the rav‘s psak that the fruits are clean stands. So what is meant by “the rav must pay from his pocket”? What must he pay? He has caused no damage at all! Even if the fruits were subsequently mixed with other clean fruits, and even if the rav mixed them with his own hands as the Gemara says later in Sanhedrin, no damage was done.

Possibly the answer is that the idea that shikul hadaas is not reversed is only the opinion of Rav Sheishes, and Rav Sheishes would indeed say that the rav who mistakenly ruled “clean” would never have to pay; the words “he pays from his pocket” are referring only to the other cases in the Mishnah. The idea that the rav pays because he mixed the unclean fruits (thought to be clean) into the other fruits is said by the Gemara only according to Rav Chisda.

To clarify: in Sanhedrin 33a, the Gemara poses a contradiction between the Mishnah in Bechoros, which says “what was done was done,” and the Mishnah in Sanhedrin, which says that a dayan can change his mind and reverse a monetary ruling. Three answers are provided; one of them is Rav Sheishes, who differentiates between an error in an explicit Mishnah and an error in shikul hadaas. But there are two other Amoraim, Rav Yosef and Rav Chisda, who answer the contradiction. According to them, perhaps all mistaken rulings, even in shikul hadaas, are reversed. So we can propose that it might be only Rav Sheishes who says that in the Mishnah in Bechoros, the psak is not reversed.

How do we pasken? The Shulchan Aruch in Choshen Mishpat 25:2-3 seems to go like the other Amoraim, not Rav Sheishes. (In S’if 2 he brings Rav Yosef and in S’if 3 he brings Rav Chisda. The only cases where the Mechaber says the psak stands are where the psak cannot be reversed, or where the dayan carried out the psak himself.) According to this, in our case psak would be reversed and the children would be mamzeirim. (Ironically, the Gemara says according to Rav Sheishes that shikul hadaas means ruling like one opinion while the sugya d’alma, the general consensus, is like the other opinion – and Rav Sheishes himself fits this description. He is one out of three opinions in Sanhedrin, yet the Gemara in Bechoros quotes only Rav Sheishes, without mentioning his name!)

The Shach (s’if katan 14) does pasken like Rav Sheishes, but says the psak only stands in the case where the rav erroneously forbade something, because of the principle of שויא חתיכה דאיסורא – the rav, so to speak, created a new issur. But if he erroneously permitted something, even Rav Sheishes would agree that the psak is reversed. Based on this as well, the children in our story are mamzeirim.

One could still argue that perhaps our case is not שיקול הדעת because there is no halachic consensus on the subject. Thus, who is to say the rav’s second psak is correct – perhaps the first was correct? Add to that the concept of לא בשמים היא – when a rav rules on an actual case in a gray area, Hashem agrees that the halacha follows his ruling. (See the Hakdama to Igros Moshe v. 1.) The first time, the rav was approached by a woman who needed a get. He paskened on an actual case, so we say לא בשמים היא and that is the real psak on the case. But the second time, he was just learning on his own and thinking, so his conclusion may not have the status of a psak.

Finally, if this rav really held that his first psak was null and void and the children were mamzeirim, why didn’t he go and inform them? Perhaps he was following the Gemara in Kiddushin 71a, משפחה שנטמעה נטמעה – “A family that has a mamzer mixed into it, and no one knows, is permitted.” This is brought down by the Rema in Even Hoezer 2:5. The Rema explains that the one who knows about it should not marry into the family, but he should not publicize it to others. He also stipulates that this is only after they have mixed in (i.e. married into other families), but as long as they are not mixed, he should publicize it. Accordingly, it may be that all the children of the woman in our story were already grown up and married, and therefore the rav did not inform them or spread the word.]