Safek in Bedikat Chametz (9b-10a)

  • Rav Ezra Bick
Dedicated by Mr. and Mrs. Leon Brum for the Refua Sheleima of
Dana Petrover (Batsheva bat Gittel Aidel Leba)
and Marvin Rosenberg (Meir Chaim ben Tzipporah Miriam)
In memory of six friends and family, 
strong pillars of the Montreal Jewish community, 
who have left us in the past 7 years. 
All were אוהבי עם ישראל, אוהבי ארץ ישראל, אוהבי תורת ישראל.
Joseph (Yosie) Deitcher
Avrum (Avy) Drazin
Rabbi Joseph Drazin
Leibel Frisch
Israel (Mutch) Yampolsky
Dr. Mark Wainberg
This shiur will deal with the issue of safek specifically in regard to bedikat chametz.  The shiur will refer to the gemara beginning with the colon on 9b ("teisha tzibburim") and will cover the cases up to "safek al" (10a "...safek maza tum'a tamei").  The basic problem is to devise a consistent set of principles when one goes "li-kula" and when "li-chumra" in bedikat chametz.
The basic sources are:
1. Tosafot 9b s.v. "heinu."  Tosafot solve the problem by referring some of the cases to biur chametz (d'oraita) rather than bedika (d'rabbanan).
2.  Ra'avad (on the Rif, 4b). [until the end of the first column on 5a].
3.  Rabbeinu David s.v. "be-shlosha" (pp. 60-68).  This is the basic source for the shiur.  For the benefit of those who don't have a copy of R. David, see here.
            The gemara (daf 9b) cites a series of cases where a safek (doubt) arises as to whether or not another bedika must be performed in a house that had been previously checked for chametz. In each case, the gemara cites a paradigm ruling from another masechta and applies it to resolve the relevant safek. At first glance, the series appears to contain a number of contradictions and inconsistencies. Above all, two well-known principles appear to be violated, at least some of the time:
1. Safek de-rabbanan le-kula - a doubt concerning a rabbinic prohibition is treated leniently.
2. Sfek sfeika le-kula - a case where a double doubt exists is treated leniently EVEN IF THE PROHIBITION IS BIBLICAL. (i.e. in a case where TWO factors must be present for something to be prohibited and it is doubtful whether EITHER of them actually occurred we rule that it is permitted.)
            There are a number of solutions to these problems, which we shall examine in detail, thereby identifying the special nature of the obligation of bedikat chametz vis-a-vis safek. The Rishonim we shall utilize are the Tosafot (and a similar approach in the Ba'al Ha-Ma'or), the Ra'avad, the Ramban, and R. David. Since all four solutions are discussed at length by R. David, who carefully analyzes each one and compares them to each other, the shiur will be based primarily on his comprehensive presentation of the sugya (RD, pp. 61-68; I shall identify the section in R. David by the first word of each section).
            [For convenience sake, we primarily shall be referring to four cases, using their paradigm names; teisha chanuyot, shtei kupot (9b), shnei shvilin, and bik'a (10a).]
1. Tosafot (9b, s.v. heinu; RD 62 s.v. ve-amru)
            The case of teisha chanuyot implies that safek is prohibited. (We shall ignore the difference between kavua and rov, which was the topic of Shiur #16.) Shtei kupot, however, implies that we can be lenient, since bedikat chametz is a rabbinic obligation. Tosafot solve this contradiction by adjusting the cases of the gemara to ensure that they follow the well-known general rules of safek. Accordingly they explain that the case of teisha chanuyot deals with the issue of EATING chametz - a BIBLICAL prohibition - and we rule "safek de-oraita le-chumra." Shtei kupot, though, relates to SEARCHING for chametz - a RABBINIC injunction - and we rule "safek de-rabbanan le-kula."
            The difficulties with the approach of Tosafot are obvious. There is no indication in the gemara that the topic of each case is different. Choosing EATING chametz as the topic of teisha chanuyot is especially difficult, as R. David points out, since it makes the details of the case irrelevant - why mention the mouse?
            In a similar vein to Tosafot, the Ba'al HaMa'or suggests that teisha chanuyot relates to BITTUL chametz, which is a BIBLICAL requirement, rather than to bedikat chametz, which is rabbinic. The Ba'al HaMa'or then raises the possibility of an additional safek: Perhaps the mouse consumed the chametz. If a double safek indeed exists (did the mouse bring chametz into the house, and, even if he did, perhaps he ate it) then even for a biblical prohibition the doubt is resolved "le-kula" (sfek sfeika). Why then do we require bittul? He answers that the sugya is according to R. Zeira (above), who holds, according to the Ba'al HaMa'or, that a mouse DEFINITELY leaves over and does not wholly consume the bread in its mouth. Thus, there is only ONE safek (whether the mouse brought chametz into the house) and we follow the double rule of "safek de-oraita le-chumra, safek de-rabbanan le-kula." (Checking back to the previous sugya, you will remember Rashi's explanation that R. Zeira maintains that a mouse MAY leave over; i.e., there IS a safek. The Ra'avad immediately attacks the Ba'al HaMa'or for this "misunderstanding" of the previous sugya.)
            R. David (s.v. ve-tzarikh) offers another possibility how to eliminate this second safek (i.e., perhaps the mouse ATE the chametz). Both the safek whether the mouse BROUGHT in chametz, and, whether he subsequently ATE it, although temporarily distinct and referring to different occurrences, are logically identical and may be subsumed in one logical category of safek: Is the chametz IN the house or not. In other words, PRACTICALLY the second safek DOES exist (as opposed to the claim of the Ba'al HaMa'or). LOGICALLY, however, it is NOT a SEPARATE safek. Therefore, the principle of "sfek sfeika le-kula" does NOT apply.
             R. David suggests the following argument as an archetype of this principle: In every one of our cases, we could argue that maybe the mouse LEFT the house with the chametz, and even if he did not, maybe he consumed it. Thus, every case could be considered one of sfek sfeika, a conclusion that is clearly incorrect. In truth, there is only one safek with two variant means of reaching the same logical possibility that the chametz is not in the house.
            R. David subsequently continues to treat our case as a sfek sfeika, offering alternative explanations why it is not decided "le-kula". Therefore, we must understand why in fact this comparison is not valid. The answer, I think, is obvious: In his archetype case, the two possibilities are two different means of reaching the same logically significant goal - there is no difference for us if the chametz left through the front door or through the mouse's digestive tract. The same would be true if I suggested all of the following possibilities - maybe the mouse took the chametz out through the front door, or the back door, or he threw it in the fire, or ate it, or gave it to his girlfriend and she ate it. We are artificially dividing up one halakhic safek into a number of different cases. In our case, however, the two possibilities are whether the mouse brought the chametz INTO the house, and whether he took the chametz out (by eating it). While the final halakhic result is whether we need worry about the presence of chametz in the house or not, this is dependent on two distinct halakhic possibilities which are not merely artificial distinctions within a larger halakhic possibility.
            (The need that the two possibilities of a sfek sfeika be independent possibilities is universally accepted; it is not always so clear, however, whether this condition is being met. There is a great deal of discussion of this point among the commentators, which is one of the reasons why sfek sfeika is a difficult principle to apply halakha le-ma'ase).
2. Ra'avad (4b -5a on the Rif; RD 63, s.v. Veda'at)
            The Ra'avad introduces a new principle, which distinguishes bedikat chametz from all other cases. He claims that although bedikat chametz is only mi-de-rabbanan, in cases of safek we rule le-chumra as "bedikat chametz was enacted (specifically) for cases of safek." The original rabbinic decree that one must search the house for chametz was not because we KNOW for certain that chametz is there. Rather, the POSSIBILITY of chametz in a house requires that it be searched, so that this possibility be eliminated. In other words, one safek in bedikat chametz is the condition of obligation, as "vadai" is the condition of obligation or prohibition in other cases. To put it even clearer (provided if you read this line slowly), safek chametz in a house, vadai obligates you to search it. Hence, one safek alone does not eliminate the need to do bedika (which is the only topic being discussed in all the cases under review).
            The Ra'avad distinguishes, however, between different kinds of safek. If something has definitely come into the house, but I do not know what, I am obligated to perform bedika. If there is a safek if anything has come into the house, then I apply the principle of safek de-rabbanan le-kula. (I would like you to see where this is in the Ra'avad - we are on 5a [Rif pagination] lines 10-16).
            The reasoning is clear. The obligation to do bedika applies to a "dirty house," one whose cleanliness (from chametz) has not been established. My house, in which I lived and ate all year, is in this category. Once I do bedika, the house is "be-chezkat baduk," meaning it is presumed clean. If a mouse is roaming around outside, but I have not seen it enter, the presumption of "cleanliness" remains, and even though a POSSIBILITY that chametz might have come into the house exists, safek de-rabbanan le-kula. However, if the mouse has definitely entered the house with something in its mouth, even though the statistical possibility that there is chametz in the house is identical to the previous case, the cleanliness of the house has DEFINITELY been disturbed, so that the house is back in the state it was before I performed bedika. Hence, I am obligated to search in order to re-establish the status of a "clean" house.
            The second case, where I am not obligated to perform bedika, is called "safek al" (a doubt if the mouse entered), and is the case of shtei kupot, as well as shnei shvilin and bik'a.
            One problem still remains. What about the second safek - that the mouse may have eaten the chametz, assuming it was chametz, after he entered the house. Since in principle safek de-rabbanan le-kula applies to bedikat chametz, this should eliminate the need to search the house. In fact, this should combine with the safek whether the mouse brought in chametz or matza to form a sfek sfeika, so that even bittul, which is a biblical obligation, should not be required. The Ra'avad answers and says: "We are dealing with a case where there was not sufficient time for the mouse to eat it." The Ra'avad seems to be limiting the ruling in the gemara to a specific case where the second safek does not exist. This is how R. David interprets the Ra'avad (RD 63, s.v. ve-ata). 
            In effect, the Ra'avad's solution of the second problem is very similar to that of the Ba'al HaMa'or (who established our gemara according to R. Zeira only). Obviously, this is a very difficult solution. If there has not yet been enough time for the mouse to eat the chametz, we can just wait a few more moments and then there will be an added safek, which should eliminate the need to search the house.
3. Ramban (RD 64, s.v. u-le-inyan ha-sugya)
            R. David cites the opinion of the Ramban, who accepts the Ra'avad's first point (that bedikat chametz was enacted for a safek) but advances an alternate explanation why the safek that the mouse ate it does NOT serve as a second safek thereby eliminating the need for bedika: Once the positive obligation to do bedika is in effect (because of a safek that chametz entered) one cannot fulfill one's obligation except by actually searching. In other words, one is obligated to eliminate the possibility that the chametz is there. Thus, EVEN THOUGH HAD WE CONSIDERED THE SECOND SAFEK ORIGINALLY, WE WOULD NOT HAVE OBLIGATED THE HOUSE TO BE SEARCHED; now that the house became obligated (at the time when the mouse entered) only a vadai cleansing (i.e. elimination of the possibility that chametz is present) is sufficient to satisfy the rabbinic enactment of bedikat chametz.
            Examination of the text of the Ra'avad, I believe, will convince us that this is, in fact, the intention of the Ra'avad himself when he wrote that there has not been enough time for the mouse to eat the chametz. The use of the word "kegon" (for example) seems to imply that the Ra'avad attributes the case to where there has not yet been enough time, as opposed to other cases. However, the continuation of the sentence makes it clear that he is in agreement with the Ramban. Since originally there was a period when the mouse could not yet have eaten it, even afterwards the obligation is not annulled until a bedika is performed. This is the entire argument in the Ra'avad:
            We shall answer by saying that it is a case where there
            was not sufficient time for the mouse to eat it, and we
            derive that he must perform bedika and he may not wait
            until it will be eaten. Since the house has been
            established before him as one that requires bedika, it
            returns to the status of a house that has not been
            searched... and safek akhila is not a leniency, unlike
            the houses of goyim (9a) and (the ditch with) the
            embryo (9b) where there was not a particular hour where
            it was established in our presence that there was a
            corpse there.
            I think it is clear that the Ra'avad mentions the fact that the mouse has not had an opportunity to eat the chametz, only in order to establish a prima facie obligation IMMEDIATELY when the mouse is seen entering the house. Then he explains that this obligation, once established (at what R. David calls "the time of dragging"), cannot be annulled except by a bedika.
            (Note: I do not think that R. David and the Ramban misinterpreted the Ra'avad. The Ra'avad is famous for emending his commentaries. In an earlier version, only the expression "it is a case where there was not enough time for the mouse to eat it" appeared, and the Ra'avad in fact offered the explanation attributed to him by R. David. Subsequently, he changed the explanation, but did not cross out this line, which, in light of his new explanation, is not perhaps phrased in the most felicitous manner, merely adding on a fuller explanation. This is born out if we compare the text of the Ra'avad published in the "Katuv Sham" (The Ra'avad in the back of the gemara was extracted from the "Temim Dei-im", a collection of responsa by the Ra'avad and other medieval authorities. Some seventy years ago, a manuscript of the Hasagot HaRa'avad was found and published under the title "Katuv Sham" with some significant changes, as well as whole sections not found in the original printed version). The corresponding section reads as follows:
            We shall answer by saying that SINCE THE PRESENCE OF
            CHAMETZ IS ESTABLISHED THERE, it is like a house that
            has not been searched... unlike the houses of goyim....
            This is a later version, where the Ra'avad has finally used an eraser to eliminate the misleading opening sentence.)
            To sum up this solution: Bedikat Chametz has two distinguishing marks, relative to other obligations and prohibitions:
1. One safek (within the house) engenders the obligation, even though it is de-rabbanan;
2. Once established, the obligation to check cannot be annulled by a further safek.
            R. David (65 s.v. aval tzarich) principally attacks the second point and asks: "Why should the degree of obligation of bedikat chametz be frozen at the 'time of dragging?' After a few moments a sfek sfeika will exist as to the presence of chametz in the house. Surely this should suffice to free one of the obligation to search - even if the question was originally asked at an earlier time? After all, sfek sfeika is treated leniently for all laws!"
            The only possible answer - or so he assumes - is that this is part of the special rabbinic nature of bedikat chametz. However, this would imply, that in our case, one would be obligated in BEDIKAT chametz but not obligated in the biblical obligation of BITTUL chametz - a conclusion that is inherently illogical according to R. David.
            I think the answer to this question is that the second peculiarity of bedika is not rooted in the fact that it is rabbinic per se, but in the fact that it is an OBLIGATION, a mitzva. Once the Sages decreed in a given situation that one is obligated to perform bedika, meaning that one is obligated to eliminate the chametz, merely lowering the statistical probability is not sufficient. It is true that had this low probability been presented originally, the Sages would not have created the obligation. However, once the obligation is present, and the obligation is to SEARCH, meaning to positively eliminate the possibility of chametz, positive and final elimination is required. This would be true of bittul (de-oraita) as well, if it were construed as a positive obligation. Apparently the Ramban considers bittul to be only a means to avoid the prohibition of having chametz (compare to the Ramban on bittul we discussed several months ago). Hence, it is indeed true that if the possibility of chametz now is negligible (sfek sfeika), there is no reason to do bittul. However, one must still do bedika since the obligation has come into force. This is not different from the obligation to perform bedika, even where the house has been thoroughly cleaned during the preceding week. Since it is a mitzva, I must perform it properly.
            R. David is also concerned about the way the continuation of the gemara proceeds according to the Ramban (and the Ra'avad). The case of teisha chanuyot requires bedika where there is one safek about the substance brought into the house. All the remaining three cases illustrate the opposite principle - that "safek al," where the house has a chezkat baduk, does not require bedika. Why are all three cases necessary, especially since it is the third one where the principle of "safek al" is explicitly formulated?
4. R. David (RD 66, s.v. ve-nir'eh)
            R. David elevates the first principle of the Ra'avad, that bedikat chametz was instituted specifically in a case of safek, to the de-oraita level and applies it to the mitzva of bi'ur chametz. Bi'ur chametz, he claims, is different from other prohibitions and obligations. The correct understanding of this mitzva is not that the existence of chametz requires me to burn or eliminate it, but that the doubtful status of the house requires me to ensure that all chametz be positively eliminated. Adding to what he explicitly says, I think he is claiming that the object of bi'ur chametz is not chametz, but the house. "Tashbitu se'or mi-bateikhem" means clean out your houses from any possibility of se'or or chametz. 
            Hence, the de-oraita obligation obtains with definite status (be-torat vadai) even if there is a safek if chametz is present. While all Torah obligations would obtain in a case of one safek (safek de-oraita le-chumra) that would only be be-torat safek. In R. David's words: "The Torah commanded us for other prohibitions only in regard to vadai, but because of the severity of Torah prohibitions the law is to rule strictly concerning sfeikot." However, the Torah refers to a case of safek when it commands bi'ur. The difference becomes apparent when there is a second safek. In all other cases, sfek sfeika is permitted. But in bi'ur chametz, the first safek does not count at all. Since the Torah positively commanded us to eliminate safek chametz, if there is one more safek, that becomes a case of safek de-oraita le-chumra. Hence sfek sfeika in bi'ur chametz is the equivalent of one safek in other Torah prohibitions. 
            Now, R. David adds a second principle: Wherever one is obligated to do bi'ur, one is automatically obligated to do bedika (even though bedika is rabbinic and hence the rule of safek de-rabbanan le-kula should have applied), since the nature of the rabbinic law of bedika is not an independent obligation, but a rabbinic instruction how to fulfill the obligation of bi'ur. In his language: "(In the case of one safek) one is obligated by the Torah either to search or do bittul (the two acceptable means of bi'ur), and the second safek (maybe he ate it) is treated as a first safek would be in a Torah obligation. Thus, he must search or do bittul. However, it is a rabbinic enactment to not rely on bittul, but to search every unsearched house." In other words, bedika is a restriction on the de-oraita of bi'ur, not an independent mitzva; hence, in effect, it follows the rule of safek de-oraita le-chumra.
            Continuing in the gemara, R. David (67, s.v. aval nomar) accepts the distinction of the Ra'avad and the Ramban between safek chametz (in the house) and safek al (where it is doubtful if the actually entered the house or not). The principle explained above applies only to the former. In fact, even the de-oraita obligation of bi'ur is not applicable if there is one safek al, since the house has a chezkat baduk. (Tosafot had raised this point, and that is why they explained the gemara as referring to EATING, and not to BI'UR, unlike the Ba'al HaMa'or, thereby eliminating the chezkat baduk. The unknown piece of food has no chazaka at all.) This explains the case of bik'a. But, in contradistinction to the Ra'avad, R. David (68, s.v. ve-zeh) argues that shtei kupot is NOT a case of safek al. Since something has gone into each house, there is a safek in each house. Hence, the de-oraita of bi'ur applies. However, here an exception is made to the second principle. Since there are TWO houses, and one is in any event before bedika, we can say IN REGARD TO THE RABBINIC OBLIGATION ONLY, that we refer the chametz to the unsearched home and the matza to the searched one. This ("tolin") is a special leniency applicable to rabbinic laws, and hence here one may rely on bittul without bedika. So teisha chanuyot is based on the double principle of safek in bittul-bedika; shtei kupot is a special distinction in that principle allowing "tolin" for the rabbinic obligation of bedika; bik'a is the principle of "safek al" and shnei shvilin is the application of "safek al" in a mutually contradictory manner.
            R. David (67, s.v. ve-hanachon) makes one further point. The scenario mentioned in the gemara (9b) just before the cases we have been discussing, is where one puts aside ten pieces and subsequently finds only nine (hereafter referred to as "ten-nine"). R. David is of the opinion that Rava does NOT require bedika in this case, whereas Rav Mari DOES (see Rashi). The Ramban claims that all the subsequent cases of the gemara are according to Rav Mari, since this case is based on the principle that bedika will not be annulled by a subsequent safek akhila. Rava apparently disagrees with this principle, and, therefore, does not require bedika, relying on the chance that a mouse ate the missing piece. R. David is not willing to allow any opinion to disagree with the principles he derived from the continuation of the sugya. He, therefore, advances a novel interpretation of Rava, restricting his permissible ruling only to the case of ten-nine. I leave this explanation for you to work out on your own. It is a very subtle point, and Rav Mari of course disagrees totally.
5. Summary - We have basically four possible opinions:
            1. Tosafot and the Ba'al HaMa'or - the obligations of bi'ur and bedika follow the usual rules of safek de-oraita le-chumra, safek de-rabbanan le-kula, and sfek sfeika le-kula.
            2. Ra'avad (as understood by R. David and the Ramban) - Bedika, even though it is rabbinic, is obligatory even where there is a only a safek chametz. This does not apply to a safek al.
            3. Ramban - agrees with the position of the Ra'avad (2), and adds a second peculiarity of bedika: If one is obligated in bedika at a certain time, a SUBSEQUENT safek will not annul the obligation.
         4. R. David - a) Agrees with the first principle of the Ra'avad and the Ramban, but applies it even to bi'ur de-oraita - one does not destroy chametz, one cleans out one's house. Hence, sfek sfeika is equivalent to a usual single safek and is le-chumra.     b) Bedika applies automatically wherever bi'ur applies. Hence, in effect, sfek sfeika is le-chumra even for bedika.
Next week's shiur by R. Gigi is on the topic of Kesher Aniva.
1. Pesachim 11a "ve-khol heikha de-lo badil minei... be-kshira lo michlefa."
2. Shabbat 113a Mishna "Koshrin de'li" and the gemara.
3. Shabbat 111b Mishna "Ve-eilu" and the gemara. Rashi s.v. ve-eilu.
4. Rambam Hilkhot Shabbat 10:1 & 5
5. Eiruvin 97a "ve-hashta de-tani... ke-ein keshira didhu."
6. Sukka 33b "ve-ein mema'atin be-yom tov... bechada". Rashi.
A. What is the definition of a kesher as opposed to aniva?
B. What is the basis of the argument between:
- R. Meir and the tana kama?
- Rashi and the Rambam?
- R. Yehuda and tanna kama?