Statistics is part and parcel of Halacha. We can eat from a cow without checking its kidneys or liver because of statistics, we can fly on a plane high in the sky over an ocean because of statistics, etc. The statistical odds of Eliyahu coming today seem vastly more remote than the odds of a cow being shected this afternoon having a defective liver, or that tonight's flight will crash, and yet the Gemora says we should not burn sofek terumah because maybe Eliyahu will come - a Halachic decision that seems to ignore statistics. I would appreciate your explaining when statistics does and does not play a role in Halacha.
Avrohom Meyer Kohn, Los Angeles, CA USA
The Gemara means to say that at some time in the future Eliyahu may come and be Metaher the Terumah (Rashi 15a). The "maybe" does not refer to Eliyahu's coming-we know he will come (Malachi 3:23)-but to the possibility that he will rule on the Terumah that it was pure.
One may still ask from the Gemara on 13a, where Rebbi Elazar Ben Yehudah said to the Chachamim that they should forbid the burning of Terumah Teluyah on the day before Erev Pesach when Pesach falls on Shabos, since Eliyahu may come and rule that it may be eaten. There the Chachamim must be asking that Eliyahu may come before Sof Zeman Achilas Chametz. However we can answer that there Rebi Elozor was telling the Chachamim that if they prohibit burning Terumah for all the reasons mentioned there even though they are highly unlikely, then they should also worry for Eliyahu.
In Halachah we rely on statistics based on the Posuk of "Acharei Rabim le'Hatos". Statistics can be considered a Rov because, for example, there is no reason for us to assume that any said cow should be different than other cows of which we know the majority are kosher. As far as Eliyahu is concerned we know that he will come one day. Our question is whether this day is the day. There is no Rov to clarify this, as we have no reason to assume that today will be the same as previous days.
Dov Freedman
Thank you very much for your comments.
However, going back to my example, virtually none of the cows that will be shected next week will have their kidneys or livers checked. The cows will be assumed to be kosher (after their lungs, and only their lungs, are checked) because someone collected statistics in the past and noted that in the past, a majority of cows have healthy enough kidnes and livers, and so a Rov was established from the past that now affects the future, and so future cows are now kosher without any checking of the kidneys or livers. Using the exact same methodology, next week's sofek terumah should be burned without concern for Eliyahu based on the assumption that Eliyahu will not come next week, based on the past track record. What is the difference? If the past can create a Rov when it comes to future cows, why can it not do so when it comes to Eliyahu?
Avrohom Meyer Kohn
Firstly, we explained that the Gemara does not mean to say that we must worry for Eliyahu's arrival on a specific day, but that we must worry for his arrival at some time in the future. Since we know without doubt that he will come one day it is not possible to use Rov to discount this happening. Consequently we can conclude that the Gemara never worried for something which was statistically unlikely.
Secondly we added that it may be possible, Halachically, to consider the possibility of Eliyahu arriving on a specific day, even though this may be considered unlikely by a statistician. This is because Rov is not a method of determining Halachic questions with the use of statistics, as we see from the fact that the Torah says Kol Kavuah k'Mechtzah al Mechtzah. Rather it is a Din Torah with its own rules. So something may be statistically unlikely but still not ruled out by Rov. (See Sha'arei Yosher (3:1) for a lengthy analysis.)
The Rov which would be relevant in our case (and in the case of the cows) is called a "Ruba d'Leisa Kaman", in other words a Rov that is dictated by nature and not by adding up a certain number of samples that are in front of us. The logic for Rov in such a case is that since all cows are created by the same method, and we know that most cows created are not Treifos, we can assume that any specific cow is not a Treifah. (It may be noted that this remains true even if a person eats pieces of many cows at once in a way that statistically he may be assumed to be eating a Treifah.) However in the case of Eliyahu, what we need to know is, if this day is the same as the other days on which he did not appear. We have no reason to assume that it will be the same since we know that one day he will appear. Since we do not know that this day will be the same, nor do we have any reason to allow us to assume that it will be the same, we cannot use other days to dictate what we should expect to happen on this day. The fact that a statistician would consider his chances of appearing as remote does not affect our judgement.
Dov Freedman
I think that what Rabbi Freedman it is trying to say is that this is a Chazakah ha'Asuyah Leshishtanos - a Chazakah that we expect to change at any moment. This can be compared to a the Chazakah that a 12 yr. old boy will not be growing Simanim, since he has not had them until now. Such a Chazakah is not a valid Chazakah, Halachically (see Kidushin 79a, Rashi DH Hacha and Tosfos DH Mi).
M. Kornfeld