Getting back to the world of rov, last post (which seems to have been written eons ago) explained that if a person bumped into a pile of 9 frogs and one sheretz, without any other information we have a right based on the probabilistic halacha of rov to assume that a frog and not a sheretez was touched and the person is tahor. However, if a Navi or bas kol told us that in fact the one sheretz was touched, the person should be tamei. Probability cannot trump the testimony of the Navi as to what actually occurred. Why then do the Chachamim reject the evidence of the bas kol which declared (Bava Metziya 59) with certainty the halacha follows R’ Eliezer and instead rely on the probabilistic evidence of rov and pasken like the dissenting majority?
The answer posed in a few of the comments is that the question of what was touched – a frog or sheretz – is a question of fact, of metziyus, of that which can be measured concretely and tangibly, if we just had more information. The bas kol or Navi simply provides that extra factual information. The question of who halacha should follows is not an objective quantifiable or factual issue. Perhaps by definition halacha simply is the consensus of the majority. As R’ Elchahan puts it in Koveitz Divrei Sofrim, there is a difference between “holchin achar harov”, the probabilistic rov of the first sort, and “hilchisa k’rabim neged hayachid”, the rov of halachic verdict of the second sort.
The answer is neat, but it’s perhaps too good. The only source for the principal of rov in the Torah is the pasuk of “acharei rabim l’hatos”, which appears in the context of psak of beis din. If probabilistic rov and the rov of beis din’s verdict or halachic conclusion are two different animals, how does the same pasuk serve as the source for both? (Compare this question with R’ Chaim’s kashe that was discussed in this post.)
More on this topic in the future bl"n.