The Gemara is prone to challenge the specific language used by a rabbi in order to sharpen our understanding of the statement and its implications. Sometimes, the analysis reveals a need for clarification and the statement is subsequently amended. Other times, it turns out the original language is stronger than it first appears to be.
We see an example of the latter on today’s daf:
Rav Huna said that Rav said: A widow who is waiting for her yavam and then died before he could take her in levirate marriage, the yavam is permitted to marry her mother.
When a man dies childless, a new status falls on his brother (the yavam) and his widow (the yevama). The question raised by Rav’s teaching is this: How substantial is this new status? Is it akin to an engagement, which in Jewish law is a substantial enough bond that it requires a formal Jewish divorce to dissolve? Or is it something less than that?
Rav’s teaching suggests that he believes the pre-marital bond between a yavam and a yevama is not a substantial one. What’s the logic here? Well, a man is prohibited from marrying both a woman and her daughter — either simultaneously or even after one of the women has divorced or died. If the connection between the yavam and yevama is a substantial one (i.e., akin to an engagement), it would seem that the man would be forbidden from marrying the woman’s mother if she dies before the yibbum marriage is effectuated. But since Rav permits this, he must believe that the relationship between a yavam and a yevama, prior to their marriage, is not a substantial one.
The Gemara raises an objection about this inference: If Rav truly thought the bond was not substantial, why wouldn’t he have just said so?
The Gemara goes on to explain that if he had, his position in the case of a woman who dies before her yibbum marriage can be performed would not have been clear. Why not? Because one could argue that the yavam-yevama bond is insubstantial only in cases where there are two surviving brothers. This is because when there are two remaining brothers, either could fulfill the obligation to perform yibbum and so neither is certain to be the yavam.
But in a case when there is only one surviving brother, we know he is definitely going to be the yavam, so we might think the link between him and his widowed sister-in-law is substantial. Had Rav merely stated his position on the pre-marital bond, it would not have been clear that he believed even a single surviving brother is permitted to marry the mother of his deceased brother’s wife.
You can probably see where this is going now. The Gemara next wonders why Rava didn’t say that explicitly — that the pre-marital levirate bond is insubstantial even in the case of one brother. Turns out, that has problems of its own — namely, one might have then inferred that Rav believes the bond is so insubstantial that the brother is permitted to marry his yevama’s mother even while the yevama is still alive, potentially in a case where the brother chooses to marry the widow’s mother in order to get out of performing yibbum with her daughter. But we know that’s not allowed — the yibbum obligation cannot be canceled that way.
And so, the Gemara concludes, Rav’s initial language turns out to be stronger than we first thought. By choosing to teach as he did, Rav embeds his full opinion in his statement, even though we need to do a little logical legwork (with the Gemara’s help) to figure it out. In short, Rav’s position is that the levirate bond is not substantial and that even a single surviving brother may marry his sister-in-law’s mother — but only after she has died, and only in cases where yibbum has yet to take place.
The Gemara next turns its attention to Rav Yehuda, who takes the opposite position from Rav, prohibiting the yavam from marrying his yevama’s mother if she dies before yibbum takes place. Could he have made his position clearer by using different language? That’s the subject of tomorrow’s daf.
Read all of Yevamot 17 on Sefaria.