The gemara is trying to prove that maamar is not a form of erusin, and brings a proof from the fact that the shomeres yavam can annul her nedarim, but a na'ara hameorasa needs both the father and the husband. But what is troubling about this proof is, what does this have to do with us, are we only talking about a case of na'ara, a very limiting application of the sugia? Tosafos deals with the issue and in fact seems to suggest that in the end this is the only application.

Why can't you read the gemara in the following manner (I am wondering if you've ever seen such a suggestion):

If we treat maamar as erusin, then in our case where she has a zika from two people - and one with some form of erusin (i.e. if we consider ma'amar similar to erusin), why don't we compare this to the case of na'ara where she has a tie to both her father and to the husband and in that case both are required for hafara. And since we see that the yavam alone can be mefer, that is a proof that she has the status of a nesu'a! And, maybe this is exactly what the gemara is saying when it answers - that in fact what the gemara meant was that the yavam can be mefer with shutfus. That the yavam can be mefer with the shutfus of the other yavam (not with the girl's father as rashi reads it, which again would have very limited application). Would love to hear your thoughts about this radical reading of the gemara, but one which to me makes a lot more sense than the alternate reading which doesn't have much to do with our case.

Mark Wagshul, Stamford, CT