Yevamot 31a (לא.) — the Talmud in English

Rava said to him:

Since you require

“the rival wife of an erva” to undergo

ḥalitza—

it is obvious that

the betrothal of the erva is not a valid betrothal, and the reason her rival wife may not perform yibbum

is merely a stringency,

and she is not in fact the rival wife of an erva.

Therefore, people will not come to err if he himself betroths her sister, or if another person comes and betroths her.

Abaye said to him:

If so,

since it is possible to institute a stringency without fearing that it will lead to a mishap, because people see the distinction between ḥalitza and yibbum,

with regard to divorce as well, let the Mishna teach

in our Mishna “a case of uncertainty with regard to divorce,” in a manner similar to “a case of uncertainty with regard to betrothal,” such as where he threw the bill of divorce and it is uncertain whether it was closer to him or closer to her,

and let it require ḥalitza

as a stringency.

And there is no need to fear that the stringency will lead to leniency, as people might mistakenly infer from the fact that she performs ḥalitza that the divorce of the erva was a complete divorce; for if it were not a complete divorce, she would be the rival wife of an erva and would not require ḥalitza.

And as a result of this error—that they will think her divorce was a complete divorce—they will come to stumble: If, after this divorce, one man betroths her, and afterward that same man goes and betroths her sister, people will think that the betrothal of her sister is not valid, since they will assume that his betrothal took effect with regard to the first sister.

But in truth, this woman is not divorced, and his betrothal of her is not valid, whereas the betrothal of her sister is valid. Consequently, as a result of this error, we would permit a married woman to the world.

There is no need to fear this, since

people know,

from the fact that her rival wife does not perform yibbum,

that

her ḥalitza is merely

a stringency,

and the divorce of the erva is not a valid divorce!

And why did the tanna not teach in the Mishna “a case of uncertainty with regard to divorce,” similar to “a case of uncertainty with regard to betrothal”? Rava said to him: The Sages were concerned that

if you say

as a stringency that the rival wife of an erva in a case of uncertain divorce

performs ḥalitza,

they will come to say that she also

performs yibbum.

And they will come to perform yibbum with the rival wife of an erva, thereby causing them to transgress a prohibition punishable by karet.

For in truth, we leave the erva who was divorced with an uncertain divorce in her presumptive status, that she is not divorced but remains a married woman. Thus, the rival wife is the rival wife of an erva, and she is forbidden to the yavam by the karet-level prohibition of a brother’s wife.

Abaye said to him:

Here too,

with regard to uncertain betrothal of the erva, you should be concerned that

if you say that she performs ḥalitza,

they will come to say that she also

performs yibbum!

For it is possible that her betrothal was valid, making her the rival wife of an erva!

Rava said to him:

Let her perform yibbum; there is nothing wrong with that,

since

she is presumed

not to be the rival wife of an erva

on the basis of her presumptive status.

Abaye raised another difficulty against Rava:

We learned later: If one was married to two women, a non-relative and his niece, and

the house collapsed upon him and upon

his wife who was

his niece.

And both of them died.

The non-relative, who had been the rival wife of his niece, remained, and she now falls before the yavam for yibbum,

and it is unknown which of them died first:

Perhaps the husband died first and his niece died afterward, in which case the niece fell for yibbum together with the non-relative. Since the niece is the yavam’s daughter and is an erva forbidden to him, the non-relative, her rival wife, is exempt from both ḥalitza and yibbum.

Or perhaps the niece died first and her husband afterward. In that case, at the time of the brother’s death, the non-relative was not the rival wife of the yavam’s daughter, since his daughter had died before his brother died; and the status of a rival wife of an erva applies only at the time of death. Therefore, she is fit for yibbum.

Since the matter is uncertain,

her rival wife performs ḥalitza but not yibbum.

And

why

does she perform ḥalitza but not yibbum?

According to your position,

here too let us say: This woman,

the non-relative, before the house collapsed upon her husband and her rival wife,

was presumed to be permitted to the world,

like the rival wife of an erva.

And because of the uncertainty,

that perhaps her rival wife died before her husband,

you are coming to forbid her.

Leave her in her presumptive status, that she is permitted to marry outside the levirate bond, and

do not forbid her because of the uncertainty!?

And if you would say that here too

they decreed upon her

as a stringency

that she should perform ḥalitza but not yibbum,

this cannot be the answer, because

it is a stringency that leads to leniency.

For if you say that she performs ḥalitza—

she will perform yibbum!

And if she performs yibbum, you will thereby have violated the prohibition involving the rival wife of an erva, which is punishable by karet!

Rava said to him:

Divorce, which is common,

as it is an ordinary occurrence,

the Sages issued a decree concerning it

not to impose the stringency of requiring ḥalitza from the rival wife when the erva was divorced through an uncertain divorce, because of the concern that if you say she performs ḥalitza, they will err and say that she also performs yibbum.

But

a collapse, which is uncommon,

as it is not usual for people to die in a collapse—

the Sages did not issue a decree concerning it

not to impose the stringency of requiring her to perform ḥalitza because of the concern that if you say she performs ḥalitza, they will say that she also performs yibbum. Rather, they required ḥalitza as a stringency.

Alternatively,

there is another distinction:

With regard to divorce, the erva is present,

as the erva is before us,

so that it is evident—

everyone sees that there is an erva here—

and nevertheless

despite this

her rival wife—

you require her to undergo ḥalitza.

People will say,

people:

the Sages established the matter of the bill of divorce,

the Sages determined

that

this bill of divorce of the erva

is a fully valid bill of divorce,

and her rival wife is not the rival wife of an erva; therefore, she does not leave without ḥalitza.

And

as a result of this error

they will come

they will come

to perform yibbum with the rival wife,

although in truth she is the rival wife of an erva.

But

a collapse—

did the Sages establish anything concerning a collapse?

Can the Sages know who died first in a collapse?

Therefore, if the Sages require her to perform ḥalitza, people will not say that she is not the rival wife of an erva and perform yibbum with her. Rather, everyone will know that they required ḥalitza only as a stringency, because it is impossible to know who died first.

The Gemara nevertheless asks:

But

since

with regard to divorce as well, did we not teach

that they imposed the stringency of requiring her to perform ḥalitza, and were not concerned that people would say, “If you say that she performs ḥalitza, [she also] performs yibbum”?

But we learned

in tractate Gittin [78b]:

If she was standing

his wife

in the public domain, and he threw to her

the bill of divorce in the public domain—

If the bill of divorce was

closer to her—

she is divorced.

And if the bill of divorce was

closer to him—

she is not divorced.

And if the bill of divorce was

equidistant between them—

she is divorced and she is not divorced.

And we say: For what halakhic purpose

what practical consequence does the ruling that when it is equidistant she is “divorced and not divorced” have?

On the one hand, we say that she is divorced—

For if he is a priest

her husband,

she is forbidden to him

to remain married to him, because of the uncertainty that perhaps she is divorced.

And on the other hand, we say that she is not divorced—

And if she is an erva

for the yavam,

and

the husband dies after the uncertain divorce—

her rival wife requires ḥalitza,

as perhaps the erva is not divorced.

And we do not say

leave the erva in her presumptive status that she is not divorced, so that her rival wife is certainly the rival wife of an erva, and do not impose the stringency of requiring her to perform ḥalitza, because

“if you say that she performs ḥalitza,

she also]

performs yibbum”!

And if so, this is difficult: Since with regard to uncertain divorce as well we find that they imposed the stringency of requiring ḥalitza, and were not concerned that “if you say she performs ḥalitza, she also performs yibbum,” why did our Mishna not teach a case of uncertain divorce analogous to uncertain betrothal—where he threw the bill of divorce and it is uncertain whether it was closer to him or closer to her—so that if the divorced woman was an erva to the yavam, her rival wife performs ḥalitza but not yibbum, and there is no concern that “if you say she performs ḥalitza, [she also] performs yibbum”?

And we answer:

it was stated concerning it,

concerning this Mishna:

Rava and Rav Yosef both said: Here,

the Mishna in tractate Gittin

is dealing with two sets of witnesses.

When one set

says that it was closer to her

the bill of divorce,

and

a set

says that it was closer to him

the bill of divorce.

And since there are two contradictory testimonies—one of which says that a divorce occurred and the other says that no divorce occurred—it is impossible to leave her in her presumptive status that she is not divorced. For a presumption clarifies an uncertainty where there are no witnesses, but it cannot clarify the uncertainty where there are witnesses who say that a change occurred in the presumptive status.

And although one set of witnesses contradicts the other, the matter nevertheless remains uncertain.

For the divorce of the erva is

the divorce of the erva

a Torah-level uncertainty,

and therefore the rival wife of the erva is required by Torah law to perform ḥalitza because of the uncertainty.

And this is not merely a stringency, such that we could say that she should not be required to perform ḥalitza because of the concern that “if you say she performs ḥalitza, [she also] performs yibbum.”

But

if

the Mishna here,

if we wish to establish it as referring to uncertain divorce, as in the case of betrothal, we would have to establish it

as involving one set

of witnesses, where one witness says that the bill of divorce was closer to him and a second witness contradicts him and says that it was closer to her. And one witness who is contradicted by a second witness does not prevent us from leaving the matter in its presumptive status. Therefore, the rival wife of the erva should be left in her presumptive status—that she is presumed permitted to the world.

And since

it is

such an uncertainty with regard to divorce merely

“a rabbinic uncertainty,”

since by Torah law we leave her in her presumptive status of being permitted to the world, we should not impose the stringency of requiring her rival wife to perform ḥalitza, because of the concern that “if you say she performs ḥalitza, [she also] performs yibbum.”

And

therefore, we did not teach such a case of uncertain divorce in our Mishna.

And this is the distinction between betrothal and divorce.

For when there is uncertainty with regard to the betrothal of the erva because one witness contradicts another, we leave the erva’s rival wife in her presumptive status of being permitted to the yavam and say that she is not the rival wife of an erva. Nevertheless, we impose the stringency that she perform ḥalitza but not yibbum. There is no need to fear that this will lead to the mishap of “if you say she performs ḥalitza, [she also] performs yibbum,” because according to the fundamental law she may indeed perform yibbum.

But with regard to uncertainty concerning the divorce of the erva, we leave her rival wife in her presumptive status of being permitted to the world, and we should not impose the stringency of requiring her to perform ḥalitza, because this could lead to the mishap of “if you say she performs ḥalitza, [she also] performs yibbum.” Here, yibbum would involve a karet-level prohibition, since the presumption indicates that she is the rival wife of an erva! And the Gemara proves:

But from where do you know that the Mishna here

must be established

as involving one set

of witnesses, one contradicting the other?

It is analogous to betrothal—

just as with betrothal

the case is

with one set, so too with divorce it is with one set.

But this is difficult:

With regard to betrothal itself—

from where do you know that it is with one set

of witnesses, one contradicting the other?

Perhaps we are dealing with two sets of witnesses

that contradict one another!?

And the Gemara answers:

If we are dealing with two sets of witnesses

that contradict one another, why should the rival wife perform ḥalitza and not yibbum?

Let us leave her in her presumptive status that she is permitted to the yavam, and if so,

let her perform yibbum; there is nothing wrong with that!

Rather, it must be that we are dealing with one set consisting of two witnesses, in a manner where one witness contradicts the other. Therefore, there is concern that perhaps the erva was betrothed, and consequently her rival wife performs ḥalitza but not yibbum.

The Gemara is puzzled: How can it be said that when two sets of witnesses contradict one another, she should perform yibbum and there is nothing wrong with that!?

The Gemara’s question is in the form of a dilemma: What is the law in every case of contradiction between two sets of witnesses [briefly called “two against two”] where there is a presumption?

Is “two against two” in the presence of a presumption a Torah-level uncertainty, meaning that the uncertainty remains on the Torah level and the presumption does not decide it, because where witnesses contradict one another, the presumption is not determinative?

As the Gemara said above, Rava and Rav Yosef established the Mishna in Gittin as referring to two sets of witnesses who contradict one another, “for it is a Torah-level uncertainty.”

Or, “two against two” in the presence of a presumption is a rabbinic uncertainty: By Torah law, we leave the matter in its presumptive status, but the Sages imposed a prohibition because of the uncertainty.

1. If we say that “two against two” is a Torah-level uncertainty, as Rava and Rav Yosef maintain, this is difficult:

For

there are

two

witnesses, and they say

that the betrothal money was

closer to her,

so the erva was betrothed and her rival wife is the rival wife of an erva, forbidden to the yavam by a karet-level prohibition,

and

if

you say, “Let her perform yibbum and there is nothing wrong with that”!?

For when two sets of witnesses contradict one another, the uncertainty remains on the Torah level, and there is no place for a presumption to decide the matter when the uncertainty before us results from a contradiction in the testimony of the witnesses!

2. It should follow that “two against two” is only a rabbinic uncertainty, and that by Torah law we rely on a presumption even in a case of “two against two.” Therefore, this too is difficult:

Furthermore, in a case of two sets of witnesses

with regard to divorce,

also

it is difficult: Why do you require ḥalitza

of

the rival wife because of the concern that perhaps the erva was divorced?

For in every case of “two against two” where there is a presumption,

it is only a rabbinic uncertainty,

since by Torah law we rely on the presumption where two sets of witnesses contradict one another!

For

by Torah law

we say: Set the two

witnesses who say that she was divorced

against

against

the two

witnesses who say that she was not divorced,

and the woman

the rival wife of the erva,

leave her in her presumption

of being permitted to the world!

Why, then, do you forbid her on the basis of the “rabbinic uncertainty” concerning the erva’s divorce?

And the proof that “two against two” is a rabbinic uncertainty, and that by Torah law we rely on a presumption, is

as is the case

as we find

with regard to the property of Bar Shatya,

the property of a man called Bar Shatya, who was at times mentally incompetent and at times mentally competent.

Bar Shatya sold his property,

he sold his real estate.

Two witnesses came,

two witnesses came

and said: “When he was of sound mind,

mentally competent,

he sold

his property, and the purchaser acquired it.

And another two witnesses came

other,

and said: “He sold it when he was insane,

and the sale of an insane person is not a valid sale, so the purchaser did not acquire the land!

And Rav Ashi said: Set the testimony of the two against the testimony of the other two,

place the testimony of one pair opposite the testimony of the pair opposing them.