Terumah (5765)

From the book Mida Tova: Articles on the Hermeneutical Principles by Rabbi Michael Avraham. Translated from Hebrew using gpt-5.4 (reasoning_effort=high, batch API).

---

With God’s help — Midah Tovah — Sabbath eve of Parashat Terumah, 5766

Questions

1. What is a mufneh gezerah shavah (a verbal analogy based on a textually superfluous term)?
2. Must there be a connection between the superfluous term and the word from which the gezerah shavah is actually derived?
3. Does a mufneh gezerah shavah require a tradition?
4. Why is a gezerah shavah stronger than a hekesh (an analogy based on textual juxtaposition)?
5. When a gezerah shavah is said to be “mufneh on both sides,” must it really involve both sides?
6. Are there gezerah shavah between different words?

The Hermeneutical Principle

Gezerah shavah.

A. Summary of Last Year’s Article

“And you shall make the altar of acacia wood, five cubits long and five cubits wide; the altar shall be square, and its height shall be three cubits.”
(Exodus 27:1)

“And you shall make an altar for burning incense; of acacia wood shall you make it. A cubit shall be its length and a cubit its width; it shall be square, and its height shall be two cubits; its horns shall be of one piece with it.”
(Exodus 30:1-2)

“And he made the altar of burnt offering of acacia wood, five cubits its length and five cubits its width, square, and three cubits its height.”
(Exodus 38:1)

Rabbi Yehudah follows his own view, for he said: The altar that Moses made was large, as it is taught in a baraita: “Five cubits long and five cubits wide” — the words are to be taken literally; these are the words of Rabbi Yosei. Rabbi Yehudah says: Here it says “square,” and there, in Ezekiel 43, it says “square.” Just as there one measured from the midpoint, so here too one measured from the midpoint. And how do we know that there? Because it is written: “And the altar-hearth shall be twelve cubits by twelve cubits, square to its four sides.” But perhaps it means only twelve by twelve? Since it says: “to its four quarters,” this teaches that one measures from the center. And Rabbi Yosei? When he learned the gezerah shavah, he learned it with regard to height,

as it is taught in a baraita: “And its height shall be three cubits” — the words are to be taken literally; these are the words of Rabbi Yehudah. Rabbi Yosei says: Here it says “square” and there it says “square.” Just as there its height is twice its length, so here too its height is twice its length. Rabbi Yehudah said to him: But has it not already been said: “And the courtyard shall be a hundred cubits, and its height five cubits,” and so on? Is it possible that a priest would stand on the altar with the service in his hand and all the people would see him from outside? Rabbi Yosei said to him: But has it not already been said: “And the hangings of the courtyard and the screen for the gate of the courtyard, which are by the Tabernacle and by the altar”? Just as the Tabernacle was ten cubits high, so too the altar was ten cubits high. And it also says: “And hangings of fifteen cubits on each side.” Then what is the purpose of “five cubits”? It refers to the space from the lip of the altar upward. And what is the purpose of “and its height shall be three cubits”? It refers to the space from the surrounding ledge upward. And Rabbi Yehudah? When he learned the gezerah shavah, he learned it with regard to width.
(Babylonian Talmud, Zevachim 59b)

In the first verse cited above, Moses is commanded to make the altar of burnt offering, that is, the bronze altar, or outer altar. The plain sense of the verse instructs him to make the altar with an area of five by five cubits and a height of three cubits. The same appears in chapter 38, in the description of the actual construction. By contrast, in chapter 30 Moses is commanded to make the golden altar, that is, the inner altar or incense altar, and there the area is one cubit by one cubit and the height is two cubits.

Rabbi Yehudah and Rabbi Yosei disagree about the height of the altar of burnt offering. According to Rabbi Yehudah, “the words are as written,” that is, the height of the altar of burnt offering was indeed three cubits. According to Rabbi Yosei, it was ten cubits high — twice its width, on the basis of a gezerah shavah of “square”-“square” from the golden altar — and what is written in the verse, “its height shall be three cubits,” refers only to the portion from the surrounding ledge upward.

At the beginning of the sugya there is a parallel disagreement between Rabbi Yehudah and Rabbi Yosei, this time regarding the length and width, and there their positions are reversed. According to Rabbi Yosei, “the words are as written,” that is, the altar was five by five cubits. According to Rabbi Yehudah, the altar was ten by ten, by a gezerah shavah of “square”-“square,” though for him the source is the altar in Ezekiel’s Temple.

We noted that both tanna’im derive a gezerah shavah from the word “square,” and both thereby remove the verse from its plain sense. Both parts of the sugya begin specifically by detailing the view of the one who says “the words are as written,” and only afterward is the view presented that departs from the plain sense.

In the second part, after the two positions have been presented, Rabbi Yehudah’s attack on Rabbi Yosei likewise does not rely on the fact that the plain sense of the language supports his own view. Rather, it relies on a collateral consideration, namely the modesty of the priest performing the service.

All this clearly shows that the disputing tanna’im shared a common datum: the word “square” serves as the basis for a gezerah shavah, and this gezerah shavah removes the verse from its plain sense. They disagree only about what is removed from its plain sense: the height or the width.

We saw that there are two possible reasons for the a priori compulsion to perform a gezerah shavah and to remove the verse from its plain sense:

1. The tanna’im had before them a traditional datum that there is here a gezerah shavah that removes the verse from its plain sense.
2. There is some textual hint that instructs us to remove something in the verse from its plain sense, or that instructs us to derive a gezerah shavah from the word “square.” We noted that in a mufneh gezerah shavah tradition may perhaps not be necessary, since the availability is its basis. The plain sense of the Gemara in Babylonian Talmud, Niddah 22b-23a and elsewhere does not seem to support this. Below, in this week’s essay, we will establish the point more firmly.

In the continuation of the essay we examined several midrashim and saw, on the basis of a combination of considerations, that some assume the gezerah shavah here rests on mufneh, whereas others rest on tradition. We distinguished between the command verses, where one may assume that the text gives a minimal measure and not necessarily an exact and binding one; in that case the word “square” is not superfluous, and the gezerah shavah is not mufneh. But one can also understand the measure as exact and obligatory; then the word “square” is superfluous, and the gezerah shavah is based on mufneh. In the execution verses, after the measures of length and width have already been given, the word “square” is certainly superfluous.

In the second part of the essay we analyzed several halakhic (legal) expressions of this distinction. We cited Rabbi Aryeh Leib Malin, who proves from a contradiction in Maimonides’ rulings that there are two legal aspects to the altar: it is a place required for the validity of offering sacrifices, and it is also a vessel that forms part of the Temple building. With respect to building the Temple, the measurements are exact, and therefore the word “square” is superfluous even in the command. But with respect to offering sacrifices, the measurements are only minimal — and according to Maimonides perhaps not necessary at all; see our discussion carefully — and therefore the word “square” is not mufneh.

At the end of the essay we raised the possibility that the gezerah shavah is based on mufneh, but the superfluous expression is not “square” but rather “five cubits by five cubits.” The word “square” is needed in order to teach that, whatever the dimensions, the altar must be square; yet the gezerah shavah itself is derived from the word “square.” This is a major novelty, because it implies that the textual compulsion to perform the gezerah shavah need not be the availability of the very term from which the derivation is made. It is enough that the verse hints, by some sort of superfluity, that a gezerah shavah is to be performed, and we may perform it from another word — not necessarily the available one.

At the end of our remarks we dealt with the implications for our conception of the hermeneutical principles in general. We saw that if mufneh is only a textual hint that a derivation is to be made, then the derivation itself can take a variety of forms. The choice among them is apparently based on substantive similarity. In our case, the choice of the word “square” was made because of a substantive resemblance between the two altars, regardless of the textual compulsion that led us to expound.

B. Does a Mufneh Gezerah Shavah Require Tradition?

Introduction

In last year’s essay we encountered a special type of mufneh in gezerah shavah. We saw a case in which the available words are not the ones actually employed in the gezerah shavah itself. The logic of such a structure is clear: mufneh is no more than a textual hint to interpret, and once that hint exists, we apply it according to the exegete’s judgment and according to his reasoning. If our claims are sound, other forms of mufneh are also possible, serving as textual hints for carrying out a gezerah shavah. Seemingly there should be no limitation on the kinds of textual irregularities that can function as such hints. To be sure, if the irregularities are not located in the word actually expounded, it is not entirely clear how we know at all that the hint instructs us to perform specifically a gezerah shavah rather than a derivation by some other hermeneutical principle. Perhaps, in that context, this is simply the only midrashic route the exegete finds. In other words, the exegete’s own reasoning is involved here as well.

However, as we saw on the page for Parashat Miketz, 5765, some element of gezerah shavah is supposed to be handed down by tradition. It may be that the tradition determines that a gezerah shavah is to be performed here. On the other hand, we also raised the possibility that when some kind of mufneh is present — that is, an irregularity that hints textually that interpretation is called for — no tradition is needed in order for us to perform a gezerah shavah.

In this week’s essay we will deal with these two points: does mufneh eliminate the need for tradition? And what kinds of mufneh hint to us that a gezerah shavah is to be performed?

Does Every Gezerah Shavah Require Tradition?

As is well known, the accepted rule is that a person may not derive a gezerah shavah on his own, as stated in Babylonian Talmud, Pesachim 66a and parallels. Rashi on Babylonian Talmud, Sukkah 11b, s.v. “Lo Yalfinan,” explains this as follows:

“We do not derive” — they did not learn it; and a person may not derive a gezerah shavah on his own, unless he learned it from his teacher, and his teacher from his teacher, back to Moses our teacher, for perhaps these verses were not written in order to expound a gezerah shavah from them.

That is, according to Rashi the meaning is that in order to perform a gezerah shavah one requires an unbroken tradition from Moses at Sinai down to us. This is the accepted approach to interpreting this rule. Maimonides, however, in the second root of Sefer HaMitzvot, maintains that every law derived by one of the hermeneutical principles is of rabbinic status, and this apparently includes gezerah shavah as well. Therefore the author of Kena’at Soferim, on the second root, explains that according to Maimonides the meaning is that an ordinary person cannot derive a gezerah shavah; only the Great Court can do so.

We have already seen several times in the past, for example on the page for Parashat Miketz, 5765, and elsewhere, that the medieval authorities wrote that the tradition concerning a gezerah shavah does not necessarily transmit all its details. Sometimes the word from which the gezerah shavah is derived is what is transmitted to us; at other times the law derived by it is what is transmitted, and in such cases the exegete must complete the picture.

The Requirement of Mufneh

It is accepted that gezerah shavah is expounded only when the word is mufneh, that is, superfluous. Some derivations are mufneh on both sides, and some are mufneh on only one side. There are also a number of methodological differences between them, for example regarding the manner in which one may refute them, and so on. See on this the page for Parashat Ki-Tavo, near the end, 5765. In his book, Chernik discusses the development of the concept of mufneh and the special requirements attached to gezerah shavah; see the essay on Parashat Hayyei Sarah, 5766.

Does a Mufneh Gezerah Shavah Require Tradition?

In last year’s essay we noted that there are two possibilities that can hint to us of the need to perform a gezerah shavah: a tradition that determines that a certain verse contains a gezerah shavah, or the availability of a word. If availability is a sufficient hint, perhaps it can serve as an adequate alternative to tradition. The textual hint provided by availability can serve as the basis for performing a gezerah shavah, and the author of the Torah can rely on it and dispense with an oral tradition that would instruct us to expound a gezerah shavah in the verse under discussion.

The sugyot of the Talmud do not imply this, since on the one hand there is a general requirement that every gezerah shavah be mufneh, and on the other hand there is a general requirement that a person not derive a gezerah shavah on his own. It therefore seems that all gezerah shavah must be mufneh, and yet one is still not supposed to derive them without tradition.

However, there are remarks that qualify this picture. For example, the Shelah, in the section on the Oral Torah, in the chapter on gezerah shavah, asks: why is there a rule that one may not derive a gezerah shavah without mufneh, if all gezerah shavah were given to us as a tradition from Sinai? The tradition itself should determine where a gezerah shavah is expounded and where it is not.¹ Yet in light of what we said above, this difficulty is not hard, for we have already seen that there is not necessarily a tradition instructing us to derive a gezerah shavah from a particular place. Sometimes, for example, the law itself is transmitted to us by tradition, and we then find for it an anchor in a gezerah shavah. In such a case there is room for the instruction not to derive a gezerah shavah when there is no mufneh.

There are quite a number of midrashim from which it appears that the gezerah shavah is innovative and was not transmitted by tradition; see Entziklopedia Talmudit, entry “Gezerah Shavah,” notes 46-94.

Another example is what we saw in last year’s essay, and in the second point of the summary above: if there is mufneh, then the derivation is not based on tradition. We also saw that there are midrashim that understood the gezerah shavah in the exposition above in this way.

A possible solution to all these problems is that mufneh gezerah shavah do not require tradition, and the requirement of tradition applies only when the gezerah shavah is not mufneh. Below we will see several sources from which such an approach emerges.

Mufneh Eliminates the Need for Tradition: Rashi on Babylonian Talmud, Gittin 41b

In Babylonian Talmud, Gittin 41b, the Gemara discusses whether one can free half a slave:

The Sages taught: If one frees half of his slave, Rabbi says: he acquires his freedom, and the Rabbis say: he does not acquire it. Rabbah said: Their dispute is with regard to a deed. Rabbi holds: “She has not been redeemed, nor has freedom been given her” — the verse juxtaposes deed to money; just as money can redeem either all or half, so too a deed can free either all or half. But the Rabbis derive it from woman: just as a woman cannot be divorced halfway, so too a slave cannot be freed halfway. But with money, all agree that he acquires his freedom, for she is both redeemed and not redeemed. Shall we say that this is their dispute: one Master holds that hekesh is preferable, and the other holds that gezerah shavah is preferable? No. Everyone agrees that gezerah shavah is preferable, but here it is different because one can refute it: what applies to a woman, who does not go free with money, can you say the same concerning a slave, who does go free with money? Rav Yosef said: Their dispute is with regard to money. Rabbi holds: “She has not been redeemed” implies she is redeemed and not redeemed, whereas the Rabbis hold: the Torah speaks in human language. But with a deed, all agree that he does not acquire his freedom.
(Babylonian Talmud, Gittin 41b)

At one stage the Gemara explains the dispute between Rabbi and the Rabbis as turning on the question whether a hekesh is preferable to a gezerah shavah, or a gezerah shavah to a hekesh. It then rejects this and says that everyone agrees that gezerah shavah is preferable.

Rashi there explains the conclusion that gezerah shavah is preferable as follows:

Everyone agrees that gezerah shavah is preferable, for when it is mufneh for exposition it is as though it were written explicitly in the body of the verse itself, since the verse is superfluous. And even when it is not mufneh, one cannot refute it, since we learn it from Sinai, for a person does not derive a gezerah shavah on his own. Therefore it is preferable to hekesh.

It seems that his intention is to explain why gezerah shavah is preferable to hekesh, whether it is mufneh or not. The explanation is as follows: if the gezerah shavah is mufneh, then it is preferable because it is considered as though written explicitly in the verse itself.² The available word is superfluous at the level of the plain sense, and therefore its meaning lies on the plane of exposition. In such a case, the exposition itself is considered explicit in that word of the verse. And in a gezerah shavah that is not mufneh, Rashi explains, there is still an advantage over hekesh, because it is handed down to us by a tradition from Sinai, unlike a hekesh.³

Why does Rashi not explain with respect to a mufneh gezerah shavah as well that it is handed down to us by tradition from Sinai, instead finding himself compelled to treat the gezerah shavah as though written explicitly in the verse? It seems, at least prima facie, that Rashi holds that a mufneh gezerah shavah is not transmitted to us from Sinai, and that tradition is required only for a gezerah shavah that is not mufneh.

Mufneh Eliminates the Need for Tradition: Ritva on Babylonian Talmud, Bava Metzia 41b

We likewise find in Babylonian Talmud, Bava Metzia 41b, that the Gemara derives from the verse “Then the master of the house shall come near to God” that an unpaid bailee who claims that the object deposited with him was stolen, and witnesses then come and show that he lied and that the object is in his possession, must pay double like a thief — that is, he is one who makes a thief-claim. The Gemara asks whether an oath is also required in order to obligate him in double payment, or whether the claim alone suffices. It derives from a gezerah shavah of “putting forth the hand”-“putting forth the hand” that an oath is indeed required in order to obligate one who makes a thief-claim.

In the new Ritva there, s.v. “Ve’idakh,” the following difficulty is raised:

And if you ask: since we have already used one of the verses to teach that we do not say “it is enough for that which is derived by inference to be like that from which it is derived,” how can we return and expound it for a gezerah shavah, seeing that we require mufneh? Moreover, according to Rava, who expounds both as above, how can this gezerah shavah be derived? It may be answered that this gezerah shavah here is a received tradition, and in any such case we do not require mufneh at all, as I have explained in several places.

The Ritva asks how a gezerah shavah can be derived from this verse, when an earlier part of the sugya makes clear that it teaches us something else — namely, that we do not apply the principle “it is enough for that which is derived by inference to be like that from which it is derived.” If so, it is not mufneh. The Ritva answers that this gezerah shavah is a received Sinaitic tradition, and in a gezerah shavah that is received by tradition, mufneh is not required at all. He adds that he explained this principle in several other places in his novellae.

From the Ritva’s words it emerges that a gezerah shavah transmitted by tradition does not require mufneh. But his words clearly imply the converse as well: a mufneh gezerah shavah does not require tradition. For if it did require tradition, then all gezerah shavah would be transmitted to us by tradition, and there would be no gezerah shavah in the world that requires mufneh.

Thus, from Rashi and the Ritva we learn two things:

1. A gezerah shavah that is received by tradition does not require mufneh in the verse.
2. A mufneh gezerah shavah does not require tradition.

One may also formulate the matter differently: a mufneh gezerah shavah is likewise transmitted to us from Sinai, but the mode of transmission is through the superfluity in the verse itself, rather than through an explicit accompanying statement instructing us to perform a gezerah shavah. Such a gezerah shavah is transmitted from Sinai just as the Torah itself — and what is explicitly written in it, such as a mufneh gezerah shavah — is transmitted from Sinai.

C. Unusual Forms of Mufneh

Introduction

In last year’s essay we saw that there can be a situation in which the gezerah shavah is mufneh through one word, or several words, but the derivation itself is carried out by means of other words — not the available ones. In the previous chapter we saw that mufneh is required only for gezerah shavah that were not transmitted to us by tradition. This strengthens the claim that the role of mufneh is to prevent mistakes in gezerah shavah derivation; therefore, when there is tradition there is no need for mufneh, and vice versa. But this very principle also explains why one word can be available while the gezerah shavah is performed by means of another. All we need is a textual hint to prevent error, and therefore it need not come specifically by way of the superfluity of the expounded word; it can also appear in other forms. In this chapter we will see several such forms.

The Interrogation of Witnesses in Capital Cases

The Mishnah in Sanhedrin 40a states that witnesses in capital cases must be examined through seven formal interrogations, concerning the time and place of the event. The Gemara there asks what the biblical source is for this rule, and brings an exposition from the verses dealing with the idolatrous city, the individual idolater, and conspiring witnesses.

With regard to the idolatrous city it is written in Deuteronomy 13: “And you shall inquire, investigate, and ask thoroughly” — these are three interrogations. With regard to an individual idolater it is written in Deuteronomy 17: “And you shall inquire thoroughly” — these are two more interrogations. And with regard to conspiring witnesses it is written in Deuteronomy 19: “And the judges shall inquire thoroughly, and behold, the witness is a false witness” — here there are two more, making a total of seven interrogations.

The Gemara asks: how do we know to combine all of them? Why should we not say that in each offense the number of interrogations required is exactly the number stated there — three in the case of the idolatrous city, two in the case of conspiring witnesses, and two in the case of the individual idolater? If the Torah had wanted to teach that seven interrogations are required, it should have written all the interrogations in one context, and we would then have learned from that one case to all capital cases that seven interrogations are necessary. The Gemara answers that because in each of these passages the word “inquire” appears, this indicates that they are meant to be learned from one another and combined.

It is not clear whether this is a gezerah shavah. Seemingly it is, on the basis of the word “inquire” that appears in all of them. Yet the Gemara states that the word “inquire” causes them to be considered as though they were all written in one place, and therefore we combine all the interrogations and require seven in total. An ordinary gezerah shavah creates a similarity between contexts, but it does not merge or conjoin them.

The Gemara then objects that there are differences among these three contexts: each contains stringencies not found in the others, and vice versa. If so, how can one learn one from the others?

Here too the objection is unclear. We are not learning one from another; rather, we are combining them. If so, why does the Gemara assume that they must all resemble one another? It seems that we are indeed performing an ordinary gezerah shavah here, by which the three contexts are learned from one another.

It appears that the Gemara’s intent is to say that we equate these three contexts because of the common word “inquire.” But an ordinary comparison cannot work here, for in a standard gezerah shavah one context contains an explicit rule while the other contains a lacuna. The gezerah shavah fills the lacuna by comparing the obscure passage to the explicit one. Here, however, there is no lacuna in any of the three contexts. In the idolatrous city there is a requirement of three interrogations; in conspiring witnesses there are two; and in the case of the individual who worshipped idolatry there are likewise two. How then can one be equated with another? Which one would prevail over the rest? The only way this gezerah shavah can equate the three contexts is by assuming that all these requirements must be combined, yielding seven interrogations, and that total must apply in each of the three contexts. For this reason, the gezerah shavah here not only compares the three contexts but also turns the situation into one in which all three are considered as though written together.

In the end, then, we do indeed equate the three cases and require seven interrogations in all of them. To this the Gemara objects: how can they be equated, when each contains stringencies absent from the others? The Gemara answers that we derive a gezerah shavah of “thoroughly”-“thoroughly,” and because it is mufneh it cannot be refuted.

Up to this point it seems that the gezerah shavah of “inquire”-“inquire” was not mufneh, and therefore one could challenge it with a refutation and thereby collapse the derivation of seven interrogations. The Torah therefore adds the mufneh gezerah shavah of “thoroughly”-“thoroughly,” which is immune to such refutation. We thus seem to have two gezerah shavah here, one mufneh and one not. It is not clear why two are needed. Three possibilities suggest themselves here:

1. By this stage of the sugya, the first gezerah shavah is no longer being used at all.
2. It may still serve to unite the three passages as though they were written in one place. Against that there is no room for refutation on the ground that they are not similar, because we have not yet compared them. Afterward comes the gezerah shavah of “thoroughly”-“thoroughly,” whose function is to equate them. That gezerah shavah could indeed have been challenged on the basis of the stringencies that show they are not alike. But it is a mufneh gezerah shavah, and therefore one cannot refute it.
3. If the gezerah shavah from the word “inquire” had been received by tradition, there would likewise have been no room to refute it at all, as Rashi in Gittin, cited in the previous chapter, states. This proves that it was not received by tradition. But how can there be a gezerah shavah that is neither mufneh nor based on tradition? Perhaps the availability of the word “thoroughly” is the solution: it is required only to hint at the gezerah shavah of “inquire.” It cannot itself serve as the basis for a gezerah shavah, since the word is not superfluous. Yet the altered diction instructs us to perform a gezerah shavah from the word “inquire.”

The Gemara then asks what the availability is here. It answers: with conspiring witnesses and with the individual idolater there is availability because the Torah should have written “and they shall inquire and investigate,” yet it writes only “inquire thoroughly,” without “investigate.” Likewise, with the idolatrous city the Torah ought to have written “and investigate, investigate, and inquire,” and we would thereby have learned three interrogations. From the fact that the Torah uses the expression “thoroughly” in all three cases, this change of language constitutes mufneh.

It should be noted that in all three cases the word “thoroughly” is not superfluous, since from it we learn the obligation of an additional second or third interrogation. Even so, the very fact that the Torah changed its wording instructs us to use the word both for adding another interrogation and for creating a gezerah shavah between the contexts.

From here we learn several novel points, some of them only as possibilities:

1. A word that is not superfluous can serve in a gezerah shavah. A change of wording functions like mufneh.
2. A gezerah shavah can conjoin contexts and not merely compare them. Perhaps such a gezerah shavah cannot be challenged even if it is not mufneh, because it does not assume similarity between the contexts. A refutation that points to the absence of similarity obstructs only the comparison between the contexts, not their conjunction.
3. One may derive a gezerah shavah that is not mufneh; the difference is only that such a gezerah shavah is subject to refutation. A mufneh gezerah shavah is not subject to refutation.⁴
4. According to the last line of approach proposed above, this is a case of superfluity in one word — “thoroughly” — which generates a gezerah shavah from another word — “inquire.” This is exactly what we proposed in last year’s essay.

A Note from Babylonian Talmud, Sotah 16b

There is another example of a gezerah shavah based on a change of word rather than on superfluity, in Babylonian Talmud, Sotah 16b. There the Gemara compares the dust placed in the bitter water of the sotah with the ashes placed in the purification water of the red heifer.

The Gemara cites a baraita that discusses the order in which the water and the dust are placed. The dust should be placed on top of the water, but even if the order is reversed it is valid. This is learned from a gezerah shavah of “dust”-“dust.” The Gemara explains that in the case of the red heifer the substance is really ashes and not dust, and that the Torah changed its wording and wrote “dust” so that we could perform a gezerah shavah. This is therefore another example of a change of language without superfluity, by reason of which we perform a gezerah shavah.

To be sure, it is not stated there that the gezerah shavah is considered mufneh, because there no refutation of the gezerah shavah is raised. As noted above, the mere performance of a gezerah shavah does not in itself require mufneh; see conclusion 3 above. The change from “ashes” to “dust” is required simply so that a gezerah shavah between the two contexts can exist at all, since without the change there would have been no similarity between them. In this respect the situation here differs from the previous example, for there the gezerah shavah could have been performed even without the change of wording, and the change was needed only to create mufneh. Moreover, here in the sugya in Sotah, even if this were considered mufneh, it would be only one-sided mufneh, whereas there the word “thoroughly” is available on all sides.

It may indeed be that such a change does not make the gezerah shavah mufneh, and this is why it does not emerge from the Gemara there that it is considered mufneh. Only a superfluous change, beyond the gezerah shavah that already exists, makes it mufneh. This strengthens what we saw above: the change of “thoroughly” makes the gezerah shavah of “inquire” mufneh, rather than the gezerah shavah of “thoroughly”-“thoroughly” itself. There a gezerah shavah already exists, and the entire purpose of the change is simply to make it mufneh; see lesson 4 above.

We may strengthen our claim by noting that Tosafot, s.v. “Shina,” there in Sotah, raises two objections to the change from “ashes” to “dust”:

1. This is not really a change, since in Scripture ashes and dust are interchangeable.
2. That change is already required by Beit Hillel, in Babylonian Talmud, Hullin 88b, to teach that ashes are also called “dust.”

On our view, this is readily explained. The change indeed is not a change in the sense of mufneh. But in order for a gezerah shavah between these two contexts at all — sotah and red heifer — the Torah had to use the same word. And in fact this is a gezerah shavah that is not mufneh.⁵

Two Superfluous Words on One Side of the Gezerah Shavah

We find in the Talmud another phenomenon that seems to bear on the possibility of separating the available word from the word used in the derivation. The Gemara in Babylonian Talmud, Yevamot 70b derives that an uncircumcised person may not eat terumah (the priestly portion) by means of a gezerah shavah of “resident and hired laborer”-“resident and hired laborer” from the Passover offering. In the course of the discussion it becomes clear that there is a refutation available against this gezerah shavah, and therefore it must be mufneh in order to remain valid. According to Rabbi Eliezer, it must be mufneh on both sides. The Gemara explains that in this case the words “resident and hired laborer” in the Passover passage are necessary, but in the terumah passage they are superfluous. The problem is that both are found on one side only, and that is not two-sided mufneh. The Gemara explains that because there are two superfluous words here, we treat this as double availability, and accordingly there is here a gezerah shavah that counts as mufneh on both sides.

A priori there are three ways to understand this gezerah shavah:

1. The gezerah shavah is “resident”-“resident,” or “hired laborer”-“hired laborer,” and the above consideration serves only as the basis for regarding it as mufneh on both sides.
2. There is here a gezerah shavah of “resident and hired laborer”-“resident and hired laborer.”
3. After the transfer, there is here a gezerah shavah of “resident”-“hired laborer,” when one of the two available words in the terumah passage is shifted over to the Passover side and serves as one side of the gezerah shavah.

The first two possibilities indicate that there may indeed be a separation between the availability and the gezerah shavah, as we suggested above. In both of them, the available word is not the word actually serving in the gezerah shavah itself, though in different ways. The third possibility — the only one that denies the option of separating availability from gezerah shavah — is problematic in its own right, since it proposes a gezerah shavah between two nonidentical words.

A glance at the sugya immediately shows which possibility is correct. We saw that the number of instances of availability equals the number of superfluous words. One might have thought that since the gezerah shavah rests on the similarity of a pair of words, as in possibility 2, a case in which both words are available would still amount to only one instance of availability. The Gemara’s treatment of this as two instances apparently indicates that the gezerah shavah is not made from pair to pair, but from one word to its parallel. This seems to rule out possibility 2.

The Gemara itself says: “Cast one to the lamed and cast one to the melamed” — that is, to the lamed, the passage to which the rule is being applied, and to the melamed, the passage from which the rule is learned. This seems to indicate possibility 3. One might still say that the intention is really possibility 1, and that the “casting” is stated only with respect to the considerations of availability — namely, that this is treated as though there were one availability in the lamed and one in the melamed. This is suggested somewhat by Rashi’s language there, for he immediately explains that the discussion concerns availability, though this is not conclusive. Add to this the difficulty noted above with possibility 3 — that the gezerah shavah would then be between nonidentical words — and we arrive at the conclusion that this is probably the Gemara’s intent.

However, Tosafot there, s.v. “Shadi,” explicitly explains that this is considered as though “hired laborer” were written here unnecessarily and “resident” there unnecessarily. They go on to ask how one can perform a gezerah shavah between nonidentical words, and from this it is clear that they understood the sugya in accordance with possibility 3. They answer that sometimes a gezerah shavah is indeed performed between nonidentical words, if they are similar in meaning and occur in the same context. For example, “this returning is that coming,” in Babylonian Talmud, Yevamot 17b, from the verses “and the priest shall return” and “and the priest shall come” in Leviticus 14.⁶

The author of Yavin Shemu’ah, principle 102 — see Entziklopedia Talmudit there, note 113 — brings proof for this principle from another sugya, in Babylonian Talmud, Sanhedrin 64a. There the Gemara derives “garment and skin” said concerning a corpse, in Numbers 31:20, from “garment and skin” said concerning a creeping creature, in Leviticus 11:32. There, although both words are superfluous, the Gemara treats them as only one instance of availability. The author of Yavin Shemu’ah explains that these words — “garment” and “skin” — are not one matter, and therefore they are not to be regarded as two separate instances of availability.

His intention is apparently that since these words are not one semantic matter, then even though both are superfluous and there would therefore seemingly be a double availability, one cannot transfer one of them to the melamed side and perform a gezerah shavah of “garment”-“skin,” because these words are not fit to be related by gezerah shavah. The problem is not the availability but the ability to compare these words by means of a gezerah shavah.

Yet one may still ask why we should not leave them in place and perform a gezerah shavah of “garment and skin”-“garment and skin,” while still counting this as double availability. In other words: what gezerah shavah is ultimately being performed in the sugya in Sanhedrin?

In light of the foregoing analysis, it is clear that the gezerah shavah there is indeed “garment and skin”-“garment and skin,” and therefore it cannot be regarded as double availability, since two-sided availability must be divided between the lamed and the melamed. According to our proposal, there should have been no obstacle to making a gezerah shavah of “garment and skin”-“garment and skin” and nonetheless treating the fact that both words on one side are available as double availability. From here we see that one does not do this.

The seemingly unavoidable conclusion is that there must be some connection between the availability and the words being compared. Yet perhaps one might still reject this and say that the requirement is only that the availability be distributed between the two sides — the lamed and the melamed — exactly as the Sages said: “mufneh on both sides.” That still does not necessarily require identity between the available word and the word actually used in the derivation. The matter still requires much further investigation.

Footnotes