Falsehood and Retroactive Transgression in Halakhah (Column 520)

With God’s help

Disclaimer: This post was translated from Hebrew using AI (ChatGPT 5 Thinking), so there may be inaccuracies or nuances lost. If something seems unclear, please refer to the Hebrew original or contact us for clarification.

Dedicated to my daughter–study partner, Rivka, may she live long,

May your strength be for Torah, and may you attain great things.

A few days ago I studied with the aforementioned rebbetzin a sugya in Tractate Shevuot that deals with defining a false oath (shevuat sheker) versus an oath in vain (shevuat shav). These concepts are rather vague, and their meaning depends on how we understand the very notion of an oath. From there broader connections arose for us that touch on retroactive “corrections” in Halakhah, which I wish to present here.

Vow and Oath

As is known, Halakhah recognizes two kinds of personal acceptances through which a person creates new prohibitions for himself (without transgressing “do not add”): a vow (neder) and an oath (shevuah).[1] The Gemara at the beginning of Tractate Nedarim (2b) distinguishes between them in halakhic character: a vow is a status on the object (a prohibition rests on the item itself—k’asar chefza a’nafshai), while an oath is a status on the person (a person binds himself relative to the object—k’asar nafshai min chefza). I will not enter here into precise definitions (see Column 230). What they share is that a person’s declaration creates a prohibition to do or not to do something. Typically, a vow forbids benefiting from or using a given object, while an oath forbids performing or refraining from a given action, not necessarily connected to a specific object.

At the beginning of Parashat Mattot, these two laws appear in one verse (Numbers 30:3):

“If a man vows a vow to the LORD, or swears an oath to bind a binding upon his soul, he shall not profane his word; according to whatever proceeds out of his mouth, so shall he do.”

The verse treats both vow and oath as a self-imposed “binding,” obligating him not to profane his word—that is, to fulfill what he vowed or swore. Seemingly, the parallel in the verse suggests they are two shades of the same genre of prohibition.

There are additional differences between vows and oaths (most of them disputed), the most prominent being invocation of the Divine Name. An oath requires swearing by God’s Name, whereas a vow does not. True, a few Rishonim waive the need to utter the Name explicitly in an oath; yet even they, according to most, explain that via the laws of “handles” (yadot) we treat it as though he did invoke the Name. Hence the Gemara treats violating an oath as a problem of profaning the Divine Name—bearing it in vain or falsely—and not merely profaning one’s speech as with a vow. The implication is that these are not two shades of the same thing but two essentially different prohibitions. In a vow, the point is to keep what one said (or accepted—even toward oneself). In an oath, the prohibition is the profanation of God’s Name, and perhaps in addition, a duty to keep one’s word. The question is what the relationship is between these two dimensions in an oath: is the duty to do what I swore grounded in preserving the honor of the Divine Name, or in the obligation to keep my commitment (not to profane my speech)?

This brings us to additional kinds of oaths. For our purposes the key one is an oath in vain (shevuat shav). A vain oath is swearing to something everyone knows is untrue (Shevuot 21a: “He swears to change what is known to people”)—e.g., swearing that the heavens are below and the earth above, swearing that a man is a woman, and the like. See Rambam, Laws of Oaths 1:4. Here there is no dimension of “keeping one’s word,” for this oath is not a commitment to do or not do something. Regarding this it makes no sense to say “he shall not profane his word,” nor “as whatever comes out of his mouth, so shall he do.” There is nothing to fulfill or profane, for we speak of facts, not an action-bound commitment. The very existence of shevuat shav shows clearly that oaths carry another dimension—profanation of the Divine Name—and not merely profanation of speech. Once we’ve seen that with a vain oath, we may infer that false oaths (shevuat sheker) likewise have this dimension (in addition to a duty to keep one’s word). Note that with vows there is no vow about the past at all—only about the future—since there the entire matter is guarding one’s speech.

What Is a False Oath?

The Gemara in Shevuot 20b–21a discusses how to define a false oath (as opposed to a vain oath). I preface that there are four types of oaths: about the past and about the future, and within each, an oath to do (e.g., “I will eat”) or not to do (e.g., “I will not eat”). The sugya cites two Amoraic views. The first is Rav Dimi in the name of Rabbi Yoḥanan:

“When Rav Dimi came, he said in the name of Rabbi Yoḥanan: ‘“I will eat,” and “I will not eat” are false oaths (sheker), and their warning is from here: “You shall not swear by My Name falsely” (Leviticus 19:12). “I ate,” and “I did not eat” are vain oaths (shav), and their warning is from here: “Do not bear the Name of the LORD your God in vain” (Exodus 20:7). As for ‘konamot’ (vow-language), he transgresses “he shall not profane his word.”’”

Rav Dimi holds that an oath about the future is a false oath, whereas an oath about the past is a vain oath (like swearing to something known to all to be untrue, which is certainly a vain oath). From his conclusion it seems “he shall not profane his word” applies only to vows, not to oaths at all—even though we saw in the verse above that “he shall not profane his word” speaks of oaths too.

What is the difference between an oath about the past and one about the future? An oath about the future is a commitment to do something; an oath about the past asserts a state of affairs that already occurred (no action-commitment). An oath about the past is a vain oath, because he simply swore; anyone hearing it knows it is not and cannot be true. By contrast, an oath about the future is a false oath. On its face this is odd, for the past-directed oath looks more like a lie—at the moment he spoke, it was false. Whereas a future-directed oath is not a lie—when he spoke it was not yet clear he would not uphold it. The statement was not false; it is thus less likely to interpret that as “false.”

Rashi writes:

“And its warning is from ‘You shall not swear by My Name falsely’—meaning, do not swear by My Name in a way that will result in falsehood later; and the warninger must warn him that by his subsequent act he will be violating ‘You shall not swear by My Name falsely.’”

He explains that a future-oriented oath becomes false by virtue of the act later done or not done. This raises a question about hatra’ah (forewarning), for at the time of the oath there is no falsehood, hence no transgression. Moreover, the act itself is not a transgression per se, for he is otherwise permitted to do it. The problem is that by doing it he turns his prior statement into a lie. The transgression lies in what he said earlier (the oath), not in what he is doing now. Therefore Rashi stresses that the forewarning is issued before the act, not before the oath; but in content, the warning must be that through his act he will render his earlier oath false—not that the act itself is prohibited. As a result of the act, his earlier statement becomes, retroactively, a lie.

Note that Tosafot s.v. “konamot” there write:

“‘Konamot’—he transgresses ‘he shall not profane his word.’ So too in an oath he transgresses ‘he shall not profane his word,’ for it is written (Numbers 30): ‘or he swears…, he shall not profane his word’; and in Nedarim 16b it is explicitly taught that there is ‘he shall not profane’ for oaths as well. And here the point is that with ‘konamot’ there is only ‘he shall not profane,’ and nothing else.”

They maintain that one also violates “he shall not profane his word” with an oath, for as we saw, both are written in the same verse before “he shall not profane.” Rav Dimi’s point is that in konamot there is only that prohibition, whereas it exists in oaths too. Of course, that applies only to an oath about the future, for regarding the past, “profaning his word” has no meaning, as I explained above. The verse at the start of Parashat Mattot concerns vows and oaths about the future, and for both there is “he shall not profane his word” upon doing something that profanes the earlier statement.

Rashi, however, seems to learn otherwise and holds that “he shall not profane his word” does not apply to oaths at all, not even to future-oriented oaths. He writes explicitly in the next comment:

“‘Konamot’—this is ‘he shall not profane’—that is, ‘he shall not profane,’ which implies future, does not come to warn about oaths regarding the future; for future-oriented oaths are derived from ‘You shall not swear by My Name falsely’; and when ‘he shall not profane’ arrives, it comes to warn regarding vows—excluding Rav Dimi, who said the warning for future-oaths comes from ‘he shall not profane.’”

The Ritva explains that indeed “he shall not profane” in the verse refers to vows, not to oaths. So what is the prohibition in an oath? Rav Dimi himself states it: Leviticus 19:12—“You shall not swear by My Name falsely.”

If so, at least according to Rashi, oaths are fully distinct from vows. In a vow, a person is required to keep his commitment (even a self-commitment), i.e., not to profane his speech. With oaths there is no independent interest in keeping the commitment; the concern is not to profane the Divine Name by which he swore. In a vow the act profanes the speech, but the transgression is doing an act that profanes the speech. The essence of a vow is a binding to do or not do something. With an oath there is no intrinsic binding to act or refrain; the transgression is profaning the Name, and that is only a result of some act.

Why should one not also be required to keep his commitment in an oath? How is it different from a vow? It seems Rashi understands that there is no halakhic obligation to keep one’s words as such. In a vow there is a status-change in the object, creating a new prohibition by which the act becomes forbidden. But in an oath, where there is no status-change in the object, only a personal duty, what remains is mere speech—and there is no true obligation to keep mere speech. All that exists in an oath is the profanation of the Divine Name. Tosafot apparently understood that an oath includes both elements (due to the verse in Mattot).

In light of the above one may wonder about a vain or false oath: how is one flogged (lokeh) for it, when there is no act—only speech? True, he performs an act when he violates his oath, but as we saw, the prohibition is the profanation of the Divine Name, or the earlier false statement, not the act itself. The Gemara continues that a special derivation is needed:

“For Rav Idi bar Avin said in the name of Rav Amram, who said in the name of Rav Yitzḥak, who said in the name of Rabbi Yoḥanan: Rabbi Yehudah says in the name of Rabbi Yose the Galilean: Every ‘do not’ in the Torah—if it involves an act, one is flogged for it; if it does not involve an act, one is not flogged for it—except for one who swears (falsely), one who substitutes (sanctity), and one who curses his fellow with the Name. From where [do we know] swearing? Rabbi Yoḥanan said in the name of Rabbi Shimon bar Yoḥai: The verse says, ‘Do not bear the Name of the LORD your God in vain, for the LORD will not hold him guiltless’ (Exodus 20:7). The heavenly court does not clear him, but the earthly court flogs him and thus clears him.” Abaye said to Rav Pappa: Perhaps the verse means he will not be cleared at all? He answered: Had it written ‘he will not be cleared,’ as you said; now that it says ‘the LORD will not clear,’ it is the LORD Who does not clear him, but an earthly court flogs and thereby clears him.”

A special verse teaches that one is flogged for a false or vain oath. Hazal derive from the verse that only the heavenly court does not clear one who bears the Name in vain, but an earthly court clears him—by flogging. Note the verse addresses a vain oath, yet the Gemara extends it to a false oath as well. Evidently, the foundation in both is the same: profaning the Divine Name. One who bears His Name in vain or falsely is not cleared above. This teaches us that even in a false oath, where there is a transgressive act and not merely speech, the transgression is not the act but the profanation of the Name that results—just as in a vain oath where there is no act at all.

Further on the Gemara reports that Ravin disputes Rav Dimi, also in the name of Rabbi Yoḥanan:

“When Ravin came, he said in the name of Rabbi Yirmiyah, who said in the name of Rabbi Abbahu, who said in the name of Rabbi Yoḥanan: ‘“I ate,” and “I did not eat” are false oaths, and their warning is from “You shall not swear by My Name falsely.” “I will eat,” and “I will not eat” transgress “he shall not profane his word.” And what is a vain oath? One who swears to change what is known to people.’”

On his view, an oath about the past—even if it is not obviously false to all (e.g., “I did”/“I did not”)—is falsehood; a future-oriented oath falls under “he shall not profane his word” (as per the verse we saw). According to him, a vain oath is only a past-oriented oath about a matter known to everyone (“swearing to change what is known”).

We have, then, an Amoraic dispute regarding past-oriented oaths: according to Rav Dimi it is a vain oath (and the future-oriented is false), whereas according to Ravin it is a false oath (and the future-oriented is “he shall not profane his word”).

The Rambam’s Position

The Rambam, Laws of Oaths 1:2–3, rules:

2) A shevuat bitui (oath regarding speech/behavior) applies only to matters one can do—whether for the future or the past. How so for the past? “I ate,” or “I threw a stone into the sea,” or “So-and-so spoke with so-and-so”; or “I did not eat,” or “He did not throw a stone into the sea,” or “So-and-so did not speak with so-and-so.” How so for the future? “I will eat” or “I will not eat,” or “I will throw” or “I will not throw a stone into the sea.” These are two [types] for the past and two for the future.

3) If one swore any of these four types and did the opposite—for instance, he swore not to eat and ate, or that he would eat and did not eat; or [he swore] “I ate” when he had not eaten, or “I did not eat” when he had eaten—this is a false oath (shevuat sheker). About this and the like it is said: “You shall not swear by My Name falsely.” If he swore falsely intentionally, he is flogged; if inadvertently, he brings a variable sin-offering, as it says, “and it was hidden from him and he knew and became guilty,” etc.

He rules that both past- and future-directed oaths can be false oaths (clarification: transgressing a shevuat bitui is a false oath). Seemingly this fits neither Rav Dimi nor Ravin. The Kesef Mishneh notes this:

“We must say our master (the Rambam) explains that although Rav Dimi, when he ascribes ‘he shall not profane’ to konamot, means that they transgress only ‘he shall not profane’; and when Ravin ascribes ‘I will eat/I will not eat’ to ‘he shall not profane,’ he means they transgress that as well—for it cannot be that they transgress that alone, since shevuat bitui is explicitly discussed in the verse regarding ‘I will eat/I will not eat,’ and about it the verse says ‘you shall not swear falsely.’ Therefore, ‘I will eat/I will not eat’ are false oaths by all opinions; and he ruled like Ravin because he was stated last, and often we rule like the latter version, and moreover the general sugya follows Ravin.”

He infers from the verse in Parashat Mattot that certainly “he shall not profane” applies to future-oriented oaths (the verse there addresses only future oaths—otherwise the notion of profaning by a subsequent act would not apply). If so, that same verse shows that future oaths are false oaths; thus no one disputes that. The debate between Rav Dimi and Ravin is whether there is also false oath for the past; on that, the Rambam rules like Ravin.

According to this, even if he swore he would eat and did not eat—where there is no act—he would be flogged, since he profaned the speech, and for that a court is to “clear him” (by flogging). As noted, in an oath one is flogged for profaning the Name, not for the act; hence it is irrelevant whether his violation entailed an act. However, regarding “he shall not profane,” he would not be flogged in such a case, for there the lashes relate to speech.

Subsequently, however, the Kesef Mishneh takes the opposite tack and, citing Maharik, offers a puzzling explanation:

“What our master wrote—‘If he swore falsely intentionally, he is flogged’—does not refer to one who swore he would eat and did not eat, for that is not a false oath, since it was not false at the time of the oath, as our master writes at the end of ch. 4. Rather it refers specifically to one who swore ‘I did not eat’ when he ate, or ‘I ate’ when he did not, for that was a false oath at the moment of swearing; and all the more so, one who swore not to eat and ate—since he performed a full act. So wrote Mahari Kolon, root 134.”

He assumes that even the Rambam agrees there are no lashes for “I will eat”/“I will not eat.” Apparently he claims there is no falsehood there—but that is not the Rambam’s view, as he writes in 4:20–21:

20) “If one swore, ‘I will eat this loaf today,’ and the day passed without his eating it—if unintentionally, he brings a variable sin-offering; if intentionally, he is not flogged, for he did not perform an act—even though he violated a false oath.”

21) “Why is he flogged if he swore ‘I ate’ when he had not, or ‘I did not eat’ when he had—despite not performing an act? Because at the moment of his oath, he swore falsely. But if he swore to do something and did not do it, that is not a false oath at the time of swearing.”

We see that there is a false oath, but without an act there are no lashes. It seems that what determines “a prohibition with an act” is the physical aspect: if the violation is carried out via an act, lashes apply—even though the essence of the transgression is the prior false speech (the act merely rendered that speech false).

In any case, it appears he holds that to incur lashes, one of two conditions must hold: either the speech was false at the time spoken (and not merely retroactively revealed as false)—hence lashes for a vain oath and a past-oriented false oath; or the violation involved an act (then lashes apply even if it was not false at the time spoken). Thus, if one swore to eat and did not, there is falsehood but no lashes: no act, and no falsehood at the time of speech.

The link between these two criteria is unclear. If the transgression lies in the falsehood of the oath itself, why should he be flogged for “I will not eat”/and then he ate? What if there is an act? The reason he is not flogged (in the no-act case) is not due to lack of act alone, but because there was no falsehood at the time of speech. It seems that for the Rambam, rendering speech retroactively false does not suffice to flog him for false speech, since at the time he spoke, he did not lie; it cannot be counted as lying. Perhaps the lashes there are for “he shall not profane his word”—i.e., for the act, not for profaning the Name (as we saw in the Kesef Mishneh that according to the Rambam, “he shall not profane” applies to oaths too)—and for that, an act is required. I will return to this difficulty at the end.

What Is the Dispute About?

Analyzing the Rambam may illuminate the Amoraic dispute. Seemingly, Rav Dimi views a future-oriented oath as a false oath because the later act turns it into falsehood; he is untroubled that at the time of swearing it was not false, since ultimately it is clarified retroactively that it was false. But this likely is not his view, since he classifies a past-oriented oath as a vain oath, not a false one. There, the speech was certainly false as spoken; yet in his view it is not “falsehood” but “vanity.” It seems he indeed sees the transgression of a false oath in the future act. That is, he considers oaths partly like vows: the focus of the prohibition is the act that makes the oath “a lie.” On his view the transgression in a false oath is the act, not the oath. When the lying occurs in the oath itself, that is a speech-violation, not an act—hence a vain oath.

Practically, we rule like Ravin. He defines only the past-oriented oath as a false oath, for he is unwilling to label an oath that becomes false retroactively as a “false oath,” since at the time of swearing he did not lie. That, for him, is a vain oath. One cannot flog someone for a statement that was not false when uttered and only later became so. A past-oriented oath is, of course, false when spoken.

Between “Retroactively” and “From Now On, Retroactively”[2]

In simple terms, the act retroactively turns the earlier speech into a lie, thereby profaning the Name. In the straightforward sense there is an igla’i milta le-mafrea—at some future point, it becomes clear what was the case in the past (which we did not understand). It resembles ordinary information discovered about the past: e.g., a historian uncovers a fact, shedding new light on history; but that discovery does not change the past, only our knowledge of it. A gap in our information was filled. Sometimes no human knew it (save the Omniscient), and still the later discovery does not alter the past, only reveals it after the fact.

Consider the claim I make today: “Tomorrow, so-and-so will violate his oath.” If tomorrow he does violate it, it turns out the claim was already true today. Our ignorance did not change that fact. There is no causation flowing backward in time—only completion of missing information. One can frame it thus: the truth-value of a statement is set by its match to the reality it describes. If there is a match, the statement is true—even if we do not know it; if not, it is false. If I say “There are one billion ants in the world,” there is no practical way to verify it; but if that is indeed the number, the statement is true, because its content corresponds with the state of affairs in the world. So too with “So-and-so will violate his oath”: our present ignorance does not affect its truth-value. This is the mechanism of completing information—what Halakhah calls igla’i milta le-mafrea.

Yet in Halakhah there are mechanisms where the future not only reveals missing past information, but actually produces a past status causally. This is what R. Shimon Shkop calls “from now on—retroactively” (mikan u-lehabah le-mafrea; see, for example, R. Shmuel Rozovsky’s Shiurim to Makkot §490, and elsewhere). This is the case with the annulment of vows (hafarat nedarim), where the annulment uproots the vow retroactively (not merely reveals that it never took effect). R. Shkop in his Kuntres ha-Tena’im (at the end of his ḥiddushim to Gittin) shows that this is also how conditional acts function.

With this in mind, consider an oath to perform something in the future that is then violated. Factually, it is clarified that it was false from the outset. We did not know it then, but in reality there is no match between its content and what happened; hence it is false. Halakhically or normatively, however, we may ask whether one can flog a person for an act that was permissible at the time and only ex post facto became an offense. The fact that it is false is not enough to designate a transgression.

This question may be asked whether the future act merely reveals that the oath was false, or whether it makes it false. Even if it merely reveals it, perhaps he “lied” already then; but can lashes be given for an act that at the time was no sin, and only something later rendered it so? We will now explore this briefly.[3]

“Saving a Blessing”

It is known that there is a prohibition against a blessing in vain (berakhah le-vatalah; Berakhot 33a; Rambam, Laws of Blessings 1:15; Shulḥan Arukh, OḤ 215:4, and more). Some hold it is a biblical prohibition (“Do not bear [the Name]”—uttering the Name in vain); most hold it is rabbinic (with “Do not bear” as an asmakhta).

Now, there are situations in which a person recites a blessing, and ultimately something happens that shows it to have been in vain. For example, one blesses over fruit but decides in the end not to eat it (nimlach). Another case: one separated terumah and then consulted a sage who annulled the designation, reverting it to tevel. It turns out his blessing over separating terumah was in vain. Similarly, one who blessed over searching for leaven (bedikat ḥametz) but found nothing—some hold that in such a case his blessing was in vain. A similar question arises regarding rabbinic annulment of betrothal (hafqa’at kiddushin), where it emerges retroactively that the blessings of betrothal and marriage were in vain. In all these situations one may ask: was the blessing indeed in vain? And if so, is there a prohibition to bring about that result (e.g., to ask a sage to annul terumah; to annul betrothal; not to eat after a blessing; etc.)?

In this context note the Rema, OḤ 432:2, regarding searching for leaven:

“Gloss: And the practice is to place tidbits of ḥametz where the searcher will find them, so that his blessing not be in vain (Mahari Brin). But if none were placed, it does not prevent [the search], for every person has in mind with the blessing to eliminate [leaven] if it be found (Kol Bo).”

He records a custom to place crumbs so the blessing not be in vain. He himself does not deem this essential, since the blessing is over the act of searching; therefore, even if he finds nothing, the blessing is not in vain. But even if we ruled like that custom, it would not prove our point. That discussion concerns whether to ensure ahead of time that the blessing not be in vain; our discussion is whether there is a duty to “save” an already-recited blessing. In such a case it may be that what has been said, has been said; there is no obligation to change matters to “save” it.

A foundational source is the Ritva, Ḥullin 106b, who writes:

“If one washed his hands for eating and blessed ‘al netilat yadayim’ and then changed his mind and did not eat now—there is no problem, and we do not require him to eat so that his blessing not be in vain; for since he washed his hands, the blessing over the washing is complete—for that is what he blesses—and at that moment his intent was to eat. So I presented to my teacher, may he live, and he agreed.”

He holds there is no transgression at all, and no obligation to go back and eat. This seems to show there is no duty to “save” deeds from becoming transgressions. There is a prohibition against a blessing in vain, but no duty to save a blessing that, when uttered, was legitimate from becoming one in vain. Presumably, if lashes existed for a blessing in vain, he would surely not be flogged; the implication for our false-oath discussion is evident.

On a second reading, however, the picture is not so simple. His reason is that there is no transgression here, since he blessed over the washing, and the washing occurred. It implies that if there were a transgression, there would be a duty to “save” the blessing. That is, he is not claiming there is never a duty to save a transgression, only that there is none here (nothing to save). This suggests that where the blessing had nothing to rest on—e.g., he blessed over a fruit and then decided not to eat it—then if he does not eat, the blessing is indeed in vain. It follows that, in such a case, he would have an obligation to eat the fruit to “save” the blessing.

Responsa Amudei Esh §2, sub-§33, raises a similar doubt about blessings over mitzvot:

“We may inquire about one doing a mitzvah that requires a blessing, and he blessed—if he is able to refrain from doing the mitzvah, perhaps we should say he may not withdraw, lest the blessing be in vain.”

He indeed cites the Ritva and rules accordingly:

“Therefore, it seems in our case that if he has a great need, his blessing is not in vain though he does not do the mitzvah immediately; however, certainly, we would not permit him to do so without cause.”

Yet in light of what we saw, this conclusion is not correct. If one blessed over a mitzvah and did not perform it at all, that is a blessing in vain even according to the Ritva. As noted, the blessing over washing is not in vain because the washing occurred. Our case is closer to one who blessed over fruit and changed his mind not to eat.

Additional Implications

At the start of his piece, the Amudei Esh cites a source from Responsa Sefer Yehoshua which rules:

“I saw in Responsa Sefer Yehoshua, in ‘Piskei u-Ketavim’ (§515), that the effectiveness of she’elah (asking a sage) to annul terumah or ḥallah is only when one separates along with the dough in question another dough, but where the separation is on that dough alone, if he asks [and annuls] it emerges that retroactively his blessing was in vain; this is akin to vow-like prohibitions which we do not permit.”

“In my view this is incorrect, and not analogous to vow-prohibitions; there the prohibition lies in the vow itself, and we are commanded to separate a person from prohibition—how can we permit him? Not so here, where at the time of the blessing he blessed permissibly; if later he wishes to ask [and annul], his blessing is not in vain…”

The Sefer Yehoshua rules that she’elah for terumah works only when one designates something else that remains terumah; otherwise the blessing becomes in vain—thus it would be prohibited to ask (since it retroactively renders the blessing in vain). He likens it to a vow: once a person vows, we certainly do not permit him to violate it, for that would make the vow-speech itself prohibited. This brings us right back to our topic (he assumes even in vows that the transgression is the profanation of the speech, not the act—somewhat like oaths).

But the Amudei Esh rejects this, arguing that in vows we indeed separate a person from transgression—the transgression is the act of violating the vow here and now, not the earlier vow-speech. In contrast, in creating a blessing in vain, one is turning something said earlier into a transgression, and that is neither prohibited nor necessary to prevent. Again we see that, in his view, there is no duty to “save” prohibitions, only not to commit them. If a vow turned the vow-speech into a prohibition, it would be permitted to violate it (at least for great need). From this it follows that for oaths—where we saw the prohibition is the profanation of the Name in the oath—rendering it retroactively false cannot be considered a “false oath.” Hence it is clear why a future-oriented oath is a vain oath, not a false one.

In Mishpat Kohen (by R. A. I. Kook) §39, this foundation is discussed at length, including the question of annulling betrothal. He asks how the Sages could annul betrothal when that would render the blessing a blessing in vain. He writes that the Sages may “uproot” Torah matters passively (shev ve-al ta’aseh), and asks whether such uprooting is classed as active (kum ve-‘aseh) or passive. The transgression of a blessing in vain is an active one; yet the act of blessing already occurred, and the Sages, through annulment, merely turn it into a transgression. He therefore analyzes whether this should be considered passive (shev ve-al ta’aseh), which is within their power. This is a middle view between the two possibilities we raised: the transgression is indeed the earlier act (the blessing), not the current annulment; but there is another way to “violate” it—passively, not actively.

For more sources, see note 218 of the editor to the Ritva there; Tevu’ot Shor §19:4; Mishpat Kohen §39; Yabi’a Omer vol. 2, YD §5; and others. Note that all these assume it is obvious there is a duty to “save” the blessing; their discussion is whether there is anything to save (i.e., whether in the given scenarios a transgression—a blessing in vain—exists). As we saw, this is not simple. It may be there is no such duty at all; and even if there is—perhaps it is only a passive duty.

It seems that underlying the claim that there is only a passive transgression here lies the conception that the transgression is that the earlier act was not “saved,” and not the original act of transgression itself. That is, where I blessed over fruit and then decided not to eat it, or dug earth on Yom Tov and then changed my mind not to slaughter, the “transgression” is not the blessing in vain or purposeless digging, but the failure to save the earlier act. When the act was done, it was done permissibly. Later circumstances that render it prohibited cannot transform the original act into a prohibition—its status depends only on the circumstances then prevailing. At most, there is now a duty to save the already-done act; failure to discharge that duty constitutes a transgression now (apparently the same nominal transgression, like a blessing in vain) but passively. Indeed, if there is a duty to save the earlier act, one who fails does transgress—but now, not then. If so, regarding a future-oriented oath, it is reasonable to say the later act is, at most, a failure to save the speech from becoming a lie and the Name from being profaned, i.e., at most a passive violation; but there is no active “false oath.” Some even hold there is no transgression at all, for there is no duty to save an earlier act from becoming a transgression. This could explain Ravin’s view that refuses to regard a future oath as a false oath. For the Rambam we saw that it is a false oath, but apparently without lashes—perhaps because it is a passive violation; this reconciles Maharik’s consideration in the Kesef Mishneh that mixed the “act” factor with the absence of falsehood at the time of speech.

Perhaps There Is an Amoraic Dispute About This Regarding Vows

I open with a question I have pondered for years. Suppose a man divorced his wife on condition she not drink wine for ten years. She divorced, remarried, and had children by the second husband. Now, after nine years, she wonders whether to drink wine. If she drinks, it will emerge retroactively that the divorce from the first husband was no divorce; her relations with the second were promiscuous, and her children by him are mamzerim. Seemingly, it is prohibited for her to drink, for it is forbidden to engage in promiscuity and to produce mamzerim. But on reflection, we can question this: it is permitted for a woman to drink wine—the first husband cannot turn a permitted act into a prohibition for her. If she drinks, the condition will undo the divorce, and the problems will arise—but that happens of itself; seemingly there is no prohibition on her part. Perhaps there is a bein adam le-ḥavero issue (imposing mamzerut on the children), but not the prohibition of promiscuous relations. Her relations were permitted at the time; if now they become forbidden, they cannot be considered “promiscuity.” Put differently: there is a prohibition to have promiscuous relations, but no prohibition to turn relations that were permitted into promiscuous ones.

In ch. 12 of my aforementioned book I suggested that the Amoraim dispute this in Nedarim 14b. The Mishnah there states:

“Mishnah: ‘Konam if I sleep, if I speak, if I walk… One who says to his wife, ‘Konam if I have relations with you’—he transgresses ‘he shall not profane his word.’”

The Mishnah teaches that konam can take effect on things that have no substance (not an object but an action).

And the Gemara:

“It was stated: ‘Konam that my eyes sleep today if I sleep tomorrow.’ Rav Yehudah said in the name of Rav: He should not sleep today, lest he sleep tomorrow. But Rav Naḥman said: He may sleep today; we are not concerned that he might sleep tomorrow. And Rav Yehudah concedes that if he said, ‘Konam that my eyes sleep tomorrow if I sleep today,’ he may sleep today; for with regard to the condition, people are not careful, but with regard to a prohibition, they are careful.”

If one forbids himself something today if a different event occurs tomorrow, Rav Yehudah and Rav Naḥman dispute it. But if he forbids tomorrow on the basis of today, there is no prohibition to create today a situation in which a prohibition will take hold tomorrow, since tomorrow he can guard himself against it.

This asymmetry in Rav Yehudah is unclear: why the difference? He could equally sleep today and be careful not to sleep tomorrow. Indeed, Rav Naḥman equates the two cases.

The difference may be between a condition and a prohibition. In the “tomorrow is only a condition” case, the person may think: today I will sleep—no prohibition; tomorrow I will sleep—again no prohibition. But tomorrow’s sleeping will clarify that yesterday’s sleeping was prohibited. Is such sleeping prohibited? Seemingly not. The act of sleeping is not itself forbidden; it only retroactively turns yesterday’s sleeping into a prohibition—but that has already happened. Thus there is no prohibition here at all, and therefore we fear it may happen. This is exactly like the criminal-liability analogy we discussed. There is no prohibition to perform an act that turns earlier permitted acts into prohibited ones. True, one might ask why the Gemara is concerned if there is no prohibition; apparently there is still a problem—perhaps only a passive one, not the vow-violation itself. The Gemara’s phrasing suggests this is about human psychology: people err in thinking there is a difference between a condition and a prohibition; they will take care not to violate a prohibition but will not be careful about a condition that merely turns an earlier act into a prohibition. On that understanding, halakhically there is a prohibition also to sleep tomorrow. Indeed, the Rishonim and Aḥaronim on this sugya assume otherwise. Still, on reflection, there is room to argue that halakhically there is truly no prohibition here. On our suggestion, there is no prohibition to turn yesterday’s permitted act into a prohibited one; hence it would be permitted to sleep tomorrow even if he slept today.

Back to Oaths

Note that all this discussion concerned vows, and as we saw, the straightforward view in vows is that the transgression is the act, not the earlier speech. True, the Torah says that doing the act profanes the speech, but the prohibition is the act, not the former speech. In oaths, by contrast, the prohibition lies in the speech, and the act merely reveals that the speech was a lie, thereby retroactively profaning the Name. There it is all the more reasonable to argue that we cannot treat this as a “false oath,” since falsehood must exist at the time of speech. This is likely Ravin’s position, refusing to view a future-oriented oath as a false oath.

However, there may be a distinction between oaths and “saving” a blessing or a vow. With oaths the Torah innovated that one must carry out what he swore; if he does not, he profanes the Name uttered in falsehood. That is precisely the Torah’s novelty: although the act is not the primary problem, but the false speech, still the Torah expects us to avoid acts that will produce the profanation. There may not be a separate “he shall not profane” prohibition on the act (as with vows), but the idea remains—that to avoid profanation we are commanded regarding the act. If we did profane, the transgression is in the profanation of speech; but there is a duty to avoid acts that will bring that about. That is, the oath-parashah innovates that even if elsewhere in Halakhah there is no duty to save from transgression and no prohibition to turn a permitted act into a prohibited one, with oaths such a prohibition exists. Therefore, here it will be an active transgression as well, even per R. Kook.

This may be why the Rambam (and Rav Dimi in the Gemara) sees a future-oriented oath also as a false oath. True, there is only a retroactive transformation of an act into a transgression; but with oaths, such transformation is treated like actual falsehood. We can now understand the Rambam’s difficulty noted earlier. We saw that for him two parameters govern lashes: to be flogged for a false oath, either the statement must be false at the time spoken, or the violation must involve an act. We wondered about the connection. Now we can say: if there is no falsehood at the time of speech, then the violation is passive (since he did not lie—he merely performed an act that turned non-false speech into falsehood). But if it is done with an act, then there is an active profanation of the Name, and lashes apply. Importantly, “active” here need not mean a physical deed per se; it means an active (as opposed to passive) violation. Consequently, speech can also count as active, even though generally in Halakhah speech is treated as lacking “an act.”

To sharpen this, consider again divorce conditional on the woman’s not drinking wine. There the act of drinking wine is entirely neutral and cannot be said to be wrong; a woman may drink whenever she wishes. True, it happens that the act nullified the divorce and rendered the children mamzerim, but those are contingent consequences; hence the drinking cannot itself be seen as a transgression. At most we can ask whether there is a duty to save earlier relations. But with oaths, the Torah itself requires us not to perform acts that profane our speech (and the Divine Name); i.e., even in oaths there is an aspect of prohibition on the act that profanes the speech. If it occurs, the ultimate transgression is defined as profanation of the Name; nevertheless the act is not neutral—it is not like the woman’s wine. The Torah explicitly warned us off that act; the resulting profanation is not an accidental consequence. Therefore, there one may say that the prohibition of lying in the original speech (as in a past-oriented or vain oath) also includes prohibiting turning the original speech into a lie. According to Rav Dimi, this may even count as active falsehood; according to the Rambam’s understanding of Ravin, it is active only if the violation entails an act rather than mere inaction. According to other Rishonim on Ravin, it is not falsehood at all—at most a profanation of speech (“he shall not profane his word”), not a false oath.

At least per the Rambam, it is now clear why the oath appears in the verse in Parashat Mattot, which speaks of future vows and oaths and includes them in “he shall not profane his word.” Even if we do not say like the Kesef Mishneh (derived precisely from that verse) that there is an explicit prohibition of profaning speech, still the act that profanes the speech (which in vows is the transgression itself) is part of the very prohibition of false speech in oaths. Without that, we could not view a future-oriented oath as a false oath.

[1] I am ignoring here bans (ḥerem) and naziriteship, which likely are not truly separate categories.

[2] See on this Column 465, and more.

[3] I discussed this at length in the fourth volume of my series on Talmudic logic, Logic of Time in the Talmud, chapters 12 and 20. Here I bring only a few among many examples.

[4] In chapter 20 there I also discussed whether the duty to save an act may be discharged even at the cost of committing a prohibition. That, too, has implications for our topic.