NEW YORK (VINNEWS/Rabbi Yair Hoffman) The laws of doubts are deep, and they come up everywhere in Torah. There is sfeika d’oraisa (a doubt in a Torah prohibition), sfeika d’rabbanan (a doubt in a Rabbinic one), doubt in money matters, doubt in ritual impurity, doubt about life and death and more. To learn a sugya the right way means learning it b’iyun — in depth. What follows is a step by step guide to studying the topic of doubts in a Torah law and doing so in the specific methodology of Iyun. Rav Elyashiv zt”l used to remark that the essential growth in Torah ins through Iyun. Some of the material here is based upon the ideas of Rav Henoch Leibowitz zt”l and much of it is based on the work of Rav Achikam Keshet shlita.
The difference between beki’us and iyun is simple. Beki’us asks “what”: what is the Gemara saying? Iyun asks “why”: why is this the law? Why did each side of the argument not accept the other side? Why did Rashi explain it this way, and use these exact words? Why wasn’t the first answer good enough? Why did the Rambam rule exactly the way he did?
Iyun is not really about knowing lots of commentaries. It is about using them to understand more deeply. True, Chazal said that Sinai — broad knowledge — is better than “one who uproots mountains” (sharp analysis). But that is only because someone with broad knowledge can also analyze; a person who is only sharp is not called Sinai. In every generation — Tannaim, Amoraim, Rishonim, Acharonim — the chachamim analyzed, asked questions, and answered them. The whole point of iyun is to stay with a question and think it through, instead of rushing to the next topic.
Everyone knows the rule that a sfeika d’oraisa is treated strictly (l’chumra). That is exactly where iyun begins — with the question fluency never stops to ask: how do we even know this rule, and what is it really?
Setting the Question
It is well known that a person may not say Krias Shema (or daven, and so on) next to excrement, or next to people whose lower body is uncovered. The Gemara asks what happens in a case of doubt:
Rav Yehuda said: A doubtful case of excrement is forbidden; a doubtful case of exposed nakedness is permitted. (Berachos 25a)
Why the difference? The Gemara explains. The rule against Krias Shema near excrement comes straight from the Torah — “v’haya machanecha kadosh” (Devarim 23:15), “and your camp shall be holy.” But the rule against Krias Shema near an uncovered body (once it has reached the ground) is only Rabbinic. And the general principle is: when there is a doubt, we are strict in a Torah matter (sfeika d’oraisa l’chumra) and lenient in a Rabbinic one (sfeika d’rabbanan l’kula).
This rule comes up in countless cases. The classic example in the Gemara and the poskim: a piece of meat, and we are not sure if it is shuman (kosher fat) or chelev (forbidden fat, which carries kareis). Because it is a doubt in a Torah prohibition, we are strict — and it is forbidden.
Now here is the question that will drive the whole sugya:
Question. Reuven ate a full meal, so he has to say Birkas HaMazon (bentch). But he is not sure if he already bentched. Let’s say the rule of “strict in a doubt” means he has to bentch again. Is that new obligation from the Torah, or only Rabbinic? The answer matters for other areas too — the laws of kavanah in mitzvos, the laws of a choleh (sick person), and more.
To answer this, we have to dig into the sugya itself. And as iyun always does, we start at the source.
From Where? The Source of Sfeika D’Oraisa
Iyun starts by asking where a law comes from and how it is defined. So let’s ask plainly: how do we even know that a sfeika d’oraisa is treated strictly? Did the Torah itself forbid the doubt? And if so, where does it say that?
The Gemara never says it outright. But there is one Gemara that deals with a doubt in a Torah law — the case of a mamzer. The Torah says (Devarim 23:3):
“A mamzer shall not enter the congregation of Hashem; even to the tenth generation he shall not enter the congregation of Hashem.”
Now, what about a doubtful mamzer — for example, someone whose father is unknown, so we are not sure if he is a mamzer or not? The Gemara (Kiddushin 73a; Yevamos 49a) says:
“A mamzer shall not enter” — a vadai (definite) mamzer shall not enter, but a safek (doubtful) mamzer may enter.
In other words, the prohibition only covers a definite mamzer. A doubtful mamzer is allowed to marry into the community. And that raises a sharp question: how can this Gemara teach us anything about sfeika d’oraisa in the rest of the Torah — where the rule seems to be the exact opposite (strict, not lenient)?
A Second Aspect of Iyun: The Shift Between the Question and the Answer
Here is another important tool in iyun, on top of asking “why.” It is to catch the Gemara in the middle of changing its mind. First, figure out the assumption the Gemara started with — usually the simpler, more natural one. Then find the exact thing that made it give up that assumption and go a different way.
The key point is this: that first assumption (the hava amina) is often not a mistake at all. It is the plain, obvious way to read things. The depth of the sugya is in seeing what knocked it down. And very often the thing that knocks it down is a chakira — a new way of looking at the whole issue. Interestingly, different Rishonim sometimes understand this “change of mind” in different ways.
Let’s try it on our mamzer Gemara. What did the Gemara assume at first?
The natural way to read the pasuk is: “lo yavo mamzer” means what it says — a mamzer stays out. If that’s what it means, then a doubtful mamzer should be an even better reason to keep him out, not a worse one. After all, when we’re not sure, the natural instinct is to play it safe and keep him out. That is exactly the kind of case the prohibition should be catching.
There is even a technical reason that points the same way. A shetuki (the doubtful mamzer) is what we call kavu’a — his status is “fixed in place” (not like something that got mixed up and we lost track of it). A doubt that is kavu’a is treated as a full 50/50. And a 50/50 doubt in a Torah prohibition would normally be strict. So everything points one way — the plain meaning, the play-it-safe instinct, and the rule of kavu’a — all say: keep the doubtful mamzer out. That is where the Gemara starts, and it is the more logical starting point.
But the Gemara ends up ruling the opposite way: the doubtful mamzer may enter. So what was the one thing that flipped such a natural starting point?
It is a change in how we read the word “mamzer”. Instead of reading it as a loose description (that a doubtful case might sort of fit), the drasha reads it as an exact label: only someone who is definitely a mamzer is the one the Torah kept out. Once you read the word that way, the doubtful case isn’t a “lighter version” of the prohibition — it was never included in the prohibition in the first place. The doubt has moved. It used to be a doubt inside the prohibition (“is this forbidden case here or not?”). Now it is a case that sits outside the prohibition altogether. The Gemara reads “congregation” the same way: only someone of clearly-known lineage counts, so a doubtful lineage is also outside the picture.
Here is the big idea: this “change of mind” is itself a chakira — and it is the very same chakira the rest of the sugya runs on. The question is: is a safek (doubt) a weaker version of the forbidden thing — so the doubt is “inside” the prohibition and we have to weigh it? Or is it something the prohibition never covered at all — so there is nothing to weigh? The Gemara started by assuming the first, and ended by choosing the second. That is the whole move of the passage in one sentence.
And this is why it matters for what comes next. We can understand the Rambam and the Rashba as arguing about this exact move — about how far it goes. The Rambam says the Gemara revealed a general rule: a doubt, by its nature, sits outside a Torah prohibition, and the mamzer case just shows it clearly. The Rashba says the move only worked here, because there was an extra word in the pasuk that let us read “mamzer” so narrowly. Without that extra word, the natural starting assumption stays — and a safek d’oraisa is still “inside” the prohibition, and forbidden. Looked at this way, the machlokes between the Rambam and the Rashba (coming up next) is not a brand-new argument. It is a disagreement about how much of the Gemara’s own move we should apply to the rest of the Torah.
The Machlokes Rishonim: Rambam and Rashba
Two Rishonim read this same Gemara in opposite ways.
The Rashba — Forbidden From the Torah
The Rashba (Kiddushin 73a; and Toras HaBayis) writes:
It appears correct to me that when Chazal said a sfeika d’oraisa is treated stringently, this is Torah law — for a sfeika d’oraisa is like a certainty from the Torah.
So for the Rashba, the Torah itself forbids even the doubt. His well-known way of putting it: sfeika d’oraisa l’chumra mid’oraisa — a Torah doubt is strict, on a Torah level.
The Rambam — Forbidden Only Rabbinically
The Rambam (Commentary to the Mishnah; and see Hilchos Tumas Meis 9:12) says the opposite:
All these doubts that were not mentioned in the Torah — that is, that arise only from the way a verse is expounded — are permitted from the Torah. Only that which they said is forbidden regarding a certain matter is forbidden; and that which they said, that a sfeika d’oraisa is stringent, is forbidden only by their words.
So for the Rambam, a Torah doubt is actually permitted on the Torah level — it is the Sages who forbade it. His way of putting it: sfeika d’oraisa l’chumra mid’rabbanan — a Torah doubt is strict, but only Rabbinically. One important note: this whole argument is only about what the Torah level is. In practice, both agree that a Torah doubt is forbidden — the only question is whether that comes from the Torah or from the Sages.
How Each One Reads the Mamzer Case
Both bring the doubtful-mamzer Gemara, and each one turns it to support his side. The Rashba first explains how the Rambam reads it, and then answers him:
From “a doubtful mamzer may enter” the Rambam was medayek that our general rule — a sfeika d’oraisa is stringent — is Rabbinic; for were it Torah law, the Torah would not have permitted a doubtful mamzer. And so he ruled in his great compilation.
So the Rambam argues: since a doubtful mamzer is permitted, that shows every Torah doubt is permitted on the Torah level. The Rashba answers by pointing to the other side of the same case: a doubtful mamzer is permitted only because there was an extra word in the pasuk (“congregation”) to teach that heter. Everywhere else, where there is no extra word, the doubt stays forbidden on the Torah level. In other words, the mamzer is the exception that proves the rule — it needed a special source precisely because otherwise the doubt would be forbidden.
The Proof from Asham Talui
The Torah never spells out the rule of sfeika d’oraisa directly — if it did, the Rishonim wouldn’t be arguing about it. But there is a related law the Torah does spell out, and it also deals with a doubt: the asham talui, a special korban brought when someone is not sure if he sinned. The Torah says (Vayikra 5:17–18):
“If a person sins and does one of the commandments of Hashem that may not be done, and he did not know, and he becomes guilty and bears his iniquity — he shall bring an unblemished ram… as an asham to the kohen… and he shall be forgiven.”
So someone who is in doubt about a kareis-level prohibition — say, he ate a piece of meat that was maybe shuman, maybe chelev — brings an asham talui (Kereisos 17b–18a). The Rashba uses this to challenge the Rambam:
This is a strong objection on the Rambam: on his view the law of asham talui would be nullified. For the Torah obligated a korban over the doubt — and if the doubtful act were permitted outright, how could a korban be imposed for it?
The challenge is simple: if a Torah doubt is really permitted (as the Rambam says), then why does the Torah make you bring a korban for doing the doubtful act? To answer this, we have to define the Rambam’s “permitted” much more carefully — and that is exactly the kind of thing iyun is for.
Heter Vadai or Heter Safek? The Chakira on the Rambam
A good way to sharpen a definition is to test it against tricky cases. Here are two:
Case 1. Reuven knows for sure the piece is chelev. Shimon is in doubt. According to the Rambam, Shimon is allowed to eat it. But is Reuven — who knows the truth — allowed to hand it to Shimon? Or does he break “lifnei iver lo sitein michshol” (do not place a stumbling block before the blind)?
Case 2. Shimon ate it. According to the Rambam, the act was permitted on the Torah level. But then Reuven comes and tells him it was really chelev. Does Shimon now have to bring a korban to atone? Or, since he did nothing wrong at the moment he ate it, is he off the hook?
Think of it like the line between day and night. It seems obvious — until you ask about bein hashmashos (twilight), and suddenly you have to decide whether the line is really sunset or the coming out of the stars. In the same way, these cases force us to pin down what the Rambam’s “permitted” actually means. There are two ways to understand it:
Reading 1 — Heter vadai (a real, clean permission). The Torah flat-out permitted the doubt; there is no worry of any prohibition at all. So Reuven may hand Shimon the piece, and even if Shimon later finds out it was chelev, he needs no atonement — because the Torah simply permitted the act.
Reading 2 — Heter safek (permitted only because of the doubt). The Torah did not really “permit” the doubt; it just depends on how things turn out. If the piece was really shuman, he did nothing wrong. If it was really chelev, he did commit an aveirah. On this reading, a G-d-fearing person should stay away from a sfeika d’oraisa so he doesn’t stumble. Reuven may not hand over the piece. And if Shimon later finds out it was chelev, he does have to atone — because he really did eat chelev.
The Acharonim argue about which of these the Rambam means. The Shaarei Yosher (Sha’ar 1, by Rav Shimon Shkop) notes that some understood the Rambam as holding the first reading (a clean permission), but the Shaarei Yosher himself concludes the Rambam means only the second (permitted because of the doubt).
How This Answers the Rashba’s Challenge
Once we take the second reading, the asham talui problem goes away. Even for the Rambam, if the act actually hit a forbidden thing, it was a real aveirah. So the korban makes sense: it atones for the doubt, in case the person really did eat chelev. If he did, he needs atonement. (If you take the first reading — a clean permission — you need a different answer, which we’ll leave aside here.)
Issur Vadai or Issur Safek? The Parallel Chakira on the Rashba
The same kind of question runs on the Rashba’s side too. Go back to the two cases, but now with his view that the doubt is forbidden from the Torah:
Case 1. Reuven knows it is chelev; Shimon is in doubt, so for the Rashba he is forbidden on the Torah level. May Reuven hand him the piece, or does he break lifnei iver?
Case 2. Shimon ate it; for the Rashba it was forbidden. Then Reuven tells him it was really chelev — or, in the flip case, that it was really shuman. Do we say (looking back) that he needs no atonement, or that he did commit an aveirah?
Again there are two ways to read it, and they are very different:
Reading 1 — Issur vadai (a brand-new prohibition on the doubt itself). The Torah added a fresh prohibition on every doubt, no matter what the truth turns out to be. So Reuven may not hand over the piece even though it is really shuman — because for Shimon, who is in doubt, the prohibition of “sfeika d’oraisa” applies. And even if Shimon later finds out it was really shuman, he still needs atonement, because he broke the prohibition of “being in a doubt” itself.
Reading 2 — Issur safek (forbidden only as a precaution). The Torah did not add a brand-new prohibition. The doubt is forbidden only as a precaution, so he doesn’t accidentally hit the real prohibition — and it all depends on how it turns out. If it was really shuman, we look back and say he did nothing wrong, and needs no atonement. If it was really chelev, he did commit an aveirah.
Here too the Acharonim argue about which the Rashba means, and the Shaarei Yosher again concludes it is the second (a precaution). Notice what the first reading is really saying: on top of the regular prohibition — which depends on whether he actually hit chelev — there is an extra prohibition just for putting himself into a doubt, against what Hashem wants. And that extra prohibition is a sure thing: even if it turns out he never touched the real prohibition, he still did something wrong simply by placing himself in the doubt.
Cases Where All Agree
There are some doubts where even the Rashba agrees we are lenient, and others where even the Rambam agrees we are strict.
A doubtful mamzer, as we saw, everyone agrees is permitted (on the Torah level) to marry into the community. Doubtful tumah (ritual impurity) has its own source in the Torah (Sotah 28b–29a), learned from the sotah:
“V’hi nitma’ah”… “v’hi lo nitma’ah” — if she is impure why say she is not, and if she is not why say she is? This tells you that a doubt is forbidden. From the sotah we learn to the sheretz: a matter that has “da’as lishaol” (an intelligent party who could be asked) — in a reshus hayachid its doubt is impure; in a reshus harabim its doubt is pure.
The bottom line: a doubtful tumah in a public place (reshus harabim) — everyone agrees it is tahor (pure); in a private place (reshus hayachid) — everyone agrees it is tamei (impure).
The Convert Who Gave Birth — Proof for the Rashba?
Take a convert who gave birth, where we are not sure whether she gave birth before she converted (in which case she brings no korban) or after (in which case she brings one, like any Jewish woman). The Gemara (Kereisos 7b) rules that in this doubt, she does bring the korban.
Now, the Sages don’t set up a korban on their own authority — because bringing chullin ba’azarah (a non-holy animal into the Temple courtyard) is itself a Torah prohibition. So the fact that she brings a korban looks like proof that a Torah doubt is strict on the Torah level — like the Rashba.
But is this doubt really like the others? The Maharit Algazi points out that this is a problem even for the Rambam: why should a doubtful new mother bring a korban at all? He answers that the Rambam’s leniency is only about prohibitions (things you must not do). But with a mitzvas asei (something you must do), if you’re not sure whether you did it, you are obligated by the Torah to do it — because a positive mitzvah pushes aside even a doubt. So a doubtful obligation to bring a korban is a different story than a doubtful prohibition.
Doubt in a Mitzvas Asei — a Three-Way Split
The Pri Megadim takes up this same point with a tumtum and androgynos (people whose gender is uncertain), who are obligated when in doubt about time-bound positive mitzvos. He suggests that maybe the Rambam and Rashba argue about a doubt in a positive mitzvah (asei) just like they argue about a doubt in a prohibition (lav) — or maybe, in a positive mitzvah, everyone agrees the obligation is only Rabbinic.
So for a doubt in a positive mitzvah, there are three possibilities:
• Everyone agrees it is obligated on the Torah level.
• Or the opposite — everyone agrees it is only Rabbinic.
• Or even here the Rambam and Rashba argue, just like by a prohibition.
Returning to the Opening Question
Now we can answer the question we started with. Reuven is in doubt whether he bentched. Bentching is a positive mitzvah (asei), so it falls right into the three-way split above: is a doubt in a positive mitzvah obligated from the Torah (Rashba, or everyone), only Rabbinically (Rambam, or everyone), or is it itself part of the argument?
In practice, the Shulchan Aruch (Orach Chaim 184:4) rules that someone in doubt whether he bentched does bentch again — but specifically because the mitzvah to bentch after a filling meal is written openly in the Torah (“v’achalta v’savata u’veirachta”, Devarim 8:10). Since it is a clear Torah obligation, we are strict in the doubt. But where the obligation is only Rabbinic — like the berachah before eating, or a meal that wasn’t filling enough to be a Torah obligation — we are lenient, and he does not repeat it.
Putting it back into our sugya: the ruling to repeat an explicit Torah bentching lines up with being strict in a doubt about a Torah-level positive mitzvah. Whether that strictness is from the Torah itself (Rashba) or a Rabbinic safeguard on top of a Torah mitzvah (Rambam) is exactly the question all the analysis above was built to frame. That is how the poskim can fully agree on what to do in practice, while still disagreeing about where the rule comes from.
The Four Positions at a Glance
The whole sugya comes down to two questions (the chakiros) crossed with two opinions (Rambam and Rashba). The chart below pulls it all together, tested against our two cases — may Reuven hand Shimon the piece, and does Shimon need atonement if the truth comes out later.
Shita / Chakira Din vadai reading Din safek reading If piece turns out mutar
Rambam (safek mutar mid’oraisa, forbidden only mid’rabbanan) Torah permits the safek outright; no issur at all. Reuven may hand Shimon the piece; Shimon never needs kapara. Torah introduces no new issur, but liability tracks reality. Reuven may not hand it over; if it was chelev, Shimon needs kapara. Vadai: never any aveirah. Safek: no aveirah, no kapara.
Rashba (safek assur mid’oraisa — safek k’vadai) Torah adds a fresh issur on the very act of entering safek. Reuven may not hand it over even though it is really shuman; Shimon needs kapara even if it was shuman. No fresh issur; the issur is only the chashash lest he strike the prohibition. Outcome tracks reality — if really shuman, nothing was transgressed. Vadai: aveirah on the safek itself; kapara required. Safek: retroactively clarified — no aveirah, no kapara.

The Other Rishonim in This Dispute
Where do the other Rishonim stand? This matters, because their position here affects how we understand them in other places.
The Raavad holds like the Rambam: these doubts are not forbidden from the Torah, only as a Rabbinic strictness. The Ran holds like the Rashba: a Torah doubt is forbidden from the Torah — otherwise, why would we say to be strict in the first place?
For Rishonim who didn’t state their view openly, the Acharonim argue both ways. The Chiddushei HaRadal goes through the proofs — reading Rashi like the Rashba, bringing Tosafos for both sides, and lining the Ran up with the Rambam — while the Pnei Yehoshua concludes that Rashi, Tosafos, and most of the commentators hold like the Rashba.

Now that we’ve finished the sugya, it’s time to review it. Chazara (review) isn’t only helpful for later — it builds understanding right now. It shows you the foundation of the sugya and how all the details grow out of it: what is the main point, what is a side point, and how the parts fit together into one structure.
Often, a sugya we first learned out of order only falls into place when we review it — because we picked it up piece by piece, not from the beginning. (Some people find it helps to write the review out. But a good summary should be short and organized, not just a copy of everything.)
A review should cover: what the sugya is about, how it’s defined, the different opinions, where each one comes from and its reasoning, the initial thought of the Gemorah, the reason for the the proofs, whether there’s a practical difference between them, where everyone agrees and why. In our sugya, we started with where sfeika d’oraisa comes from; we laid out the two opinions and each one’s source; then we looked at each opinion on its own — the two ways to understand it and what that means in the actual cases — and finally we brought it all back to the question we opened with.