Chullin 104a is a sugya about meat and milk — whether poultry cooked in milk is forbidden by Torah law, and whether meat and cheese may be brought up onto one table together. The Rebbe never once discusses any of that. Both of his citations go to a single question the Gemara asks in passing, and to what Tosafos does with it.

The Gemara's question is this: and from where do you say that we do not issue a decree upon a decree?^{חולין ק״ד, א} Rav Yosef had just relied on that rule to prove something, and the Gemara wants to know where the rule comes from.

Tosafos' answer is more unsettling than it first appears. He lists places where the Talmud does invoke the rule — and then concedes that there are also places where the Sages plainly did decree upon a decree, and we are not troubled by it — for one may not liken the decrees of the Sages one to another, except in those places where the Talmud itself likens them^{תוד״ה ומנא, חולין ק״ד, א}.

So the rule survives not as a limit on rabbinic authority, but as a limit on yours.

In both places he reaches for the same eight words — Tosafos', not the Gemara's — and in both places he uses them to shut down a conclusion someone would find convenient. The difference is in how he holds them. In 5716 he quotes Tosafos exactly: אין לדמות גזירות חכמים זו לזו^{תוד״ה ומנא, חולין ק״ד, א; ומובא בכרך טז, אות י}. In 5719 he does not quote at all — he states the principle in his own words, אין למדים גזירה מגזירה^{כרך כה, אות לה}, and his footnote opens with ראה, sending the reader to this Tosafos and to a second one in Niddah. One is a quotation; the other is a citation of a principle.

Shavuos 5716 — invoking it. One might reason, he says, that fences and safeguards belong only to Torah law: since rabbinic enactments are themselves decrees, there is no room to decree a further decree upon them. He answers that this is simply not so — the Sages made two decrees in many places without their counting as a decree upon a decree, כולה חדא גזירה היא^{כרך טז, אות י}, and אין לדמות גזירות חכמים זו לזו^{כרך טז, אות י}.

And the reason is not procedural. Rabbinic law is not a fence around the Torah standing at some lower grade of seriousness; per the Rambam, everything the Sages said to do and everything they warned against, Moshe our teacher was already commanded at Sinai to command us to keep^{הקדמת סהמ״צ, שורש א׳, ומובא בכרך טז, אות י}, on the strength of לא תסור. Both were given at once — כולם ניתנו מרועה אחד^{כרך טז, אות י}. Where the law does distinguish rabbinic from Torah obligations, that is the Torah's own legislation, not because rabbinic laws are weaker^{כרך טז, אות י}. From which it follows that where no fence was built, nothing may be inferred about how grave the matter is — and where a fence was built on a rabbinic enactment, that fence carries full weight: עשו משמרת למשמרתי^{מו״ק ה, סע״א, ומובא בכרך טז, אות י}.

Purim 5719 — withholding it. Three years later he asks why Purim never received a second day for the doubt of the calendar, as the festivals did. He surveys the Rishonim, and comes to the answer that a second day was instituted only for Torah festivals, not rabbinic ones — הם אמרו והם אמרו. And he rejects it, on exactly this daf:

Had the second day of the diaspora been a matter of decree, without a reason, then one could say that here they decreed and there they did not — along the lines of the principle that one does not derive one decree from another. But since it is a matter grounded in reason and understanding… then by that same reason that they established a second day for all the festivals, they should have established one for the days of Purim as well. The Rebbe · תורת מנחם כרך כה, ע׳ 132, אות לה

Same principle, opposite service. In 5716 he invokes it to forbid an inference of leniency. In 5719 he withholds it to forbid an inference of arbitrariness.

1. The rule is epistemic, not substantive — and reading the Gemara alone hides that. On the daf, לא גזרינן גזירה לגזירה functions as a premise in an argument: Rav Yosef leans on it to prove poultry-in-milk is biblical. It reads like a fact about what the Sages may do. Tosafos' answer dismantles that: they demonstrably did decree upon a decree in some places, and the rule survives only as a prohibition on comparing — אלא במקומות שהש״ס מדמה, only where the Talmud itself draws the comparison. It is a constraint on the reasoner, not on the legislator.

2. Which means it cuts both ways — and that is the half nobody uses. Because the rule is phrased as a limitation, it is naturally heard as a source of leniency: don't extend decrees. But a bar on analogy is symmetric. If you may not reason from one decree to another, then you also may not reason from a decree's absence to the lightness of the underlying matter. That second direction is exactly the move of Shavuos 5716, and it is available in Tosafos without being stated there.

3. Purim 5719 supplies the criterion Tosafos leaves open. Tosafos says comparison is permitted where the Talmud compares — which tells you when you may compare, but not what makes two decrees comparable in the first place. The Rebbe adds the test: the principle governs a decree ללא טעם^{כרך כה, אות לה} — a bare enactment with no reason given. Where the enactment rests on articulable reasoning, the principle is unavailable and consistency is owed. That is a real addition to Tosafos' rule, and the entire Purim argument rests on it: because a second day exists for a stated reason, its absence at Purim cannot be shrugged off.

The absence of a rule is not evidence that the matter is small. This is the plainest consequence and the easiest to get backwards. Where no one built a fence, the honest conclusion is that no fence was built — not that nothing was at stake. The Shavuos sicha spends its length establishing that rabbinic law is not a diminished Torah, and the practical edge is that you cannot rank your obligations by how much protective machinery has accumulated around them.

Where you can state your reason, you owe consistency. This is the sharper one, and it is the Purim move turned inward. A bare decree may be uneven; a reasoned practice may not. So when you find yourself keeping something carefully in one setting and loosely in another — and you can articulate why you keep it at all — that reason has already obligated you in both places. The exemption you are reaching for is only available to a rule that has no reason, and your rule has one.

And a question to put to the phrase “that's different.” The whole force of this daf, on the Rebbe's reading, is that אין לדמות is a disciplined refusal to compare where comparison is unwarranted — not a licence to declare any two cases incomparable whenever the comparison is uncomfortable. Tosafos permits it only where the Talmud itself compares. So the question is whether you are pointing at a real distinction, or borrowing a rule about the limits of analogy to protect an inconsistency you would rather not examine.