There seems to a lot of confusion and many different customs surrounding the choice of haftarah for Acharei Mot and Kedoshim, especially in Ashkenazi communities, and especially in a year like this year (5784) where the two sidrot are read separately, and each needs its own haftarah. (See, for example, the many variants listed on Wikipedia, which notes that "that there is some confusion over the correct Haftara".)

Please help me to make some sense of the variations.

A good answer should list as many customs as possible, along with as much data as possible for each custom, for example:

  • Which haftarah is read for Acharei Mot and which for Kedoshim?
  • What happens when the two sidrot are read together on one shabbat?
  • What happens when the two sidrot are read on separate shabbatot, but the haftarah for one of them is superseded by shabbat hagadol / erev rosh chodesh / rosh chodesh?
  • Historical provenance of the custom
  • In which communities is this the current practice?
  • Which rabbis (historical / contemporary) endorse the custom?
  • Rationales given for the custom
  • See chitzei gibborim vol 11
    – Double AA
    Commented May 1 at 14:34
  • @DoubleAA Thanks for the reference. Rabbi Duker is great
    – Joel K
    Commented May 1 at 15:44
  • See judaism.stackexchange.com/a/73031/21 . For the standard Ashkenazi minhag, it's basically the Ramah plus R' Akiva Eiger. We try to avoid the "Abominations of Bloody City" haftarah as much as possible.
    – Shalom
    Commented May 1 at 15:59
  • @Shalom but in a year like this year, which sidra get which of the two haftarot? Many chumashim (and some calendars) print the inverse of the Ramah’s order.
    – Joel K
    Commented May 1 at 16:03

1 Answer 1


The Babylonian annual Torah cycle portions Acharei and Kedoshim traditionally were associated with haftara readings from Ezekiel 22 and Ezekiel 20 respectively. Both those readings speak about sexual immorality and avoiding the "ways of Egypt", just as the latter half of both Torah sections do. (Notably, both readings also open with a double usage of the unusual word התשפט.) This is still the practice in most communities (Sefardi, Italian, Yemenite).

In Ashkenazi lands, for reasons yet unknown but presumably related to the very negative feel of the haftara (similar to R' Eliezer's opposition to the haftara for Shemot), they replaced one of those readings with a different reading from Amos 9 (not unlike what Ashkenazim did for Shemot, Tzav, or Behar/Bechukotai). Rashi and other early Ashkenazi authorities write that they read Amos 9 for Kedoshim, as did the Terumat HaDeshen. Indeed in most pre-printing-press Ashkenazi chumashim that's what is written: Ezekiel 22 for Acharei and Amos 9 for Kedoshim (call that E22-A9). (References to manuscript counts are from R' Eli Duker's survey in Chitzei Gibborim #11.)

The Mordechai (Megilla #831) writes that he found in a[n unnamed and not extant] book of local customs that when Acharei and Kedoshim are combined (ie. non leap years) we should exceptionally read from Amos even though it isn't the haftara of the second section. (Some authorities held to always use the haftara from the first of a doubled section, while the Mordechai and others held to generally use the second section's.) R' Isaac Tyrnau in his highly influential book of Polish customs, says based on this that the common chumashim referenced above have it backwards, and Amos 9 is for Acharei while Ezekiel 22 is for Kedoshim (A9-E22).

Now wait a second. The original order was E22-E20. Rashi read E22-A9 by doing one substitution. I can get that. But how did we get from E22-E20 to A9-E22? Logically, there's only a few options: 1) Someone (either the Mordechai, the custom booklet he found, or earlier) accidentally mixed up a record of Rashi's practice (especially since in non-leap years both sides suspiciously agree), 2) Someone, coincidentally in parallel to the development of Rashi's custom to set Kedoshim to Amos, thought to set Acharei to Amos and pushed off E22 to the next week because it was somehow more worth keeping than E20 (not unlike what happened with Behar), 3) Among pre-printing-press Ashkenazi chumashim with A9 for Acharei, about half have E20 (!) for Kedoshim (also a single substitution from the Babylonian practice). Perhaps the tradition in the Mordechai derives from a location that read A9-E20. The fact that both Ezekiel readings could be referred to in a source as התשפט certainly didn't help avoid mix-ups. Perhaps one day we'll find the Mordechai's referenced booklet and gain some clarity.

Despite being the apparent minority position (numerous authorities appeal only to the Mordechai for this and not their local custom) of unclear provenance, A9-E22 makes it via the great Polish decisor Rama into the Shulchan Arukh (OC 428:8, cf. Levush who accepts A9-E22 but suggests reading E22 when combined). Most chumashim though continued to print things the same way and we end up with the record in the halachic literature being at odds with the record in the synagogue literature. Surely many shuls who had knowledgeable leadership followed the halachic sources, but also many were clearly following what it says in their chumash (and unwittingly, Rashi). This continued for centuries and was a regular occurrence every leap year.

Around 200+ years ago, Rabbi Akiva Eiger of Poznań noted that if the Mordechai intended to override readings from Ezekiel when possible by reading Amos, the same ought apply when Acharei's haftara is superseded on Erev Rosh Chodesh (years when Pesach starts on Saturday in the Diaspora) freeing up Amos to be read the next week for Kedoshim. (For Rashi et al. the same would apply when Pesach starts on Thursday and Kedoshim is Rosh Chodesh, or in Israel when Pesach starts on Saturday and Kedoshim is Erev Rosh Chodesh.) This is a non-trivial novel extension of the Mordechai's position since in the case of a double parsha, the haftara of the first section is still related to something in that day's reading, whereas here Amos is not (for this position) at all related to Kedoshim. With the contemporary spread of reading Malachi on Shabbat Hagadol in leap years, and not just on Erev Pesach, (eliminating years starting on Thursday) we're left with just one year type (leap year starting on Shabbat with Pesach on Tuesday) where both Acharei and Kedoshim are available for their own haftarot, and thus in the modern day this has become quite rare indeed.

The rarer it gets the less likely people will insist on sticking to their traditional practice (if they even remember this detail ~20 years later). In the beginning of the 20th century R' YM Tucazinsky and R' YE Henkin in their influential guides to synagogue service in Israel and the USA respectively maintained the order of the chumashim (and Rashi) despite various rabbinical critiques on the basis of prominent printed halachic works. There was a big brouhaha in Jerusalem when R' Tucazinsky's guide was first published (and some there (cf. Even Yisrael 8:38) went beyond extreme and read Amos two weeks in a row to avoid Ezekiel entirely, against every inch of precedent!) R' Henkin's successors eventually gave in after his death and changed the recommendation in the newer Ezras Torah Luach. Finally in the last ~50 years ArtScroll actually went and switched the order in their printed chumash based on the Mishna Berurah.

And still many other modern chumashim maintain the old practice. That Rashi's original Ashkenazi practice at all survived centuries of prominent rabbis trying to stamp it out is nothing short of a miracle.

  • 2
    Tour de force ... bravo!
    – Shalom
    Commented May 6 at 3:01
  • It seems that some chassidic communities (notably Chabad) read A9-E20 (and A9 when acharei and kedoshim are combined). I assume this is an innovation to harmonise between Ashkenazi and Sefardi practice, rather than a preservation of the custom you described from many of the pre-printing press chumashim.
    – Joel K
    Commented May 6 at 5:50
  • Did anyone ever read E22 for Kedoshim when Acharei Mos didn't get a Haftorah (IE A9 is for Acharei, and rejecting the "non trivial extension")
    – AKA
    Commented May 6 at 8:36
  • 1
    @AKA R Isaac Tyrnau writes explicitly that he believes that one should do precisely that. If Acharei falls on Erev RC Iyar, you skip A9 that year, and read E22 for Kedoshim.
    – Joel K
    Commented May 6 at 9:11
  • 2
    @AKA Yes, that's the Rama. That's what I meant by "was a regular occurrence every leap year" Every leap year you had to decide which belonged where since there's always at least one that needs its haftara. Even RAEiger only makes his comment as a suggestion not a description of local practice.
    – Double AA
    Commented May 6 at 11:34

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .