For many years, most communities followed the haftarah readings as given in the Hertz and the Soncino (Cohen) chumashim. While these showed variations in practice between Ashkenazim and Sephardim, they didn’t show any alternative minhagim, and most people were unaware that there even were any.
This position began to change with the publication of the Artscroll chumashim and machzorim in the 1990s which gave different readings for a number of haftarot. There was initially a recognition that Artscroll represented a different set of traditions to those of Minhag Anglia, and conflicts were usually resolved in favour of Hertz/Soncino.
However, over the last 25 years or so, there has been a gradual incorporation of what we might politely call Minhag Artscroll into our own liturgy and practice, as can be seen in both the multiple revisions to the Singer’s siddur and also in the differences between the venerable Routledge machzorim and the more recent Koren Sacks Minhag Anglia series.
This trend has also affected the accepted texts for a number of our haftarah readings. Specifically, there has been an apparently uncoordinated churn of recent publications emanating from Koren, all of which are supposedly authoritative, but which in some respects not only suggest different readings from those given in the Hertz/Soncino chumashim but which are also somewhat inconsistent with each other.
The readings given here mostly follow those given in the Koren Sacks Tanakh of 2021 (Magerman edition) which appears to be the best source showing where Minhag Anglia differs from Minhag Ashkenaz (though there do remain some ambiguities). You can find the full list on the Koren Publications web site; just click the link to look inside the Tanach, and scroll down to page xxix.
There are four haftarot in particular which are likely give rise to the most confusion when members of a congregation are using different texts, making it likely that some will be unable to follow what is read. This is especially the case where the actual starting point is unclear or where the specified reading consists of non-contiguous sections for which there are different traditions. The best solution at the moment may be for a community to explain the problem before announcing which minhag will be followed.
The haftarah reading which is most seriously affected is that for Shabbat Shuvah. The Hertz and the Soncino (Cohen) chumashim give the reading as comprising three sections: Hosea 14:2-10 followed by Micah 7:18-20 and ending with Joel 2:15-27.
The Artscroll Stone chumash, published in 1996, had a different order for the three sections; instead of Hosea-Micah-Joel they gave Hosea-Joel-Micah, and the selection from Joel began four verses earlier, at Joel 2:11 rather than Joel 2:15. There were notes stating that customs varied regarding what was read and in what order, but none of these corresponded to the reading as given in the Hertz/Soncino chumashim.
The differences in order together with the addition of the four extra verses at the start of the section from Joel gave rise to confusion when the Artscroll chumash became widely used alongside the Hertz and Soncino chumashim. If the haftarah were leyned according to Minhag Hertz/Soncino, people using the Artscroll chumash needed to be agile readers if they wanted to follow what they were hearing; and if haftarah were leyned according to Minhag Artscroll, people using the Hertz/Soncino chumashim found it quite impossible to follow, no matter how agile their reading, as the four extra verses at the start of the section from Joel were completely missing in their books.
As noted above, the recent publications from Koren have just added to the confusion as to what we should be reading for haftarah Shabbat Shuvah:
The Koren Sacks Tanakh states that Minhag Anglia conforms to the Hertz/Soncino reading. So if we follow this, we’d read Hosea 14:2-10 followed by Micah 7:18-20 and end with Joel 2:15-27
The Sacks Siddur of 2023 (United Synagogue edition) includes end notes on Minhag Anglia which agree that the reading from Joel should begin at 2:15, but also state that Joel should be read before Micah. So if we follow this, we’d read Hosea 14:2-10 followed by Joel 2:15-27 and end with Micah 7:18-20.
The Koren Sacks Chumash of 2025 (Conrad Morris edition) includes a preface on Minhag Anglia which states that Joel should be read before Micah, but the text given in this chumash begins the reading from Joel at 2:11; just like Artscroll, it starts with four verses that aren’t found anywhere in Hertz/Soncino. So if we follow this, we’d read Hosea 14:2-10 followed by Joel 2:11-27 and end with Micah 7:18-20. Regrettably, the Koren Sacks Chumash lacks any indication that there might be alternative minhagim.
The Hertz chumash has a haftarah text from Kings II 11:17-12:17, with footnotes stating that while Sephardim begin at Kings II 11:17, Ashkenazim begin at Kings II 12:1. The Soncino chumash has the same text, but with footnotes only in Hebrew. Perhaps because all these footnotes are very small, many people have either ignored or not noticed them, which may be why both the Koren Sacks Tanakh and the Koren Sacks Chumash state that Minhag Anglia is now to read from II Kings 11:17 -12:17. However, the Minhag Anglia notes in the 2023 Siddur disagree, stating that we should read II Kings 12:1-17.
The Hertz chumash has a haftarah text from I Samuel 15:1-34, but with footnotes (easily missed because they aren’t on the first page) stating that Ashkenazim begin reading from I Samuel 15:2. The Soncino chumash has the same text, but with a footnote only in Hebrew. Once again, many people have either ignored or not noticed these small footnotes, which may be why both the Koren Sacks Tanakh and the Koren Sacks Chumash state that Minhag Anglia is now to read I Samuel 15:1 -34.
The options for this haftarah are possibly more confusing than those for Shuvah, because as well as different chumashim, different machzorim are added into the mix:
The text as originally given in the Routledge machzor was that we read Joshua 5:2-6:1.
While the Hertz chumash also gives the text as Joshua 5:2-6:1, following his practice of catering for Sephardim as well as Askenazim, Hertz included the single verse addition of Joshua 6:27 at the end with a footnote stating it is only read by Sephardim.
Once again, it would seem that Minhag Anglia may have evolved by ignoring the footnotes in Hertz, as the Koren Sacks Tanakh tells us that we read Joshua 5:2-6:1 and add Joshua 6:27 at the end (though it does not explicitly state that this is Minhag Anglia). There is also a bracketed note stating that some begin by reading Joshua 3:5-7 but without stating whose minhag this is.
The minhag of beginning with Joshua 3:5-7 would appear to be an Artscroll import, as both the Artscroll chumash and Artscroll Pesach machzor show this haftarah as comprising three sections, starting with Joshua 3:5-7, continuing with Joshua 5:2-6:1 and ending with Joshua 6:27.
The Koren Sacks Pesach machzor gives no guidance as to what should be read. It starts with Joshua 3:5-7, but states that only some read those three verses. It continues with Joshua 5:2-6:1, but states that only some end at Joshua 6:1, and that some others finish with the single verse Joshua 6:27. Nowhere does it state what reading corresponds to whose minhag.
Finally, the Koren Sacks Chumash of 2025 (Conrad Morris edition) starts with Joshua 3:5-7 and continues with Joshua 5:2-6:1. There is no indication of any alternative minhagim and Joshua 6:27 is not included.