Thursday, March 01, 2012

Summary of Jordan Penkower on reading Zekher or Zeikher Amalek

Summary of Jordan Penkower on reading Zekher or Zeikher Amalek

This post originally appeared here. It is a summary of Prof. Jordan Penkower's Minhag u-Mesorah: 'Z-kh-r Amalek' be-hamesh o be-shesh nekudot by his student Yosef Peretz. It appears here with permission, with some additions by the Seforim Blog editor.

It is common practice in most Ashkenazi congregations to read the words z-kh-r Amalek, in Deut. 25:19, twice: once with a tzere under the zayin (zeikher) and once with a segol (zekher). Sephardic congregations, however, who do not distinguish between the pronunciation of these two vowels, read it only once, and they point it with a tzere. What is the origin of this practice, when did it begin, and is this double reading necessary to fulfill the command of reading these verses of Parshat Zakhor, the special reading on the Sabbath before Purim?

These questions are discussed in several articles by R. Mordechai Breuer and Dr. J. Penkower,[1] two of the most eminent among contemporary scholars of the masorah and Biblical text. In their opinion, zeikher with a tzere is beyond a shadow of doubt the original and correct pointing of the text; and that is how the verse should be read in Parshat Zakhor. It must be stressed that the custom itself of a double reading is quite surprising and completely unique in Torah reading, for it was customary to decide in favor of one reading or another whenever there was a conflict between variant readings, pointing of vowels, or assignment of cantillation marks (as in cases of kri and ktiv, where a word is written one way but read another).

R. David Kimhi, in his Sefer ha-Shorashim (in the manuscript versions), was the first to note the discrepancies regarding the pointing of these vowels in Sephardic biblical manuscripts. Under the root z-kh-r he says: "'blot out the memory (zekher) of Amalek' (Deut. 25:19), with six dots (i.e., segol twice), but 'praise [in remembrance of (le-zekher)] His holy name' (Ps. 30:5), with five dots (i.e., a tzere and a segol), and this occurs nowhere else; thus it appears in some books [i.e. manuscripts of the bible containing the Masorah]. But in others, z-kh-r is always pointed with five dots." In other words, Radak noted that in some books he found z-kh-r pointed with segol and in others, with tzere.

In printed editions of Sefer ha-Shorashim, however, beginning with the Venice edition and carrying through the 19th century, Radak's remark concludes with the words "and this occurs nowhere else." In other words, he found zekher in Parshat Zakhor always pointed with double segol.[2]

Here is a typical example in a manuscript (from 1481):

And here is the Venice edition:

Here is the manuscript text restored in Biesenthal and Lebrecht's Berlin 1847 edition:

Various prayer books and Pentateuchs published in the 16th, 17th and early 18th centuries were redacted according to Radak's remark as it appears in these printed editions; for example, the siddur of R. Shabtai Sofer (in the word zekher in the Ashrei prayer and in the prayers on the eve of the New Year), and Meor Einayim, the Pentateuch published by R. Wolf Heidenheim.*

The Basel 1579 Seder Tefillot Mi-kol Ha-shanah (Ashkenaz):

The manuscript of R. Shabtai's siddur:

In 1832, about fifteen years after publication of Heidenheim's edition, a book by R. Issachar Baer appeared, entitled Ma`aseh Rav, containing a description of the practices of the Vilna Gaon. The author mentions that the Gaon's disciples disagreed over the way their teacher used to read the word zekher in Parshat Zakhor. The author attested as follows: "When he would read Parshat Zakhor, he would say zekher, with a segol under the letter zayin. R. Hayyim of Volozhin, however, whose endorsement is on the book, wrote there, `As for his writing that in Parshat Zakhor one should read [z-kh-r] with six dots, I heard from the saintly person [i.e., the Gaon of Vilna] that he read with five dots (=zeikher).' I do not know whether those hearing him were mistaken, and thought they heard segol twice, or whether he changed his practice in his later years."

Here is the text, with the footnote referring to R. Hayyim of Volozhin's testimony:

Here is R. Hayyim's testimony in his approbation at the beginning of this work:

Approximately eighty years later, R. Israel Meir Ha-Cohen, otherwise known as the Hafetz Hayyim, came out with his Mishnah Berurah on the Shulhan Arukh. Since the special reading of Parshat Zakhor is a command from the Torah, and there was some doubt as to the correct reading, he ruled that the words z-kh-r Amalek should be read twice. In his own words, "Some people say it should be read as zeekher Amalek (Deut. 25:19) with a tzere, and others say that it should be read as zekher Amalek with a segol; therefore the correct practice is to read both ways, to satisfy them both" (Mishnah Berurah 685.18). Later, a variety of customs emerged in this regard. Some readers only repeat the two words, z-kh-r Amalek, while others repeat the entire phrase, timhe et z-kh-r Amalek, and still others repeat the entire verse.[3]

Here is the Mishnah Berurah:

As mentioned above, Rabbi Breuer and Dr. Penkower note that the correct pointing of z-kh-r is with a tzere and the custom of double reading is unfounded. This follows from the arguments they cite from the masorah, from the ancient and highly precise Tiberian manuscripts and teachings of the Masoretes.

R. Jedidiah Solomon Norzi, a seventeenth century masoretic scholar, held to be the final arbiter on the text of the Bible,[4] wrote a work entitled Minhat Shai, in which he remarks on inaccuracies that entered all the books of the Bible in the 1547-1548 Venice edition of Mikraot Gedolot and other editions published around then.[5] He says nothing about zekher ,which is pointed with a tzere followed by segol, from which it follows that he agreed with this pointing and had no doubts about it being correct.[6]

Minhat Shai from the first edition (Mantua 1742-4):

The most conclusive proof is found in ancient manuscripts, dating to the time of the masorah, and held to be very precise in their pointing and cantillation marks: the Leningrad manuscript (known as B19),[7] the Sassoon 1053 manuscript and others. All point z-kh-r with a tzere under the zayin.[8]

Leningrad manuscript:

Today we have in our hands a famous ancient manuscript, the Keter Aram-Tzova,[9] the most ancient and best authorized text of the entire Bible. The pointing and masoretic annotation of this manuscript were done in Israel over one thousand years ago by Aharon Ben-Asher, considered the greatest Masorete of all generations. Due to Ben-Asher's precision and reputation, Maimonides chose to base his Hilkhot Sefer Torah on Ben-Asher's work.[10] In the 15th century, at the latest, the Keter manuscript was transferred to Aleppo, Syria, where it was stored in the Sephardic synagogue until the riots against the Jews of Aleppo (Aram Tzova) that broke out in 1948, during which the manuscript was damaged, most the entire Pentateuch being lost, including Parshat Zakhor.

Because the Keter manuscript was so special, the Jews of Aleppo did not allow others to photograph the manuscript or to examine it. Whoever wished to clarify a question of variants in the text had to write down the query and the person in charge would relay the version found in the Keter. The most famous of those addressing queries was R. Jacob Saphir, who submitted over five hundred questions, seeking to find out what variant appeared in the Keter. Fortunately, one of his questions pertained to the pointing of z-kh-r in Parshat Zakhor. The question and response were as follows: "(Deuteronomy) 25:19 z-kh-r h"n [question]. Yes [response]." In other words, is the word z-kh-r pointed in the Keter with five dots [h"n = hamesh nekudot, five dots,i.e., tzere followed by segol]? The answer provided by the keeper of the manuscript, R. Menashe Sitton, was" yes".[11]

The entry in the published version of this manuscript (Rafael Zer Meorot Natan Le-rabbi Ya'akov Saphir, Leshonenu 50:3,4 Nisan-Tamuz 5746):

Dr. Penkower examined the pointing of z-kh-r in dozens of medieval manuscripts and found that in those manuscripts reputed to be more precise (some of the Sephardic ones) it was pointed with a tzere, and in those reputed to be less precise (some of the Sephardic ones and most of the Ashkenazi ones) it was pointed with a segol. These findings led him to note, "If we were to start taking into account the pointing in manuscripts far removed from the precise Tiberian ones, and were to begin reading doubly all instances of variation between them and the precise manuscripts, the Torah reading each week would last an inordinately long time."[12]

All editions of the Bible today point z-kh-r with tzere followed by segol, leaving no vestige of the double segol variant. One cannot but agree with R. Breuer and Dr. Penkower that there is no need for Ashkenazim to read zekher, for zeikher will suffice.

To sum up:

* R. David Kimhi mentions two methods of pointing which he observed in Sephardic manuscripts: zekher and zeikher.

* The disciples of the Vilna Gaon disagreed about how their Rabbi used to read this word in Parshat Zakhor, whether with a tzere or a segol.

* Because of uncertainty as to which was correct, the Mishnah Berurah ruled that z-kh-r Amalek should be read twice, once with tzere and once with segol.

* The findings presented by R. Breuer and Dr. Penkower prove conclusively that the correct and original pointing of this word is zeikher (with a tzere). Therefore, in their opinion, one should return to the ancient practice, and all Jewish communities ought to read the word only once, as zeikher.

[1] M. Breuer, "Mikraot she-yesh lahem hekhre`a," Megadim 10 (1990), pp. 97-112; J. S. Penkower, "Minhag u-Mesorah: 'Z-kh-r Amalek' be-hamesh o be-shes nekudot" (with appendices, in R. Kasher, M. Tzippor and Y. Tzefati, eds., Iyyunei Mikra u-Farshanut, 4, Bar Ilan University Press, Ramat Gan 1997, pp. 71-127.

[2] Penkower (loc. cit., p. 80 ff.) discusses at length the differences between manuscript and printed versions of Sefer ha-Shorashim.

[3] On other practices, see Penkower, loc. sit., p. 71, n. 1.

[4] Y. Yevin, Mavo la-Masorah ha-Tverianit, Jerusalem 1976, p. 101 ff.

[5] J. S. Penkower, Ya`akov Ben Hayyim u-Tzmihat Mahadurat ha-Mikraot ha-Gedolot I-II (Dissertation), Jerusalem 1982.

[6] R. Jedidiah Solomon Norzi was preceded by another important masoretic scholar, R. Menahem de Lonzano, author of Or Torah. In his book, he remarks on the same editions of Mikraot Gedolot on the system of vowel pointing and cantillation marks, but also says nothing about z-kh-r being pointed with a tzere followed by segol.

[7] The printed Bible published for the IDF, prepared by Prof. A. Dothan, is based on this manuscript.

[8]For further detail, see R. Breuer's article (cited in n. 1), p. 110, and Penkower's article (cited in n. 2), p. 101.

[9] This manuscript is also known as the Aleppo Keter, or simply the Keter for short.

[10] Several printed Bibles based on the Keter are available today. The most important of these is undoubtedly the Mikraot Gedolot- ha-Keter, redacted and edited by Prof. Menahem Cohen, Bar Ilan University Press, Ramat Gan 1992 and following years. Thus far five volumes have been published, covering the following six books of the Bible: Joshua-Judges (including a general introduction to the edition), Samuel I and II, Kings I and II, Isaiah, and Genesis vol. 1.

[11] The manuscript containing these questions and responsa is known as Meorot Natan. See here.

[12] At the Project on Bible and Masorah at Bar- Ilan University, headed by Prof. M. Cohen, dozens of medieval manuscripts of the Bible were examined, and the findings of these studies reinforce Penkower's conclusions regarding variant pointings in the manuscripts.

* Seforim Blog editor note: the issue is discussed in the En Hakore commentary included by Heidenheim in Humash Meor Einayim, only in the parashah of Amalek in Exodus 17, not in Deuteronomy. En Hakore is a masoretic commentary by R. Yekutiel (or Zalman) Ha-nakdan who lived in Prague in the 13th century. Here is how it appears in Heidenheim's Humash Meor Einayim.

However, it appears to be a misstatement to say that Heidenheim was among those whose work was "redacted according to Radak's remark," for while there is no example of Zekher from Psalms to compare with in his Humash Meor Einayim, Zekher is also pointed with tzere , rather than segol, in the Ashrei prayer included in Humash Moda Le-vina. This would seem to indicate that he did not follow the Radak. He merely published En Hakore, which dealt with the issue raised by Radak.


  1. are the comments lost?

  2. "<span>it's great acclaim was but another product of the infamous Russian propaganda machine."</span>

    That sounds a bit ridiculous.

  3. The fact that we are not usually concerned with alternative nekudos is most likely due to the fact that most cases of Torah reading do not involve mitzvos m'doraisa; this is a good reason not to be machmir. It would be worthwhile to check is there are "accepted" differences in other sections which involve a mitzvah m'doraisa, like birchas kohanim or parshas chalitzah; according to the Mishna Berurah's reasoning, one would probably have to be machmir there as well.

  4. The question also ought to consider whether there is a real safek or not. I think that's what Penkower et al. are addressing.

  5. It's worth pointing out R. Meshulem Roth's astute remark that there is a rare form of smikhut which reads segol-segol rather than sheva-patach. Thus, in the story of Yoav and his teacher, it is possible that rather than reading 'zekhar amalek,' (sheva-patah) understood to mean 'the males of Amalek,' the teacher had 'zekher amalek' (two segols) which he mistakenly took for the rare smikhut form.

    The implication of this might be:

    1) we certainly should not consider zekher (6 points) a possible reading, since the Gemara records this terrible story which resulted all because of that reading.
    2) on the contrary, this can serve as "proof" that zekher (6 points) is an or the ancient reading. The point of the story is only to understand the pshat correctly, it means 'memory' and not 'males.'

  6. Ari Kinsberg12:35 PM


    I looked again at the section you psoted. I see now in the first paragraph where he makes a reference to the 1525 Mikraot Gedolot as being "yoter meduyak." Just note:
    1) This 1525 edition would be the one I referred to above that was published by a Protestant and edited by a meshumad.
    2) The 1525 Mikraot Gedolot is not a bad edition, but it certainly is not the gold standard.
    3) It's unclear what Rav Belsky means when he continues "u-muga al pi ha-mikhlol ve-ha-minhat shai."
    A) Mikhlol is a dikduk textbook, not a masoretic treatise. (The Minhat Shai himself was greatly indebted to Radak, yet did not hesitate to reject him in favor of masoretic evidence.)
    B) Minhat Shai was prepared in large part as a collection of corrective glosses to the Mikraot Gedolot of 1547 (published by a Protestant and edited by yet another meshumad). So I'm not sure what the reference is to a connection between the 1525 Mikraot Gedolot and Minhat Shai. And as far as the utility of later Mikraot Gedolot editions published by Jews who had access the works by Minhat Shai and others, it isn't clear (to me at least) to what extent they actually corrected the texts.

  7. akiva2:58 PM

    That is definately correct; however, there are differences between early masoretic texts as well, and that is were the issue should apply, if it has application to the average torah reading. Note the comment in the article, "These findings led him to note, "If we were to start taking into account the pointing in manuscripts far removed from the precise Tiberian ones, and were to begin reading doubly all instances of variation between them and the precise manuscripts, the Torah reading each week would last an inordinately long time". This seems to imply that Dr Penkower was not taking into account the issue that I mentioned in his analysis. As an aside, I do not know which things he would consider to be a real safek. Also, I was directing my comment at the second paragraph of the article, which reads "It must be stressed that the custom itself of a double reading is quite surprising and completely unique in Torah reading, for it was customary to decide in favor of one reading or another whenever there was a conflict between variant readings, pointing of vowels, or assignment of cantillation marks (as in cases of kri and ktiv, where a word is written one way but read another)." I was explaining why this should be a unique halacha in terms of k'rias hatorah (but see Tosafos HaRosh, Ri, Rabbeinu Peretz, etc. in Berachos 13a).

  8. Ari Kinsberg2:59 PM

    Very, very rare. I think he cites עשן but I think there is also הדר. Both appear in semikhut with the conventional and rare forms. Unfortunately זכר (males) is unattested biblically at all in semikhut. Perhaps someone has information from a cognate language?

    "2) on the contrary, this can serve as "proof" that zekher (6 points) is an or the ancient reading. The point of the story is only to understand the pshat correctly, it means 'memory' and not 'males.'"

    Please clarify. Thanks.

  9. David Zinberg3:09 PM


  10. Mar Gavriel1:54 PM


  11. Yehuda H.6:10 PM

    Given that the Ashkenazi mesorah is different than the Tiberian mesorah, and since the majority of the Ashkenazi mss. write with six dots, could that not be viewed as a possible legitimate source for six dots?

  12. Mar Gavriel3:00 AM

    Yehuda: By that logic, you should also spell your Torah scrolls differently, in hundreds of places -- or at least have a special non-Tiberianly-spelled Torah scroll for Shabbath Zakhor.

  13. Yehudha4:27 AM

    "who do not distinguish between the pronunciation of these two vowels"

    of course we do

  14. The late Lubavitcher Rebbe (in the name of his father-in-law?) once quipped on this controversy: "Zeicher, Zecher - abi opmekn!" ("Zeicher, Zecher - as long as [Amalek] is wiped out!"). [mentioned in Sefer haMinhagim -Chabad] (The debate in Chabad was, of the three times we read parshat zachor, when do you read with a tezre first & when with a segol first - not either/or.)

    I personally heard when the Lubavitcher Rebbe prayed at the amud, that when it came to Ashrei "Zecher rav...", he would mumble both Zecher & Zeicher - but this was not an instruction for the public.

  15. Ari Kinsberg1:55 PM

    MAR G.,

    But of course our sifre torah are not 100% Tiberian either. Someting I wonder about is how those (myself included) who are makpid about following the Tiberian textual mesorah in some issues, conveniently ignore it in others in other issues. For example, do you know any Tiberianite who insists on attending only a Yemenite minyan for parshat no'ah (ויהי vs. ויהיו)?

  16. Ari Kinsberg2:02 PM


    I'm not sure what you mean by "accepted differences," but yes there are differences in the hilufim lists compiled by R. Breuer, Ginberg, Biblia Hebraica, etc. (Btw, if you consider parshat para to be deoraita, if I remember correctly לאמר at the beginning in the 1524-5 Mikraot Gedolot has a segol rather than a tzere.)

  17. akiva4:38 PM

    Thank you. By "accepted", I mean differences which have halachic validity, not ones that can fairly definitively traced to a printer's error or something similar. The issue is more pressing for chalitzah and the like, which are clearly d'oraisa. The opinion of Tosafos in regards to parshas para may be a chumra, but I don't think that it is the generally accepted opinion, so for k'rias hatorah it would be a chumra on top of a chumra. In truth, even parshas zachor needs thought; it is not necessarily true that a sefer torah is needed (see Tosafos Ber. 13a, etc). Therefore, it should not be any different than yetias mitzrayim, which does not have any specific "nussach". If there is no difference in meaning, either way should be fine for zachor purposes, even if incorrect in terms of k'rias hatorah. As an aside. in a later comment you mention being makpid on the Tiberian mesorah. If this is not just a question of pronunciation, why don't you always need a Tiberian sefer for k'rias hatorah, at least l'chatchila? (Not a challenge, just a question.) Thanks again for the information.

  18. bniderlaw@gmail.com5:09 PM

    אין זכר לזכר

  19. Mar Gavriel5:48 PM


    It wouldn't just be for Parashath Noah -- it would be always. If ויהי is wrong, then the Torah scroll is pasul always, not just on the week when that word is read. (And yes, I corrected my personal Torah scroll to read ויהיו.)

  20. Oh, I see the comment I posted on Echo showed up on Blogger. Nice.

  21. Yeedle8:38 PM

    I see that the comments posted on echo show up on blogger, but not vice versa.

  22. Anonymous11:32 PM

    Can someone please explain what R' Chaim means with "אבל לגבי דידי מילתא זוטרתי הוא כי כן נתפשטו ונדפסו הנהגות מהרי"ל ע"י שמש קטן ששימש אותו". thanx.

  23. Anonymous6:04 PM

    Theoretically you are correct. However, I think that halachically it has to be assumed that we do not have full knowledge of plene and defective spellings, and thus all these differences are not me'akev bedieved.

  24. A. Schreiber11:36 AM


  25. JYTTdfdf2:11 PM


  26. A. Schreiber2:12 PM


  27. A Schreiber2:23 PM