“Due to the great multitude of our sins, in the past decade, a leprosy has spread among the cantors.”
You know when a halakhic authority starts a sentence this way, you better buckle up – because you are in for a treat. Whatever could this terrible sin be?
R. Yehiel Mikhel Epstein (Arukh Ha-Shulhan 338:8) relates that much to the chagrin of the Rabbinic establishment, the prayer leaders had begun to employ the use of tuning-forks (arguably, a form of pseudo-instrumentation) into their Sabbath cantorial routine. He continues: “Due to our terrible sins, our generation is loose and the masses support these cantors. Not only are we unable to protest, but even exiting the synagogue causes a fight, as is known.”
Not only were these cantors culpable of adding instruments to accompany the Sabbath and holiday liturgy, but they were also guilty of adding lyrics:
Further, regarding the practice of saying words, and repeating them twice and three times, and spreading notes before the platform to sing in the style of a performance – all who have awe of heaven are pained by this, and they cannot protest, for the masses are undisciplined, and they will not listen to the words of the sages in this matter! They say that this is their enjoyment of Shabbat and Yom Tov!
This is a very concerning allegation made by R. Epstein, as the practice of repeating words while singing prayers was not only prevalent in his community, but I have also observed it in contemporary synagogues of various backgrounds. With the High Holidays approaching, a time iconic for its songful liturgy, this phenomenon deserves closer scrutiny. What are the issues with repeating words during prayers, and can we find a way to justify what many seem to view as normative practice?
This essay outlines seven potential issues with repeating words from our liturgy. We will begin by addressing the technical halakhic issues before turning toward broader philosophical considerations.
(1) This discussion begins with the Mishnah and Gemara in Tractate Berakhot (33b). The Sages were concerned that in certain circumstances the repetition of words from our liturgy could bear a heretical connotation:
Mishnah: If one [in praying] says “may thy mercies extend to a bird’s nest,” “be thy name mentioned for well-doing,” or “we give thanks, we give thanks,” [modim, modim] he is silenced.
Gemara: We understand why he is silenced if he says “we give thanks, we give thanks,” because he seems to be acknowledging two powers…
(2) While the Jerusalem Talmud (Berakhot 5:3) likewise deems the repetition of the word modim as problematic, it is more concerned with the false implication of the existence of two deities rather than the unintended heretical insinuation:
“One who recites modim modim is silenced”: R. Shmuel son of Yitzhak said [the concern is based on the verse] “…the mouth of liars be shut.” (Psalm 63:12). Thus it is [only problematic when recited] in public, whereas in private they are merely supplications.
Note that the Jerusalem Talmud’s concern is further-reaching than its Babylonian counterpart. The latter would only mandate silencing the prayer leader when a repetition meets the criteria of heresy, the ultimate falsehood. The former, on the other hand, sees heresy as one example of a broader category of falsehood. Thus, something that is false, but not heretical, would still require immediate silencing.
However, not all infractions made by the prayer leader necessitate such an extreme measure. The Talmud in Berakhot (33b-34a) continues:
“We give thanks, we give thanks, he is silenced.” R. Zera said: To say “Hear, hear,” [in the Shema] is like saying “We give thanks, we give thanks.” An objection was raised: He who recites the Shema and repeats it is reprehensible [meguneh]. He is reprehensible [meguneh], but we do not silence him? — There is no contradiction; in one case he repeats each word as he says it, in the other each sentence.
The Talmud distinguishes between the more confrontational act of silencing versus a more passive condemnation of some forms of repetition as reprehensible or meguneh. However, it remains unclear as to why repetition absent heretical or false content should be problematic. While perhaps most instances do not qualify for automatic silencing, there can still be additional issues.
R. Moshe Schick, in his responsa (Maharam Schick, Orah Hayyim no. 31) propounds two additional halakhic issues with liturgical repetition:
(3) Bal Tosif: The Rabbis prescribed specific formulations for prayers and blessings. When there is a specific formula, it is forbidden “to change that which was coined by the Sages.” Hence, when one appends words to a formalized prayer, one violates the principle of Bal Tosif.
(4) Hefsek: Based on various passages in Shulhan Arukh (Orah Hayyim 104:5-6, 590:8), Maharam Schick asserts that adding words does not enhance the prayer experience, but rather constitutes a hefsek (interruption), which may compromise one’s fulfillment of prayer obligations.
R. Moshe Feinstein (Igrot Moshe, Orah Hayyim 2:22), however, claims that there is a high threshold for an addition or change to the text of prayers to constitute a hefsek. According to R. Feinstein, the only time the prayer would be invalidated (thus requiring the individual to restart the prayer anew) is either when (A) the new meaning conveys a false notion; or (B) if he deliberately alters the text in a way that renders a part of the prayer incoherent. Thus, any other forms of alteration would not constitute an authentic hefsek.
Yet even if we may obviate the technical bal tosifand hefsek considerations, there remain three broader philosophical problems with liturgical repetition. Maharam Shick proceeds to identify two ways in which such practices undermine the sanctity of prayer:
(5) How One Addresses the King: If one does not speak to a king of flesh and blood by repeating words in an odd and incoherent manner, certainly it behooves one to grant the King of Kings at least the same treatment. Maharam Schick buttresses this point with a passage in the Talmud (Sukkah 50a and Bava Basra 97b) that castigates bringing inferior quality material for an offering to God by citing the verse: “And when you offer the blind for sacrifice, it is no evil; and when you offer the lame and sick, it is no evil. Present it now unto your governor; will he be pleased with you or will he accept your person, says the Lord of hosts”(Malachi 1:8).
(6) Misappropriation of God’s Words for Personal Entertainment: The words of our Bible and liturgy are sacrosanct and should not be misused for one’s entertainment purposes. The Talmud (Sanhedrin 101a) relates how the Torah laments before God: “Master of the Universe, Your children have rendered me like a harp on which clowns play.” Maharam Schick asserts that when one adds words to the liturgy they are essentially transforming the prayer service into a concert meant for mundane gratification.
Similarly, R. Yosef Engel (Responsa Ben Porat no. 7) raises the alarm on cantors who add lyrics to “enhance” their performances and flaunt their voices. For one is supposed to concentrate on God when praying, but instead this cantor distracts the masses by refocusing the congregation’s attention from God to the cantor himself!
(7) Along a similar line of reasoning, we may derive a third potential philosophical issue from a responsum of Rashba (1:215) chastising cantors who cause a tirha de-tzibbura (imposition on the congregation’s time):
And it was taught in a baraita that we do not stand in prayer from a place of conversation nor from a place of frivolity nor from a place of lightheadedness nor from a place of inane words. Therefore, if the prayer leader’s intention is to make his voice heard and take joy in the fact that the congregants praise his voice, that is abhorrent. And it is a scenario like this that the verse refers to when it states in Isaiah (12:8), “she lifted her voice upon me, therefore I despised her.” In any case, it is inappropriate to lengthen the prayer services since in many places we are instructed to curtail out of concern for tirha de-tzibbura. For example, R. Yehudah recorded that the custom of R. Akiva was to shorten his prayer time when he prayed with the congregation.
Whereas R. Engel emphasized the concern of cantors distracting the congregants from thinking about God, Rashba criticizes cantors who unnecessarily impose on the congregants’ time. Following this line of thought, R. Naftali Hoffner (Dinei Tefilat Ha-Shahar 305:5, p. 394) suggests that the repetition of words would run afoul of this issue as well. Hence, even if there is nothing inherently wrong with altering the text, the gratuitous repetition of words will perforce constitute a tirha de-tzibbura.
Having enumerated seven potential issues with liturgical repetition, how do we make sense of congregations that seem to have no qualms with such practices?
Following his exhilarating rebuke, R. Epstein (Arukh Ha-Shulhan, ibid.), suggests a limmud zekhut, a strained justification for what he views as a nearly unequivocal sin:
In truth, perhaps there is no prohibition in this, but one who is good before God will flee therefrom. We have come to justify the actions of the sanctified descendants of Israel, whose eyes are sealed. Perhaps, from the fact that our Sages said that one silences a cantor only for repeating the word “Shema,” we may say that this is not true for other words that they repeat twice and three times. As to the notes they spread before the platform, we cannot present a reason to state a clear prohibition here, and so, “Let Israel practice as it will; better for them to practice in error, etc.”
Indeed, even when R. Epstein seeks to provide justification, he still concludes that the proponents of this practice are wayward and that it is better that they sin unwittingly than knowingly.
However, what if we could demonstrate that in certain instances it is not only permissible but even laudable to repeat words and verses?
Tosafot (Berakhot 33b-34a) state that the recital of “Hashem Hu Ha-Elokim” (Hashem is God) seven times at the conclusion of Yom Kippur and Hoshanah Rabbah may be regarded as a minhag kasher, a legitimate custom. By praising God seven times, we acknowledge each of the seven levels of the firmament upon which God resides. Thus, the repetition is not an act of heresy but a declaration of faith.
Unlike Tosafot, who were hesitant to extend their approval beyond the repetition of Hashem Hu Ha-Elokim, Bah (Orah Hayyim 61:7) is willing to even support repeating the verse of Shema itself. He writes that the repetition of Shema was incorporated into the selihah about persecution, for during these times of oppression and pogroms “they would sanctify God’s name by declaring aloud “Shema Yisrael Hashem Elokeinu Hashem Ehad! (Hear O Israel, the Lord is your God, the Lord is One!)” Thus, Bah contends that when the Talmudic Sages banned the repetition of Shema, they were not referring to such scenarios. For in these circumstances, the repetition of a verse does not detract from the meaning, but rather it emphasizes and reinforces the readers’ emotional commitment and dedication to God’s name.
Accordingly, we can understand why the Mishnah in Tractate Sukkah (38a) authorizes such a practice in regard to Hallel: “Where the custom is to repeat [the verses], he should repeat; [where the custom is] to say them only once, he should say them once.” Repetition that reinforces praise of and faith in God may even be deemed laudable; however, the school of thought exemplified by Maharam Schick, Arukh Ha-Shulhan, and R. Yosef Engel, who strongly opposed introducing repetitions, would likely aver that we may not introduce repetitions beyond what the great Sages of the Mishnah and Talmud deemed appropriate. Indeed, R. Yaakov Ariel (Responsa Be-Ohalah shel Torah 2:2) infers from Tiferet Yisrael (Berakhot 5:3) that one may only employ repetitions in Hallel because it is deliberately designed as a shir (song). Thus, the repetitions permitted in Hallel prove to be the exception rather than the rule.
R. Yaakov Epstein (who happens to have been a student of R. Ariel) provides a thorough treatment of this topic and concludes his analysis with a critical distinction. He writes (Responsa Hevel Nahalato 12:3):
In my humble opinion, the repetition of stanzas, sentences, and words are permissible for the sake of accommodating the tune when they are not a component of Keri’at Shema, Shemoneh Esrei, Hazarat Ha-Shatz or if it contains the Name of God – as there would be a concern for acknowledging two deities. And even the piyutim (later liturgical additions) found in Shema and Shemoneh Esrei may not be repeated. However, it is permissible to incorporate repetition into all other parts of the liturgy such as general petitions (bakashot ve-tahanunim). However, it would still remain forbidden to eviscerate or discombobulate the meaning of the prayers.
Thus, it seems that prayers instituted prior to the closing of the Talmud must remain immutable, whereas voluntary or post-Talmudic petitions would allow for a degree of malleability. Piyutim and other post-Talmudic or non-obligatory prayers, such as a mi she-beirakh and prayers for the state are subject to change. The original authors of these prayers do not possess any formal halakhic superiority to those who seek to alter them, whereas Halakhah recognizes that the laws canonized in the Talmud enjoy legal superiority and cannot be altered by ensuing generations.
We enumerated seven issues engendered by liturgical repetition. And while that may be daunting, there are potential solutions to each of them.
Regarding the first two legal concerns for uttering (1) heresy and (2) falsehood, we may suggest that many repetitions do not convey such messages. Regarding the more technical concerns such as (3) bal tosif and (4) hefsek, if one limits his freedom of repetition to post-Talmudic or non-obligatory prayers, it would not be subject to these principles. And even if one inadvertently repeats a word or phrase of a compulsory prayer, R. Feinstein reassured us that there is a high threshold for constituting a hefsek.
This brings us to the broader philosophical concerns, or rather, values of (5) addressing the King in a dignified manner; (6) ensuring that the prayer is about praising God, not personal gratification; and (7) not imposing on the congregation’s time. While these may be construed as halakhic issues, they do not possess clearly defined parameters in most instances. Thus, they may be superseded by a competing value such as reinforcing one’s relationship with God. Decisions like this carry a degree of subjectivity and, in my biased opinion, are generally best left to the intuition of the congregation’s rabbi. Regardless of who makes the decision, one must always remember that the key is for the tunes and words to facilitate a connection to God, rather than God’s word serving as a means for an ulterior motivation.
 Translations of Arukh Ha-Shulhan are from Sefaria.org, Talmudic quotes are from the Soncino edition; all other translations are mine.
 The Mishnah in Sukkah (38b) indicates that repetition is acceptable in the context of Hallel. We will address this later in the essay.
 See Tosafot (Berakhot 33b) who cite this passage. The rationale to permit repetition during one’s private prayer is that God can discern the individual’s true intentions.
 It is ambiguous from the Gemara which category requires immediate silencing and which one is merely meguneh. Rashi (s.v. Milta) claims we only silence someone who repeats an entire verse, as that appears as if one is praising two deities, whereas literally repeating “Modim Modim‘” or “Shema Shema” is simply meguneh. However, Tosafot (s.v. Amar) cite the opinion of Bahag and Rabbeinu Hananel who assert the opposite. See Mishnah Berurah (61:23), which rules that one who repeats the entire verse has fulfilled the mitzvah ex post facto, whereas literally repeating the word “Modim” or “Shema” twice might compromise one’s fulfillment of the mitzvah altogether. This is primarily an issue of hefsek which we will address below in this essay.
 The order of Maharam Schick’s concerns has been adjusted to better fit the flow of the essay. Note that he also cites the concern of declaring falsehood based on the aforementioned passage in the Jerusalem Talmud.
 See Rosh Hashanah 28b.
 See Responsa Pekudat Elazar (Orah Hayyim no. 25), which claims that altering a prayer like the Shemoneh Esrei is a far more egregious form of hefsek than doing so in a prayer such as Yekum Purkan.
 However, R. Feinstein writes later in the same responsum that even if one is not required to restart the prayer due to hefsek, gratuitously altering the text by making additions would still be meguneh. R. Shlomo Zalman Auerbach adopts a similar position, as recorded in Halikhot Shlomo (9:9).
 We may suggest that bal tosif is not a concern when we are dealing with blessings and prayers that are not explicitly formulated word-for-word. See R. Yehuda Herzl Henkin, who addresses to what extent a blessing may be altered in his treatment of shelo asani ishah (Responsa Benei Banim 4:1).
 See Berakhot 34b.
 For further analysis, see R. Ovadia Yosef, Responsa Yabia Omer, Orah Hayyim 7:4-5.
 Note that tirha de-tzibbura can also be presented as a halakhic concern. I have discussed its parameters and legal force here and here.
 It is worthwhile listening to how R. Yehoshua Grunstein passionately makes the case for why tirha de-tzibbura should be emphasized in this discussion.
 Tosafot also point out that the original verse of “Hashem Hu Ha-Elokim” found in the Bible (I Kings 18:39-40) is repeated twice. One could view the verse as precedent for permitting repetition or, alternatively, it might be the exception that proves the rule. Authorities like R. Schick and R. Epstein would likely follow the latter reading. However, we have adopted the former interpretation for this section of the essay.