“Each River and its Channel”: Halakhic Attitudes Toward Liturgy
A reply to Rabbi Aryeh Frimer, “The Wrong Changes in Jewish Liturgy”
Guest post by R. Dr. Seth (Avi) Kadish
Rabbi Dr. Seth (Avi) Kadish is a graduate and musmakh of Yeshiva University who received a Ph.D. in Jewish Philosophy from the University of Haifa. He teaches adult Jewish education in a number of frameworks including soldiers in the IDF’s “Nativ” course, the Hebrew University’s Gandel Program, and the International School at the University of Haifa. He is the author of Kavvana: Directing the Heart in Jewish Prayer (1997) and The Book of Abraham: Rabbi Shimon ben Ẓemaḥ Duran and the School of Rabbenu Nissim Gerondi (currently being published online). This essay is available under the CC-BY-SA license; attribution should point to the post at the Hirhurim blog. A continually updated and corrected version is available here: link.
A. Reading the Rambam Critically
A few months ago I arranged source-sheets for an evening of study with Israeli adults on a topic that would seem to be entirely unrelated to Jewish liturgy, namely the Rambam’s views on the afterlife and the resurrection of the dead. This was part of a year-long course of study meant to cover the “big issues” in Jewish philosophy, primarily by taking the Rambam and contrasting him to the alternative outlooks of Rabbi Yehudah Halevi in the Kuzari and others. In this case it meant studying the Rambam along with the sharp criticism of his eschatology by the Ramban in Sha`ar ha-Gemul.
The Rambam’s openly stated view was that the term Olam ha-Ba (“The World to Come”) as used by Ḥazal refers to the eternal life of the soul (or the intellect) after death without the body, while Teḥiyyat ha-Metim (the resurrection of the dead) refers to a one-time miracle in which the human being as a whole will live again in a body. This is a very straightforward and appealing view, and is probably the way most Jews understand those terms today, both due to its innate appeal and to the Rambam’s powerful influence. As always, the Rambam’s view on this issue is clearly and forcefully stated (though perhaps more stridently and apologetically than usual in Ma’amar Teḥiyyat ha-Metim), and can be read together with his general philosophical views in fascinating ways on both exoteric and esoteric levels.
What is important for our purposes here is not just that the Ramban later attacked this view, but how he carried out his attack, namely that his most powerful and convincing arguments were specifically exegetical ones: By mustering numerous proof texts, the Ramban made a powerful case that when Ḥazal used the term Olam ha-Ba they could not possibly have meant what the Rambam says they meant, but that they rather referred to the future life of the resurrected dead within bodies. In other words, the “World to Come” and the resurrection of the dead are one and the same. After studying the Ramban, one is left with a strong impression that the Rambam’s reading of Ḥazal is a difficult and problematic one that is extremely hard to justify as peshat.
B. The Rambam’s Theory on the Origin of Prayer
A similar thing is true of the Rambam’s view on the origin of halakhic prayer. In a very straightforward and appealing way, he famously explained (Hilkhot Tefillah 1:1-6) that the rabbinic decree of formal prayer was a social and pedagogical response to difficult historical realities during the return to Zion, specifically to the inability of Jews to pray to God in Hebrew in their own words. This made it necessary for Ezra and his court to legislate a basic prayer-text in Hebrew that all Jews could use. Much like the Rambam’s view on Olam ha-Ba, this is probably the way most observant Jews today understand the origin of the siddur, both due to the Rambam’s influence and to the way his explanation of history makes highly intuitive sense.
But as in the case of the Rambam’s interpretation of Olam ha-Ba, even though his view on the origin of rabbinic prayer is internally cogent, it is still not necessarily the natural way to read and understand numerous talmudic passages that may well imply the very opposite of what he said. In this case it was the Ramban’s greatest student, the Rashba—whose work represents the culmination of the Sephardic tradition of encyclopedic Ḥiddushim on the Talmud—who vividly illustrated just how difficult the Rambam’s view on the creation of rabbinic prayer is in light of the talmudic evidence (Ḥiddushei ha-Rashba on Berakhot 11a):
It seems to me that when the mishnah taught “When they said to be short one may not lengthen, and when they said to lengthen one may be short” (Berakhot 1:4), it doesn’t mean that one may not shorten or lengthen the text of the blessing, i.e., not to add or subtract words. Because if that were so, then they should have decreed the text of each and every blessing with specific numbers of words, and taught us the exact text of each blessing. But we never find this anywhere! They only pointed out words which are specifically required, such as in the arguments about mentioning rain (Berakhot 33a), dew and wind (Ta`anit 3a)…
But when it comes to the remaining text of the blessings, the rabbis never gave a measure saying that a person must say an exact number of words, no more and no less! And they never told us how many words must be said in a blessing in order for it to be called “long” or “short.” And not only this, but they explicitly said that in the eighteen blessings of prayer, if one wants to add something to any blessing in accordance with its topic he may do so (Avodah Zarah 8a). And this is true even if the words of his addition are more than the main part of the blessing, and although those blessings are “short” ones [but he makes them long]…
And even though… the blessing ha-zan is called “long” in the Tosefta and the Yerushalmi, it is nevertheless concluded later (Berakhot 40b) in case of Benjamin the shepherd that even if he said “Berikh Raḥamana Mareih de-hai pitta” [Aramaic: “Blessed be the Merciful One, the Master of this bread”] he has fulfilled his obligation! And the gemara there implies that this is even lekhateḥillah [=fully permissible] as it cites in this context: “These are said in any language…”—which is even lekhateḥillah, and the same thing is also implied about this in the Yerushalmi (6:2).
This quotation is just part of a comprehensive essay by the Rashba showing that Ḥazal never composed a fixed prayer-text when they enacted formal blessings, but rather ordained a general format, i.e. a certain number of blessings on specific themes, with set procedures for how to begin and end them. To the Rashba’s halakhic analysis of the structure of blessings based on the talmudic evidence, it is also possible to add numerous talmudic anecdotes and other statements that seem to imply the opposite of what the Rambam theorized.
The ironic thing about this whole issue is that on a truly halakhic level, i.e. when it comes to explicit talmudic rules about the structure of blessings, the Rambam generally agreed with what the Rashba would later write. And structure is the only thing about which there are any explicit rules in the Talmud! It is specifically on a non-halakhic level that the Rambam and the Rashba disagree, namely in their historical theories about what Ḥazal actually did when they enacted blessings: Did Ḥazal compose texts (as demanded by the Rambam’s theory of history), or did they rather mandate blessings on specific themes, leaving the exact words up to the person saying them each time? In fact, the Rambam himself seems to have hinted that there is no explicit talmudic basis for a prohibition against changing the words in blessings. He used an unexpected phrase to discourage it when he wrote that to do so is ein ra’uy (“not fitting”), rather than simply saying that it is a wrong or prohibited procedure (Hilkhot Berakhot 1:5).
C. A Spectrum of Opinion
Among the Sephardic rishonim there were also compromise positions. The Ra’ah leaned towards the idea that Ḥazal actually composed a text, but took that historical theory in a different practical direction by suggesting that Ḥazal had no problem with their text being changed, including major changes, as long as those changes didn’t become permanent ones in every prayer. Adding piyyutim is permitted precisely because they are not added every day. The Ritva cited the Ra’ah alongside the Rashba.
Far from being the view of “a few rishonim” who argued with the Rambam as Rabbi Frimer would have it, the Rashba’s flexible view of both history and halakhah was a popular Sephardic one, echoed by the Abudarham and adopted with some modifications by the Rashbaẓ (a Spanish exile who became the greatest posek of North Africa and an important authority for the author of the Shulḥan Arukh). Furthermore, if we take the compromise positions into account, then the Rashba’s basic overall conclusion about the legitimacy of making major changes within the blessings was shared not only by the much earlier ge’onim, but also by the vast majority of rishonim in Spain, North Africa, Provence and Ashkenaz, including a great many that it is impossible to analyze here. In Spain: Abudarham, Rabbenu Yonah (on Alfasi Berakhot 5b), Rashba, Ritva, Ra’ah, Ran, and Rashbaẓ. In North Africa: Rabbenu Hannanel and Rashbaẓ (as previously). In Provence: Ra’avad (quoted in the Tur and Shiltei Gibborim A on Alfasi Berakhot 5a) and Meiri.
In Ashkenaz the discussion began with Rashi’s interpretation of Mishnah Berakhot 1:4 (11a), which was ambigious. It seemed to pose a theoretical problem for major additions like piyyutim (even though I am unaware of any explicit evidence that Rashi actually opposed them). According to the Meiri’s interpretation of Rashi, it is forbidden to add new topics to blessings, which would exclude major piyyutim but permit modest changes. Alternatively, the Ra’ah followed Rashi in his interpretation of the mishnah, explaining “long” and “short” in terms of the length of the text, but permitted any change that was not permanent (see above). Most influential, however, was Rabbenu’s Tam’s far more lenient interpretation (rejecting Rashi), which was cited widely and approvingly among the ba`alei ha-tosafot (Berakhot 11a), Hagahot Maimoniyyot (on Hilkhot Tefillah 6:3), and many other Ḥakhmei Ashkenaz (and even elsewhere such as by Rabbenu Yonah and the Meiri).
An important exception was the Rosh, who is quoted by his son in the Tur (Oraḥ Ḥayyim 68) as having recommended the elimination of piyyutim in the blessings (“for it causes [people] to interrupt through useless talk about frivolities, and also Rabbenu Tam’s explanation that he gave to justify the custom is not correct in my eyes”). But elsewhere (on Berakhot 34a and cited in Tur Oraḥ Ḥayyim 112) the Rosh explicitly accepted the practice based on standard justifications without a word of criticism.
Indeed, Tur Oraḥ Ḥayyim 68 is highly unusual in its emphasis. As such, it is probably the best possible example of the exact halakhic attitude that Rabbi Frimer emphasized in his article. It begins with a lengthy citation of the Ramah, who described his displeasure with piyyutim and his unfortunate inability to eliminate them, followed by the Rambam. The Tur then continued by pointing out the prevalence of the custom, Rabbenu Tam’s “forced interpretation” (daḥak lefaresh) in order to uphold it, and concluded with his father’s recommendation to abolish it.
However, Tur Oraḥ Ḥayyim 68 is not in any way representative of the general opinion in Ashkenaz or elsewhere, neither in its own time nor later. The Baḥ on the Tur responded passionately in favor of the traditional Ashkenazic opinion. The Levush is notable for his succinct summary of the halakhic situation in his time, both in terms of the principles and the practical reality; he remarked, “Nahara nahara u-fashteih [Each river and its channel]: Where the custom is [to say piyyutim] then that is the custom, and where the custom is not to say them then that is the custom, for each of the two opinions has authorities to rely upon.”
D. Practical Halakhah and Halakhic Attitudes
Rabbi Frimer is quite right to say that it is wrong to tamper with the text of blessings according to the Rambam and that the Shulḥan Arukh accepted this position (Oraḥ Ḥayyim 68), apparently because it was shared by the Rambam and the Rosh. However, it is equally correct that to make changes in blessings is fully permissible according to the Rashba and many others (actually the vast majority of rishonim), and that the Rashba’s position was adopted by the Rema. What I find lacking in the Ḥakirah article is that it heavily documents and emphasizes just one sliver out of a very wide spectrum while largely ignoring the rest. It also errs by turning the Rambam’s attitude into a prism through which alternative opinions are viewed, in the sense that they must be assumed to have allowed piyyutim only for exceptional reasons, or only allowed piyyutim and nothing else.
Rabbi Frimer argues with passion that the position of the Rashba and the Rema is utterly irrelevant to contemporary changes because it was an exception made by the rishonim for piyyutim alone, and only because their recitation was already such an entrenched and highly revered custom. If correct, that would seem to settle the issue once and for all.
However, it is a mistake in my opinion to say that when it comes to piyyut, the rishonim were doing nothing more than a “limmud zekhut” in order to defend an age-old custom. It was indeed a venerable custom for them, but that was only because it was first and foremost a popular and beloved custom, both among great Torah scholars (Rambam and Ramah notwithstanding) and the people at large in synagogues. Piyyut was, more than anything else, a beloved social institution in the middle ages. This bit of historical reality must be kept in mind if one wants to read the rishonim in a balanced way and draw fair conclusions from what they wrote. The point is that they weren’t forced to justify piyyut in the middle of blessings just because it was a such an entrenched custom. Rather, they wanted to because it was a way to better serve God, and they were able to because the truth of Torah supported it.
In other words, this was not by and large a “limmud zekhut” at all. Not in Spain from the 13th century onwards, and certainly not in Ashkenaz where the ba’alei ha-Tosafot reacted to the theoretical anti-piyyut implications of Rashi’s position. Rather than a “limmud zekhut” to justify the insertion of piyyutim into berakhot, what we have here is a principled, across-the-board recognition that peshat in the Gemara simply does not preclude significant changes and additions in the obligatory blessings. This is a matter of basic talmudic truth that goes beyond piyyutim, and most of the rishonim who discussed it in their comments on the sugya emphasized the principle first and the piyyutim second. Had the majority thought it was truly forbidden they wouldn’t have allowed it or would have called it bedi`avad (just like the minority who actually opposed it). The large majority of rishonim supported saying piyyutim within the blessings because in their opinion such change is allowed lekhateḥillah as ikkar ha-din.
And now we get to the crux of the matter: Does the fact that the social institution of piyyut has been the historical test-case for this sugya for the past thousand years mean that it is the only possible application of the talmudic principle? Are we forbidden to apply it in other areas, even if matters related to avodat Hashem are directly affected by the very same sugya? I think not, agreeing here with Rabbi Sperber as opposed to Rabbi Frimer’s article.
It is very important to ask why piyyut was the only major real-life historical example of change in the formal blessings for so many centuries. Indeed, in a comment to me Rabbi Frimer made the interesting and related argument that “if things were as free as you argue, the differences would have have been much greater than they are.”
In theory I agree with Rabbi Frimer on this point. Logic might indeed suggest that people would generally take advantage of available halakhic leniencies, and that attempts would not be made to force conformity if no conformity was truly required. However, in the case of prayer, halakhic ideas and ideals have rarely had much to do with the reality of how Jews prayed throughout history (much of my book is devoted to illustrating this).
Furthermore, in this case it seems that the forces opposing change in the formal prayers were far too strong. The first force was the increasing dispersion of Israel a millennium ago, along with other historical factors that motivated the ge’onim to write the first halakhic manuals for prayer, i.e. the very first siddurim. The propagation of siddurim around the Jewish world during the times of the rishonim was a powerful historical force, one that in its very nature opposed flexibility in the obligatory blessings decreed by Ḥazal, and the siddurim fundamentally changed attitudes towards Jewish prayer by their very existence. The second force was more powerful than the first, but even less connected to the world of halakhah: The invention of printing and the large-scale production of siddurim for the masses.
However, neither of these two historical forces—the existence and spread of siddurim during medieval times and the mass production of siddurim in the age of the printing press—is halakhic. It may even be argued that they had anti-halakhic consequences (I do so argue in my book). But they have nevertheless characterized Jewish prayer for the past thousand years, and as a result the very idea of formulating an obligatory berakhah in one’s own words—something that was arguably unremarkable in the Talmud—has become unimaginable to Orthodox Jews today, even if halakhic principles would suggest otherwise.
And so we are confronted by something that is fundamentally a matter of attitude rather than halakhic principles. The reality of the past thousand years is clear, but it is not halakhic. Rabbi Frimer nevertheless argues that this reality should be treated as halakhic obligation, even at the expense of avodat Hashem. I disagree with that attitude for the numerous reasons spelled out in my book, some of which are echoed by Rabbi Sperber.
The fundamental problem is whether one should minimize or maximize the significance of the institution of piyyut from a Torah perspective. Rabbi Frimer sees their historical popularity as something that once created an anomaly or even a distortion within the halakhic framework, but it seems that the distortion has gradually been corrected over time as ever more communities have moved towards the Rambam’s approach. His approach may even imply that it is perhaps God’s will for us to fully correct that distortion once and for all. But it is also possible to suggest something else entirely: Maybe the historical significance of piyyut was actually the hand of God in history, working to make sure that halakhists would be forced to deal with the element of creativity within the formal blessings as a practical issue, so that the flexible attitude would not be lost entirely within the mainstream of halakhah, but would rather be kept alive through the positions of so many of the greatest rishonim and as codified by the Rema. Or perhaps, when God planted such a strong urge for creativity and personal expression within human beings, that in itself already meant that it would be impossible for Jews to leave the formal blessings entirely alone over the course of centuries in exile (and that this is a positive thing since they were never meant to be frozen in the first place).
Since we do not know God’s mind, these suggestions on both sides are really just matters of attitude. But both attitudes are equally halakhic and expressions of the Torah.
There is one more remark by Rabbi Frimer makes that deserves further thought: When it comes to changes in Naḥem for the ninth of Av, he mentions that the drastic historical changes that have taken place in our day might be viewed as she`at ha-deḥak (thus allowing a change in that particular blessing as a unique exception). What Rabbi Frimer means by this is obvious and correct, but the phraseology is striking: she`at ha-deḥak?! Are we to call the ingathering of the exiles and a thriving Jewish Jerusalem in a strong, independent Jewish state, and the partial realization of a great redemptive promise that God made to His prophets as it takes place before our eyes, a “situation of hardship”?
The very use of this phraseology is a clear indication of halakhic attitude, because here too the opposite attitude and phraseology are equally possible: Perhaps it not changes within Naḥem that are permitted today because of a current she`at ha-deḥak, but rather 2000 years of exile that were she`at ha-deḥak in the effect they had on halakhic prayer: When Jewish communities are dispersed around the world, living among other nations as minorities, it is natural for each long-term community in exile to settle upon a single fixed Hebrew text for prayer, and to hold fast to that exact nosaḥ over the course of many generations and centuries. In such an atmosphere, even an appropriate change to Naḥem reflecting joyous new realities seems to be deeply problematic, and that very attitude reflects the she`at ha-deḥak of exile. But when Israel dwells together as a people in its land, that is not the natural way for the nation to talk to its God. Here is how Joseph Heinemann described the way Jews said the Amidah during Second Temple and Mishnaic times (which is a way to think about how the Rashba’s idea might have worked in real life):
In early times the person who prayed was not obligated to use a certain text of a blessing fixed word for word, but on the spot he would “compose” such a text for himself as he said it, or—when he wasn’t able to do so—he used one of the popular texts that he learned from others. But even that praying Jew wouldn’t repeat this same text time after time, but would change it, lengthen it or shorten it, consciously or without realizing it, according to his needs and the amount of free time that he had. Of course, as time went on, certain phrases became widely accepted by most praying Jews, as well as mentioning various details and matters (beyond the main topic of the blessing). At this stage the halakhic scholars made mentioning these phrases and other matters an obligation. Each Jew who prayed was still able to use, for the main body of the blessing, the words that he preferred, or which he was used to, as long as he “mentioned in it” those details and phrases, which had meanwhile become the accepted custom… For instance: “One who reads Shema` in the morning must mention the Exodus from Egypt in Emet ve Yatziv…” (Yerushalmi Berakhot 1:3).
It must be stressed that the ingathering of the exiles in Israel today is shattering the legacy of frozen customs and prayer-texts (at least outside of the ḥareidi world and immigrant communities). It is not resulting in some sort of “unified custom” or single nosaḥ as people once imagined, but is rather moving towards a general tolerance for all sorts of competing customs and texts simultaneously within mixed communities and families. Within Religious Zionism, the Ashkenazic paternalism of the past is vanishing rapidly today, and that is a blessed event. Especially in the newer yeshivot and communities, customs are allowed to vary based on the preference of the shali’aḥ ẓibbur, and sometimes various minhagim are even combined eclectically. What is happening in Israel today is unimaginable from the perspective of the largely Ashkenazic monotony of Orthodox life in the English-speaking world.
E. Obligatory versus Non-Obligatory Prayers
In his review, Rabbi Frimer took Rabbi Sperber to task for not distinguishing between changes within obligatory blessings versus changes in other non-obligatory prayers. In my opinion, this major part of his essay was superfluous and entirely missed the point: The author of eight brilliant volumes of Minhagei Yisrael clearly understands that basic distinction no less well than Rabbi Frimer or myself. The obvious reason Rabbi Sperber spent so much time and energy on the non-obligatory material is simply because people’s attitudes are often far more significant than halakhah itself. American Jews used to joke about the Orthodox concession to modernity being to give up on the 2nd Yekum Purkan, and laugh about it without even realizing that Sephardim don’t say either of them! What do such Jews think when they learn that Sephardim don’t say Yizkor either, nor do they have Unetaneh Tokef on the Yamim Nora’im? What is Rosh Hashanah without Unetaneh Tokef? In fact, most Sephardim don’t even have the very concept of Yamim Nora’im, and these holidays have a significantly different atmosphere than for Ashkenazim. So can we still call Sephardim “Orthodox”? Are they really even Jews?
There is nothing halakhic about any of these tongue-in-cheek comments, but it is obvious that such attitudes color the thinking not just of the average Orthodox Jew, but even that of serious Torah scholars. If we are so exacting about non-obligatory things, treating them as if they were severe obligations down to the very letter and vowel, then how much more so must we be strict about true talmudic obligations! Isn’t that an “obvious” kal va-ḥomer? In my opinion, this attitude deeply colored Rabbi Frimer’s reading of the rishonim in his article, which is what made the Rambam’s position so unremarkable and obvious to him, and also what led him to portray piyyutim as a one-time historical exception to a very clear general rule, which was only in deference to a problematic custom. It was therefore quite right for Rabbi Sperber to dwell on the non-obligatory prayers, and it is absurd to suggest that he confused the two categories.
F. Is Ḥiddush Only for Petitions?
In his review, Rabbi Frimer further limits ḥiddush within the rabbinic blessings to the petitionary blessings in the middle of the Amidah. I analyzed Rabbi Eliezer’s statements about keva` and Gemara’s idea of ḥiddush in my book and will not deal with it here. To make the point short, it is incorrect to simply equate these statements with adding personal petitions in the middle blessings of the Amidah, to the exclusion of making changes in other blessings. If we don’t adopt the Rambam’s position, then the talmudic concept of ḥiddush is more fundamental than the narrow area of individual bakkashah.
In Berakhot 34a we are taught in the name of Rav Yehudah: “A person must never ask for his needs in the first three blessings or the last three blessings but only in the middle blessings. For Rabbi Ḥanina said: The first ones are like a servant who sets out praise before his master, the middle ones are like a servant who asks a largess from his master, and the last ones are like a servant who received a largess from his master and takes his leave.” This is a vivid illustration of the broader talmudic idea that blessings are limited to their topics and themes, such that a petition is normally out of place in a blessing of praise. However, changes and additions that conform to the theme are fine.
That is exactly how changes within blessings of praise were explained by the rishonim. First of all, they pointed out that the Gemara only prohibits “a person” asking for “his needs” to do so in a blessing of praise, i.e. it is individual and personal needs that are inappropriate (especially since the last three blessings are full of requests for public needs). Secondly, they explained that public needs are perfectly appropriate even in blessings of praise: “Together with the servant setting out praise before his master, he can also ask for public needs, for it is considered praise and honor to his master that the public depends upon him.”
To clarify: Changes and piyyutim are wrong in any blessing according to the Rambam, no matter what they ask for or who is doing the asking. What allows them to exist at all is the alternative idea shared by the Rashba and others that only the format of prayer is fixed and not its text. However, even according to the flexible position, the change or addition must conform to the theme and topic of the blessing (me`ein ha-berakhah). Therefore, petitions for personal needs are only appropriate in the middle blessings.
Rabbi Frimer is probably correct that the ever-present phenomenon of adding requests contributed to some of the many variations we find in the middle blessings. But the first three blessings exhibit more variation than Rabbi Frimer allows for, and the last three even more so (not to mention the berakhot of Shema`). The relative uniformity we find in the first three blessings and especially the first one (and it is only relative) likely derives in part from another talmudic source that limits attributes of praise to God in the beginning of the Amidah (cf. Berakhot 33b). But that is not a general rule for all blessings.
G. Two Valid Attitudes
To finish, I’d like to quote the conclusion to chapter 9 in my book (the chapter entirely devoted to the Rambam’s position). This conclusion is about the two conflicting attitudes I have been trying to describe, and shows that Rabbi Frimer’s approach has great merit precisely because it is so well grounded in the Rambam. Though I wrote it 14 years ago, it seems like it was written today as a direct response to Rabbi Frimer:
In conclusion, we have seen that each view—both those who denied that there was an official text and those, like Rambam, who believed in one—is able to explain the talmudic sources according to its own assumptions. Both approaches, then, are valid and consistent with the talmudic evidence. Both are the words of the Living God. However, this does not mean that both can easily coexist in terms of religious practice. For each implies a Torah value that is to some degree exclusive of the other.
Remember that the Kesef Mishneh, explaining the Rambam, rhetorically asked what profit there is to changing any of the words in the “official text.” He meant, of course, that there is no profit! But the opposite camp would certainly argue that it is important to avoid keva` by “saying something new” in prayer each time, and that if not doing so results in rote prayer then one has committed a serious religious and halakhic trespass (as we saw in chapters 1-2). In the middle ages, the primary way to bring ḥiddush into formal prayer was by adding piyyutim, which gave formal prayer a fresh depth of meaning that was absent without them. Thus it is no accident that for the rishonim, the matter of changing prayer-texts was tied to the issue of piyyutim on a practical level. Now, the Rambam might counter that the obligation to “say something new” was not accepted as binding halakhah. But the response would still be that it is a great Torah value to “say something new” for those who are capable of doing so and want to, even if it is too great an obligation to impose upon all.
On the other hand, denial of an official text lessens the value of custom and tradition, especially for a carefully worded text that many, like the Rambam, ascribe to Ezra and the Anshei Kenesset ha-Gedolah. Minimally, the exact words in our siddurim have been carefully preserved and lovingly recited and studied in their present forms for many centuries by men who were good and pious Jews and great Torah scholars. That value cannot be denied either.
Ultimately, it must be recognized that both sides are based on true Torah values, but that there may be no way to reconcile them in a way that is satisfactory to both. The Rambam’s view, though it was a minority opinion in the middle ages as we saw in chapter 8, has dominated Jewish life ever since (partly for kabbalistic reasons that he himself would never have accepted). But part of the reason our daily tefillot have been drained of their meaning is that the opposite value (informality and personal expression) has been almost totally denied for so very long. The time has come for us to respect the values represented on both sides, and to regard both as legitimate choices for Jews who pray. The tension between two Torah values is an argument for the sake of heaven.
 On the opposing positions of the Rambam and the Rashba, and for an exhaustive, exacting and comprehensive analysis of the entire spectrum of opinion on this complicated talmudic topic that is carefully organized and clearly written, see Rosh Devarekha by Rabbi Yiẓḥak Shilat (Ma`aleh Adumim, 1996), pp. 217-256 (chapter 9). In my opinion, no one should approach this topic without first carefully studying Rav Shilat’s outstanding essay. My own book (see n. 3) also examines this topic in great detail (chapters 8-11), including certain larger dimensions of it that go beyond the formal sugya and which Rav Shilat does not address; but in terms of a classic, thorough talmudic analysis Rav Shilat’s essay is the mature work of a world-class Torah scholar and second to none.
 On the way Rambam’s writing about prayer fits into his general halakhic an philosophical outlook, see Gerald Blidstein, Ha-Tefillah be-Mishnato ha-Hilkhatit shel ha-Rambam (Jerusalem: Bialik Institute, 1994).
 In the parallel passage in the Yerushalmi, permission is given to Benjamin (or Minyamin) to continue saying the blessing this way. As Rav Shilat points out (pp. 227-228), the Rambam would probably explain this by saying that the shepherd was incapable of learning to do otherwise.
 I analyzed several such texts at length in my Kavvana: Directing the Heart in Jewish Prayer (Hoboken, NJ: Jason Aronson, 1997), pp. 262-266. As for how the Rashba’s basic theory makes excellent historical sense, not just in terms of explaining the sources but also in light of how early rabbinic prayer might actually have worked in reality, see pp. 257-266. It is clear that all of the arguments and prooftexts that the Rashba marshaled against the Rambam’s position can be explained in accordance with the latter, and Rabbi Yiẓḥak Shilat showed exactly how this can be done in Rosh Devarekha, pp. 227-232. However, Shilat too admits that the historical evidence seems to better support the Rashba’s position (p. 228 n. 17). The real question is not whether a logical solution can be proposed in accordance with the Rambam in every case, because it is clear that each argument and text is fundamentally ambiguous and allows for both readings, but whether that solution really seems to be peshat throughout the entire talmudic sugya at hand.
 Rabbi Yosef Karo examined the Rambam’s unusual wording in the Kesef Mishneh, but Rabbi Yosef Kafiḥ offered a much clearer explanation of that halakhah. On this see my Kavvana, pp. 312-313 (n. 8). The other major rishonim who explicitly opposed piyyutim like the Rambam (though not necessarily for the same reasons!) were the Ramah and the Rosh.
 A late echo of this position can be found in Arukh ha-Shulḥan, Oraḥ Ḥayyim 119:2, which discusses the radical error of people who take printed, pre-formulated petitions provided in their siddurim as options, but say them each and every every time out of ignorance!
 This is the phrase that Rabbi Frimer used in a minimalistic fashion in order to summarily dismiss opposition to the Rambam’s view in footnote 37 of his article. That single footnote is actually the basis for most of Rabbi Frimer’s major halakhic conclusions in his article, but it unfortunately summarized a complex, nuanced topic in an overly simplistic and unfair way.
 Cited in full by the Beit Yosef (on Tur Oraḥ Ḥayyim 113) as a rejection of the views of the Ḥasidei Ashkenaz who “counted the words” in blessings. In this essay I will not deal at all with those who took the “fixed-text” position for mystical or kabbalistic reasons, beginning with the Ḥasidei Ashkenaz and culminating with the hasidic siddur (“Sefard”) and the Nefesh ha-Ḥayyim (mentioned by Frimer in n. 37, and which I discussed at great length in Kavvana, pp. 281-293). The author of Nefesh ha-Ḥayyim was operating on the same kabbalistic playing field as the hasidim with whom he argued, and with whom he shared most of his fundamental outlook. But that entire debate mostly left behind the nuanced halakhic give and take based on the talmudic evidence that was characteristic of the rishonim.
 Most important is that the Rashbaẓ polished the Rashba’s halakhic analysis by providing for a more nuanced and flexible interpretation of “long” and “short” blessings based on each specific talmudic context. In addition, he seems to have thought that Ḥazal actually did compose a text, but that they placed no limitations at all on changing it beyond the structure of the “long” and “short” blessings and the basic theme of the particular berakhah.
 In his citation of R. Isaac ibn Ghiyyat (commentary on Alfasi Rosh Hashanah 8b). However, he also cited the Ramban (see next note).
 For a full analysis of Rashi, see Shilat pp. 217-221; also see my Kavvana, pp. 268-269. Since my book was published, I have become ever more convinced that Rashi is talking about adding topics that lengthen a blessing, not changing words (as per the Meiri). In truth, there is no evidence whatsoever to suggest the latter, but there is an important hint towards the former. It should be pointed out that the Ramban (Ḥiddushim on Berakhot 46a) expresses an idea similar to that of Rashi (see Shilat, p. 218) but likewise says nothing explicit about piyyutim. However, the Ramban is also cited by the Ran (see previous note) as saying that even short additions to the first three blessings in the Amidah on Rosh Hashanah and Yom Kippur are only grudgingly permitted; however, this seems not to be because they are changes per se, but because they are petitions within blessings of praise. Cf. very similar comments by none other than the Rashba as cited in Beit Yosef on Tur Oraḥ Ḥayyim 112 (both the Ramban and the Rashba cite Masekhet Soferim; Rav Shilat doesn’t seem to have mentioned this citation of the Ramban); also see below in this essay, “F. Is Ḥiddush Only for Petitions?”
 The Beit Yosef on Tur Oraḥ Ḥayyim 112, the siman which deals with piyyutim in the first three blessings of the Amidah and justifies the custom, wrote: “In siman 68 it was already explained that the correct custom is not to say the krovetz.”
 For a full analysis of this issue as it developed in the Tur and the Shulḥan Arukh, see Rosh Devarekha, pp. 250-256 and Kavvana, p. 314 n. 9. On whether this might have any practical relevance for those who follow the Beit Yosef rather than the Rema, especially taking into consideration the consensus of most rishonim, see Kavvana, p. 295-296 and n. 61.
 It is important to emphasize that in this essay I have greatly simplified the range of approaches to this sugya among the rishonim, and painted the picture with a very wide brush. For a much more detailed and nuanced presentation, see Rav Shilat’s study.
 Former yeshivah students from my generation may have a hard time believing that people once loved piyyutim, since we generally hated them knowing that they had little halakhic basis while at the same time they were so difficult to understand and seemed like such a burden. I advise anyone who feels this way to spend a good deal of time with Sephardim.
 Joseph Heinemann, Ha-Tefillah bi-Tekufat ha-Tannai’m veha-Amoraim: Tivah u-Defuseha (Jerusalem, 1984), pp. 36-37. While making a very similar point, Shadal added the following practical idea: “That is why they instituted the silent Shemoneh Esrei, so that the shali’aḥ ẓibbur could rehearse his prayer to himself before reciting it for the congregation” (from the introduction to Shadal’s edition of the Italian version of the siddur, “Mavo le-Maḥzor Benei Roma“, in Maḥzor Kol ha-Shana kefi Minhag Italiani [Levorno, 1858; reprinted in a critical edition by Daniel Goldshmidt, Tel Aviv: Devir, 1966]). In other words, there was a need for the shali’aḥ ẓibbur to literally prepare and rehearse what exactly he wanted to say! According to this, the widespread use of siddurim has drained the silent Amidah of its primary function in Ḥazal’s decree. On the debate among modern scholars, see Kavvana, pp. 257-266.
 I should mention that I don’t know Rabbi Sperber personally.
 It is excellent that Rabbi Frimer pointed this out, as so many people are unaware of it.
 An amazing example of this are the dozens of comments to Rabbi Frimer’s essay that debated minor phenomena like the censored sentence in the daily Alenu, i.e. matters whose halakhic significance is zero. These things can certainly be fascinating, and the great interest is therefore understandable. But the fact that so many people (including rabbis and Torah scholars) take these things so seriously in practice (beyond the level of fascination), and endlessly debate which siddur is preferable in terms of its nosaḥ, is testimony to how far we are from the ideal of halakhic prayer.
 However, I will note Rabbi Frimer’s translation of le-ḥaddesh bah davar as “to insert” (!) something fresh, which reflects an inherent bias towards the attitude described above (by assuming that the Gemara speaks of a fixed text into which it is possible to “insert” something). That it is indeed how most people understand the phrase today, but simply look at Rashi on that Gemara to find a different attitude entirely.
 Even for the Rambam and the Shulḥan Arukh, the basic sense of the explicit ruling is that personal petitions only belong in petitionary blessings; cf. Frimer, n. 37.
 See Tur Oraḥ Ḥayyim 112 and the sources cited in Beit Yosef.
 I’ve copied the passage with some minor stylistic improvements. What I really wish I could do is simply post the relevant chapters online in their entirety, correct and revise them when needed (after sitting stagnant for 14 years), and then let people use them and debate them as they see fit. Indeed, if anyone has any suggestions or advice about how I might be able to do so legally for a book that was published 14 years ago, or can help in terms of how to go about getting permission from the publisher if necessary, I would be very grateful. I am the copyright holder. I would further like to ask Rav Shilat for permission to make his chapter on this topic available online. If anyone has any personal contact with him and might be willing to help, please let me know.