From Medina Raḥamim to Elul Seliḥot:
Toward a Prehistory of Nocturnal Penitential Prayer

Aton M. Holzer

Shulamit Elizur, the undisputed doyenne of piyyut scholarship, published in 2016 a characteristically magisterial study on the origins of the seliḥot liturgy.[1] This liturgical category – like the prayerbook as a whole – is not attested before the eighth and ninth centuries. The great payyetanim of late antique Eretz Israel – Yose ben Yose, Yannai, and R. Eleazar ha-Qalir – composed qinot, qerovot, and other poetic forms, yet, as Elizur has demonstrated elsewhere,[2] seliḥot cannot be ascribed to this early Eretz-Israeli stratum. Contrary to earlier assumptions that located their genesis in the pre-classical Palestinian milieu, the genre is in fact a Babylonian creation, nurtured in a setting where “primitive” poetic styles persisted well into later centuries. The available evidence indicates that seliḥot entered the liturgical repertoire of the land of Israel only after the late eighth century, and even then only in a limited and sporadic fashion. From Babylonia, the genre subsequently radiated westward, leaving its imprint on Italy, Ashkenaz, and Sefarad.

The seliḥot as they have come down to us consist of two distinct components. The seliḥot proper are payyetanic compositions structured upon the framework of a primordial abecedarian proto-seliḥah preserved in the Seder of R. Saʿadyah Gaon. This early prototype interwove biblical verses drawn alternately from Torah, Writings, and Prophets. Three verses, however – the Thirteen Attributes (Exod. 34:5–6), the Prayer of Moses (Num. 14:19–20), and the Prayer of Daniel (Dan. 9:18–19) – were singled out for repetitive or responsive recitation, while the remaining verses were gradually supplanted by confessional passages (viduy) or thematic prose paragraphs. Elizur contends that the genre most plausibly derives from a pre-classical Palestinian text embedded in the Qedushat ha-Yom blessing for Yom Kippur, where prayers composed entirely of biblical verses were customary. In Babylonia, this early piyyut – and, in due course, the expanding sequence of piyyutim modeled upon it – was transplanted into the Amidah at the selah lanu blessing, thereby establishing the formal locus of the seliḥot within the liturgy.

A second, distinct component of the seliḥot service is the raḥamim – the repetitive litanies beginning with ʾEl raḥum shimekha. This section was not initially bound to the fixed liturgy but was instead recited in assemblies convened during the ashmoret ha-boker (“the third watch of the night,” before dawn) throughout the month of Elul, the Ten Days of Repentance, and Yom Kippur. As with seliḥot, this practice finds no precedent in the late antique liturgy of Eretz Israel, but rather emerges within the Babylonian milieu. In the course of time, the raḥamim were often relocated to the taḥanun following the morning prayer, and, once the seliḥot component was excised from the Amidah, the two elements were joined. Thus, on Yom Kippur, only the evening service retained raḥamim, while the other services preserved seliḥot. In Sephardic communities, the seliḥot rites that remained independent of the fixed liturgy – those recited at ashmoret during Elul and the Ten Days of Repentance – developed in a more fluid and less regimented manner: seliḥot-type piyyutim were freely interwoven with raḥamim-litanies, precisely because their origins lay outside the structured framework of the Amidah.

When, then, did these nocturnal gatherings originate? No documentary evidence attests to seliḥot or raḥamim before the eighth century. Nevertheless, there is reason to suspect that Babylonian Jewish communities had already consecrated the month of Elul to penitential supplication by the close of late antiquity. A fascinating recent study by Michael E. Pregill traces the evolution of traditions surrounding the Golden Calf episode across late antique Judaism, Christianity, and Islam. He demonstrates that the ḥet ha-ʿegel was a charged theme in Eastern Christianity from the fourth century onward, generating a distinctive cross-pollination between Jewish and Syriac Christian traditions in the Sasanian orbit.

A striking motif emerges in the writings of Ephrem of Nisibis (306–373) – who integrates certain midrashic elements, such as the murder of Hur and apologetic defenses of Aaron – namely, the portrayal of Israel as an unfaithful bride:

“The Holy One took the Synagogue up to Mount Sinai
He made her shine with pure white garments, though her heart was dark
She whored with the Calf (bĕ-ʿeglâ gārat), and He came to despise her
He smashed the Tablets, the book of her covenant (kĕtābâ da-qyāmāh).”[3]

This imagery appears to have acquired particular resonance in the Aramaic-speaking milieu, aided by the rendering of saru (Exod. 32:8) as satu in the targumim – a translation evocative of sotah – and echoed, albeit in less explicitly sexualized form, in the Peshitta. The motif was subsequently taken up by Syriac Christian theologians and exegetes such as Theodoret of Cyrrhus and Jacob of Serugh. As Pregill persuasively argues, rabbinic sources themselves came to embrace this image, precisely as both a product of and a response to this polemical environment.

It may therefore be that the month of Elul – the period traditionally associated with Moses’ ascent to atone for the sin of the Golden Calf – was observed with particular intensity, at least on an individual level, within the Babylonian milieu, where the Golden Calf served as a focal point of interreligious controversy.[4] A possible allusion to such a practice appears in Seder Eliyahu Zuta, a distinctive rabbinic midrashic work and moral treatise,[5] likely composed in Babylonia between the fifth and ninth centuries:

“Were it not for God’s forbearance toward Israel during the first forty days that Moses was up on Mount Sinai to bring the Torah to his people, the Torah would not have been given to Israel. By what parable may the matter be illustrated? By the parable of a mortal king who wed a woman he loved with utter love. Having sent for a man to act as a go-between between him and his future queen, he showed the emissary all his nuptial chambers, his halls of state, and his private living quarters. The king then said to the go-between: Go and say on my behalf to the lady, ‘I do not require anything from you. You need make for me only a small nuptial chamber where I can come and dwell with you, so that my servants and the members of my household will know that I love you with utter love.’ Yet even while the king was concerning himself with the measurements of the nuptial chamber [she was to make for him] and while he was dispatching a messenger to convey many, many gifts to the lady, people came and told him: ‘Your future wife has committed adultery with another man.’ At once the king put aside all the plans he had in hand. The go-between was expelled and withdrew confounded from the king’s presence, as is said, While the King was thinking [about the measurements] of His nuptial chamber, my spikenard let go [and lost] its fragrance (Song 1:14)… During the last forty days when Moses went up a second time to Mount Sinai to fetch the Torah, Israel decreed for themselves that the daytime hours of each day be set aside for fasting and self-affliction. The last day of the entire period, the last of the forty, they again decreed self-affliction and spent the night also in such self-affliction as would not allow the Inclination to evil to have any power over them. In the morning they rose early and went up before Mount Sinai. They were weeping as they met Moses, and Moses was weeping as he met them, and at length that weeping rose up on high. At once the compassion of the Holy One welled up in their behalf, and the holy spirit gave them good tidings and great consolation, as He said to them: My children, I swear by My great name that this weeping will be a joyous weeping for you because this day will be a day of pardon, atonement, and forgiveness for you – for you, for your children, and for your children’s children until the end of all generations.” (Ch. 4, Braude translation)

The precedent for individual supplications is reflected within the raḥamim-style litanies themselves – for example, “He who answered Abraham our father … to Isaac his son … to Daniel … to Ezra.” Daniel Boyarin has noted that the prayers of the narrator in the extracanonical 4 Ezra (late first century CE) bear striking affinities to the litanies, confessions, and invocations of the divine attributes that later characterize the seliḥot.[6] Even if one remains cautious about retrojecting the fully developed genre – absent from both Talmuds and unattested in the Palestinian West – back to so early a period, and even if many of Boyarin’s textual parallels must undoubtedly be assigned to later strata,[7] the evidence nonetheless suggests the possibility of an early use of penitential litanies in private devotion. This, in turn, stands in contrast to their later employment within the communal ʿAmidah of fast days, as recorded in m. Taʿanit 2:4.

Other, later developments within “Babylonian territory” – regions under the halakhic and tributary jurisdiction of the Babylonian center by the close of late antiquity – may likewise hint at the existence of such practices and illuminate stages in their historical evolution.

Recent scholarship suggests that rabbinic Judaism maintained a significant presence in the Arabian Peninsula, at least in its northern regions – possibly even reflected in the toponym Khaybar (deriving from ḥaver[8]) – as well as in the Jewish city of Yathrib/Medina, which some now identify as a community of halakhah-observant, rabbinic Jews.[9] This line of research is relatively recent, its conclusions necessarily tentative, yet it opens promising avenues for further inquiry.

While the midrashic material drawn upon by the Qurʾan appears to be of Eretz-Israeli provenance[10] – attesting to the earlier dominance of the Eretz-Israeli center – the Arabian Peninsula after 570 was largely under the sway of the Sasanian Empire or its clients, with direct Sasanian control extending as far south as Najrān and Jeddah. The Ḥijāz itself is later described as long having functioned as a “backyard” of the Babylonian academies:[11]

“’Abdur Rahman bin ‘Abdul Qari said, ‘I went out in the company of ‘Umar bin Al-Khattab one night in Ramadan to the mosque and found the people praying in different groups. A man praying alone or a man praying with a little group behind him. So, ‘Umar said, ‘In my opinion I would better collect these (people) under the leadership of one Qari (Reciter) (i.e. let them pray in congregation!).’ So, he made up his mind to congregate them behind Ubai bin Ka’b. Then on another night I went again in his company and the people were praying behind their reciter. On that, ‘Umar remarked, ‘What an excellent Bid’a (i.e. innovation in religion) this is; but the prayer which they do not perform, but sleep at its time is better than the one they are offering.’ He meant the prayer in the last part of the night. (In those days) people used to pray in the early part of the night.’” (Ṣaḥīḥ al-Bukhārī, Vol. 3, Book 32, Hadith 227)

The “Prophet’s Mosque” of Medina, in its original form, seems to have resembled a synagogue, with the qibla (direction of prayer) even initially facing Jerusalem. In the first decades of the seventh century, before the coalescence of an “Islamic” identity, Jews might actually have been included in the proto-Islamic muʾminūn (“believers”) movement.[12] This hinges, to some extent, on whether the word characterizing the Jews’ relationship to the new movement is umma (“[one] community”), as Fred Donner reads it, or amāna (“secure”), as Michael Lecker would have it.[13]

Regardless, in the early period of Islam Jews and Christians seem to have joined the “community” without having to recant their previous faith or identity, and there was no “conversion” rite or procedure.[14]

The brief “honeymoon” between Muhammad and the Jewish tribes of Yathrib ended abruptly, culminating in the bloody execution of some four hundred men of the Jewish tribe of Banū Qurayẓa.[15] More recent revisionist scholarship, however, tends to regard this episode – attested only in sources written a century or more later – as altered,[16] exaggerated, or even fabricated, serving primarily the purpose of boundary-setting.[17] These retellings were likely shaped by later Muslim conflicts with Jewish communities in Damascus and Baghdad, while the original events, insofar as they occurred, are better understood as a matter of political expediency rather than religious confrontation.[18]

In any case, many Jews appear to have remained in Medina for several decades thereafter. The expulsion of non-Muslims from the Ḥijāz – the western province of Arabia – was long attributed to the second caliph, ʿUmar ibn al-Khaṭṭāb (r. 634–644), who was also credited with the shurūṭ ʿUmar (“Pact of ʿUmar”), the charter assigning dhimmī status to Jews and Christians and thereby marking clear boundaries between them and the emergent Islamic polity. More recent research, however, tends to assign both the expulsion and the Pact not to ʿUmar I, but rather to ʿUmar ibn ʿAbd al-ʿAzīz (r. 717–720).[19]

Jewish sources themselves offer little contemporary testimony datable to this formative period. One exception is an apocalyptic work composed around the mid-eighth century, Nistarot de-Rabbi Shimon bar Yoḥai, pseudepigraphically attributed to R. Shimon bar Yoḥai – cast here as the fiercest rabbinic adversary of Rome.[20] This work strikingly celebrates ʿUmar ibn al-Khaṭṭāb, conqueror of Jerusalem from the Byzantines, as a beneficent figure whose reign would prepare the ground for Israel’s redemption. The text prophesies that the Ishmaelite regime would vanquish Christendom and create the conditions for the advent of the Messiah:

“The second king who will arise from Ishmael will be a friend of Israel. He will repair their breaches and mend the breaches of the Temple; he will shape Mount Moriah and make the whole of it a level plain. He will build for himself there a place of prayer (hishtaḥavayah) upon the site of the ‘foundation stone’ (ʾeven shetiyyah), as Scripture says: ‘and set your nest on the rock’ (Num. 24:21). He will wage war with the children of Esau, slaughter their troops, and capture a great multitude of them. And in the end, he will die in peace and with great honor.”

ʿUmar acquired the epithet al-Fārūq (“the redeemer”) – from the same root as purqan – most likely through Jewish usage.[21] The early Islamic movement may at first have been perceived in terms akin to contemporary Noahides: strict monotheists who acknowledged Israel’s chosenness and with whom joint prayer posed no obstacle. ʿUmar’s acceptance of the Byzantine surrender in 637 and his subsequent activities in Jerusalem – including the readmission of Jewish residents, the clearing of refuse from the Temple Mount, and the construction of an initial “prayer-house” open to all worshippers, Jews among them[22] – could only have deepened Jewish admiration for him and for the movement he led in its formative environment.

It was during ʿUmar’s caliphate that the month of Ramadan, having been unmoored from the solar calendar after 632, fell successively in Iyyar,[23] then in Tishrei (637–639), and thereafter in Elul (640–642). Jews observing the Babylonian rite of raḥamim by night would naturally have assembled in Medina’s spacious prayer-house during the nights of Tishrei and Elul/Ramadan. Before the incorporation of seliḥot, these supplications were recited individually. ʿUmar evidently regarded the practice as salutary but preferred that it be performed collectively; hence, alongside the adoption of seliḥot, the communal ashmoret service became standard across lands under Babylonian influence. Intriguingly, however, the seliḥot embedded in these ashmoret gatherings retained an individualistic tenor for centuries – most clearly visible in the piyyutim of R. Isaac ibn Ghiyyat, composed especially for the nights of Elul. Thus, even as the ashmoret gatherings assumed a communal form, they continued to be conceptualized as assemblies of individuals.[24]

The Geonic treatment of this custom, preserved in R. Isaac ibn Ghiyyat’s Hilkhot Teshuvah (§58), records that both the ninth-century Gaon R. Kohen Ẓedeq b. Abimai of Sura and the renowned eleventh-century R. Ḥayya Gaon of Pumbeditha restricted the practice to the Ten Days of Repentance in their own locales. Yet R. Ḥayya also concedes that in “some places in Persia” the recitation of taḥanunim extended through the entire month of Elul, and R. Isaac ibn Ghiyyat notes that such was likewise the practice in his native al-Andalus. To one line of speculation, then, another may be added: the custom of seliḥot throughout Elul – attested both in Persia at the eastern edge of the Jewish world and in Spain at its westernmost reach, both within the orbit of Babylonian influence – appears to have “skipped over” the very center itself, namely, the Geonic academies of Babylonia.

As Simcha Gross has demonstrated,[25] the Iggeret of R. Sherira Gaon, in recounting the Geonic academies’ greeting to the Islamic conquerors, reflects the rising prominence of the ʿAlids as the Abbasid caliphate matured; in this recension, it is specifically Caliph ʿAlī whom the Geonim are said to have welcomed with great affection. By the late tenth century, the Shiʿite Būyids had assumed control of Baghdad and much of the Iraq–Iran region. Within Shiʿi circles, the prayer service introduced by ʿUmar – the Tarāwīḥ – was condemned as an illegitimate innovation, and both Fāṭimid and Būyid rulers actively suppressed its observance.[26] Indeed, even today Shiʿi sources continue to enumerate its perceived deficiencies: that it is performed at the wrong time (in the early evening rather than at the close of night), that it diminishes the primacy of individual prayer, and that it is conducted with undue haste.[27] It is not difficult to imagine that similar criticisms contributed to the suppression of month-long nocturnal supererogatory prayer within the Geonic academies of Baghdad and its environs – the very setting in which anti-Tarāwīḥ polemics were most vigorously advanced. Moreover, the month of Elul coincided with the kallah, when vast numbers of students converged upon the Geonic centers for intensive study, thereby directing the academies’ energies away from nocturnal penitential assemblies at precisely this season.

Notes:

Many thanks to Prof. Y. Tzvi Langermann, Prof. Shulamit Elizur, and Prof. Gabriel Said Reynolds, for their valuable comments on several of the ideas discussed in this article. Needless to say, all errors remain the sole responsibility of the author.

[1] Shulamit Elizur, “The Origins of the Selihot Piyyutim,” Tarbiz, vol. 84, no. 4 (2016): 503-542 (Hebrew), available here (https://www.academia.edu/36608670).
[2] Shulamit Elizur, “The Character and Influence of the Babylonian Center of Poetic Production: Considerations in the Wake of Tova Beʾeri’s Books,” Tarbiz, vol. 79, no. 2 (2010-2011): 229-248 (Hebrew), available here (https://www.academia.edu/36608504).
[3] Hymns on Faith 14.6, translation in Michael E. Pregill, The Golden Calf between Bible and Qur’an: scripture, polemic, and exegesis from late antiquity to Islam (Oxford University Press, 2020) 216.
[4] Fascinatingly, in Second Temple times the month of Elul appears to have been the season when numerous calves – specifically, since ovine and caprine births occurred earlier in the year, prior to the preceding goren – were brought to the Temple for the tithe of animals (maʿasar behemah). One may wonder whether a symbolic connection was drawn between this practice and the atonement for the sin of the Golden Calf observed at this time of year. See Ze’ev Safrai, Mishnat Eretz Yisrael: Tractate Bekhorot, with Historical and Sociological Commentary (Yavneh: Kvutzat Yavne, 2020), 297–298 (Hebrew).
[5] Lennart Lehmhaus, “‘Were not understanding and knowledge given to you from Heaven?’ Minimal Judaism and the Unlearned ‘Other’ in Seder Eliyahu Zuta,” Jewish Studies Quarterly, vol. 19, no. 3 (2012): 230-258, available here (https://www.academia.edu/1817892).
[6] Daniel Boyarin, “Penitential Liturgy in 4 Ezra,” Journal for the Study of Judaism, vol. 3, no. 1 (January 1972): 30–34, available here (https://www.academia.edu/36253296).
[7] Shulamit Elizur, “The Character and Influence,” p. 243, fn. 54.
[8] See discussion in Raphael Dascalu, “Revisiting the Qur’anic aḥbār in Historical Context,” Journal of Qur’anic Studies, vol. 23, no. 2 (2021): 41-65.‏
[9] See Haggai Mazuz, The Religious and Spiritual Life of the Jews of Medina (Leiden: Brill, 2014).‏
[10] Holger M. Zellentin, “Aḥbār and Ruhbān: Religious Leaders in the Qurʾān in Dialogue with Christian and Jewish Literature,” in Angelika Neuwirth and Michael Sells, eds., Qurʾānic Studies Today (New York: Routledge, 2016), 258-289, available here (https://www.academia.edu/34810735).
[11] See Robert Brody, The Geonim of Babylonia and the Shaping of Medieval Jewish Culture (New Haven: Yale University Press, 1998), 125.
[12] See Fred M. Donner, Muhammad and the Believers: At the Origins of Islam (Cambridge, MA: Harvard University Press, 2010).‏
[13] See summary in Mark R. Cohen, “Islamic Policy Toward Jews from the Prophet Muhammad to the Pact of ‘Umar,” in Abdelwahab Meddeb and Benjamin Stora, eds., A History of Jewish-Muslim Relations: From the origins to the present Day (Princeton University Press, 2013), 58-77, available here (https://www.academia.edu/37423214).
[14] Ilkka Lindstedt, Muḥammad and his followers in context: the religious map of late antique Arabia (Brill, 2024), ‏319.
[15] Meir Jacob Kister, “The Massacre of the Banū Qurayẓa: A Re-Examination of a Tradition,” Society and Religion from Jahiliyya to Islam (Routledge, 2022), VIII 61-VIII 96.‏
[16] JaShong King, “The Message of a Massacre: The Religious Categorization of the Banū Qurayẓa,” Ancient Judaism, vol. 6 (2018): 203-226.‏
[17] See Mohammed Ahmed, “The Literary Role of Jews in Qur’anic Exegesis,” (Ph.D Diss., University of Cambridge, 2025).‏

Gabriel Said Reynolds, who serves as the Jerome J. Crowley and Rosaleen G. Crowley Professor of Theology and Assistant Professor of Islamic Studies at the University of Notre Dame, expresses skepticism regarding the traditional accounts in the Sīra literature about the Jewish tribes of Yathrib and their interactions with Muhammad and his followers. Nonetheless, he emphasizes that the existence of Jewish tribes in the Hijaz is beyond dispute, a fact confirmed by the discovery of rock inscriptions in the region (personal communication).
[18] See e.g. Juan Cole, Muhammad: Prophet of Peace Amid the Clash of Empires (Hachette UK, 2018).‏
[19] Milka Levy-Rubin, Non-Muslims in the Early Islamic Empire: From Surrender to Coexistence (Cambridge University Press, 2011); and Harry Munt, “‘No two religions’: Non-Muslims in the early Islamic Ḥijāz,” Bulletin of the School of Oriental and African Studies, vol. 78, no. 2 (2015): 249-269.‏
[20] See the discussion by John C. Reeves, Trajectories in Near Eastern Apocalyptic: A Postrabbinic Jewish Apocalyptic Reader (Atlanta: Society of Biblical Literature, 2005), 76-77. Translation in ibid., p. 81-82.
[21] Suliman Bashear, “The Title ‘Fārūq’ and Its Association with Umar I,” Studia Islamica, no. 72 (1990): 47-70.‏
[22] Beatrice St Laurent, “Discovering Jerusalem’s First Mosque on the Haram al-Sharif and Capitalizing Jerusalem in the Seventh Century,” Bridgewater Review, vol. 36, no. 1 (2017): 23-28.‏
[23] Ben Abrahamson and Joseph Katz, “The Islamic Jewish Calendar: How the Pilgrimage of the 9th of Av became the Hajj of the 9th of Dhu’al-Hijjah,” Paper presented at the‏ Jamalullail Chair for Prophetic Sunnah International Conference (JCICI), Malaysia, October, 2020, available online here (https://www.alsadiqin.org/history/The%20Islamic%20Jewish%20Calendar.pdf).
[24] Ariel Zinder, “‘There They Stand at Midnight, Time and Again’: Selihot for Repentance Nights by Yitzhak Ibn Giyyat; A Critical Edition with an Introductory Essay and Literary Analysis,” (Ph.D Diss., Hebrew University of Jerusalem, 2014; Hebrew).
[25] Simcha Gross, “When the Jews Greeted Ali: Sherira Gaon’s Epistle in Light of Arabic and Syriac Historiography,” Jewish Studies Quarterly, vol. 24, no. 2 (2017): 122-144, available here (https://www.academia.edu/33854404).
[26] Christine D. Baker, Medieval Islamic Sectarianism (Leeds: Arc Humanities Press, 2019), 55. Incidentally, the Sunni call to the morning fajr prayer – al-ṣalāt khayr min al-nawm (“prayer is better than sleep”) – bears a striking resemblance to the opening of the piyyut of unknown provenance, Ben adam mah lekha nirdam; qum qera be-taḥanunim (“O mortal, why do you slumber? Arise and call out in supplication”)
[27] See, e.g., “Chapter One: The lies, innovations & conjectures behind Tarawih,” available here (https://shiapen.com/comprehensive/tarawih-a-parody-of-prayers/chapter-one-the-lies-innovations-conjectures-behind-tarawih).