A Moroccan Synagogue Service

Edwin Seroussi

This is a short extract of the article “Liturgy: An Overlooked Space in the Moroccan Jewish Musical Map.” This extract is accompanied by the recordings analyzed in detail in that article.

Research of Moroccan Jewish religious music has been largely, if not exclusively, dominated by the singing of paraliturgical Hebrew (and at times Judeo-Arabic and Aramaic) poetry. This poetry is set to Andalusian musical genres, not only to selections of Al-Ala, but also those from Western Algerian urban music called djiri or ghrnati from Oudja, Tlemcen and Wahran. Among these paraliturgical practices, the repertoire of the early morning winter vigils called bakkashot received particular attention in research as well as in the public projection of Moroccan Jewish music in Israel and international festivals.  The performance of the liturgy, that daily, weekly and yearly spiritual diet that is the duty of any observant adult male Jew, was consigned to the second tier of Moroccan Jewish music research. Although liturgical issues emerged during early fieldwork among Moroccan Jews in Israel, it was always a peripheral topic, and mostly focused on the slippage of paraliturgical music into the liturgy.

* Cover image: Ibn Danan Synagogue in Fez, Morocco.


Basics of the Moroccan Jewish Liturgy: Analysis of a Service

The Moroccan Jewish worship, namely daily, weekly and holiday prayers, has been hardly analyzed contextually. Unlike the prestigious paraliturgical piyyut, liturgical repertoires mark sonic difference between Jews and Muslims in Morocco. They comprise an intimate space of Jewish sound that links the Moroccan Jewry not so much to its immediate Muslim neighbors but rather to the diasporic Jewish commonwealth of Sephardic and Oriental pedigree and, to a certain measure, to modern colonial contexts, mostly French-Jewish. Yet, this intimate musical space of the Moroccan Jews still bears some traces of the non-Jewish surrounding. It is not deprived of “musical Andalucianisms” at different levels, some at the level of voice production only, others much deeper that comprise, in my opinion, a modern development. 

There are of course aesthetic reasons too for the lack of emphasis on the liturgical repertoires of the Moroccan Jews among the contemporary agents of musical production and music scholars. This music is highly functional; it is generated by the structures and contents of sacred texts whose ceremonial performance and clear-cut pronunciation precedes musical concerns. These texts are an assortment of very diverse sources and registers: biblical and oral law passages, post-biblical prayers in poetic prose and poems (piyyutim) spanning almost a millennium of literary history. This textual mixture creates a sonic tapestry that moves between fast non-metered mumbling on a single recitation tone to clear-cut florid melodies with fixed meter with various permutations of melodic and rhythmic formations in-between these two extremes. 

Analyzing synagogue services in detailed manner is the only method with the potential to unravel the mechanisms generating Moroccan Jewish liturgical performances.  The following analysis of a weekly evening prayer shows some analytical parameters designed to interpret the structure and aesthetics of these liturgical performances as well as the social mechanisms that generate them. I recorded this service in Tiberias on February 4, 1981, the eve of the New Moon (Erev Rosh Hodesh) of the month of Adar Alef, 5741.

Listen to this full service (total duration 21:35):


Detailed analysis of each section (by pressing "play" you reach the section of the service discussed in each paragraph; to return back to the same point use the two arrows icon). 

0:00 “Leshem yihud,” is a short kabbalistic prayer originating in the circles of the mystics of Safed. It proclaims the intention of performing the mitzvah, in this case the evening service “that Jacob our father instituted,” with the goal of stimulating the union of the spheres or emanations of the Godhead. In spite of opinions against its recitation, Moroccan Jews punctually perform it, an indication of the noticeable kabbalistic influences of their liturgical practices. The leader of the service (or of its opening section, as in this case) usually says it in a whispering recitation, almost like an incantation, around one tone and in a very fast tempo in rather soft dynamic. 

0:18 “Barekhi nafshi et Adonay” (Psalm 104) is the special Psalm for the eve of the New Moon according to the Sephardi liturgical usage. Verse 19 of this psalm, “He made the moon to mark the seasons,” is the justification for its inclusion in this service. The whole congregation sings it in unison with a two-part psalmodic formula, i.e. a musical phrase consisting two halves is repeated with each verse, its length depending on the number of syllables of each verse. There is a steady pulse and rhythm is determined by the difference between short and long notes corresponding to accented and unaccented syllables. The rather slow tempo gradually accelerates (2:04, 3:19). This melodic and rhythmic formula is identical to the one used to sing the special psalms that open the eve service of each Holiday according to the Sephardic tradition. By performing this psalm with this musical technique, the Moroccan community aligns itself with the general pan-Sephardic practice although the specificity of its intonation marks it as unequivocally Moroccan. 


5:09 “Adonay tzeva’ot imanu” (Psalm 46, 8; 84, 13; 20, 10). These three psalm verses comprise the opening formula of the daily evening prayer. They are performed with the same psalmodic formula as the previous psalm.


5:55 Half qaddish, performed by the cantor. The first part of this prayer is sung to the melody of the piyyut “Lindodkha, yedid naʿalah” by Israel Najara. This piyyut is included in the compendium Shir Yedidut, in the chapter corresponding to the Sabbath of the pericope “Vayehi” in the  tbaʿ (mode) Rasd dhil.

The original melody of the Half Qaddish, "Lindodkha yedid" sung by Rabbi David Kadosh

This melody is a favorite one among Moroccan Jews and is applicable to other piyyutim as well as sections of the Sabbath and holidays’ services.  At 6:54 the public recites the congregational response “Yehe sheme raba,” each individual with his own tempo and rhythm, thus creating a loud “sonic cloud.” The continuation of this qaddish by the cantor is performed in the customary fast non-metric recitative formula used daily, as are the other qaddishim below.


7:30 Ve-hu rahum (Psalm 78, 38; Psalm 20, 10). Two psalm verses that engulf the qaddish with the previous set of psalm verses. The first of these two verses, is a long one is sung to the melody of the previous qaddish, the one belonging to the piyyut “Lindodkha,” by the entire congregation, while the second verse, “Adonay hoshi’ah,” is sung by the cantor in a melismatic melody without clear beat that functions as a transition to the next section.

08:22 “Bareku et Adonay hamevorakh”. The call to public prayer sung by the cantor alone in slow tempo, without clear beat or meter. The entire congregation responds with a clearer beat, still without meter. The cantor then repeats the response solo, leading without any interruption to the next blessing.

08:42 Birkat hama’ariv ‘aravim, first blessing before “Shema.” Fast recitation of the cantor in loud voice, with special stress on the key words “ma’ariv aravim”. The phases “Um’sader et hakokhavim”, “U--mavdil bein yom u-bein layla” “Adonay tzeva’ot shemo” and the blessing formula itself (Barukh Ata Adonay...) are intoned by the entire congregation with a special melodic formula spanning a perfect fourth, in a call and response pattern with the cantor repeating the same words following the congregational response. This is a distinctive liturgical practice of many Moroccan Jews in Israel probably originating in the circles of the Ecole Normale Hébraïque of Casablanca since the early 1950s. 

09:32 Ahavat ‘olam, second blessing before Shema. The congregation intones the phrase “Ubahem nehege yomam valayla” in responsorial manner with the same pattern as the repeated words by the congregation in the previous blessing.

10:00 Shema Yisrael, a dense “sonic cloud” that overlaps with the last words of the previous blessing “ohev amo Yisrael” that are unheard due to the congregational irruption of the Shema prayer.  The cantor takes the lead back from the congregation on “ve-ahavta” singing the text loudly according to the cantillation signs. A much-elaborated ornamented performance of the ta’am (Masoretic accent) shnei gereshin is characteristic of the Moroccan cantillation system and is emphasized by the cantor by a slowing of the tempo. In the background, the congregation whispers the prayer parallel to the cantor. 

12:36 Adonay elohekhem/Emet ve-emunah, last two words of Shema are repeated by the cantor in a melismatic solo leading directly into the next prayer, the first blessing after Shema that is performed by a second cantor in fast tempo recitation within a narrow tonal range.

13:07 Ra’u banim, a section of “Emet ve-emunah,” is sung mostly by the children of the congregation as a choir with the rhythmic melody of the Arab song “Ya um al-‘abaya” (يا أم العباية) that was made famous in a Cairophone 78rpm recording by the singer and movie actress of Syrian-Lebanese origin Siham (or Seham) Refki (1922-2007).

The song is attributed to the Syrian composer Abdel Ghani al-Sheikh (1900-1970). However, it belongs to a genre of popular songs that draw heavily from folk styles from the Greater Syria region. It is widespread in Baghdad too, and even appears, by mistake, in the song catalogue of the Iraqi-Jewish musicians, the Al-Kuweity brothers, Daud and Salah. The section “Mi kamokha” (14:10) is also sung to the same melody by a child soloist.  This Arab song is obviously a latecomer to the Moroccan Jewish repertoire. It may have entered the Jewish repertoire through a para-liturgical piyyut, after its Egyptian commercial recording.

14:35 Malkhutkha, the second cantor returns to a fast recitation mode for the ending part of “Emet ve-emunah.” At 14:45 the words “Uge’alo mi-yad ḥazaq mimenu” are sung in a congregational response that is repeated by the cantor in the same manner as in the repeated words endowed with special kavvanot in the first and second blessings before Shema.

14:52 Hashkivenu, the second blessing after Shema is sung to the tune of the Eastern Mediterranean Judeo-Spanish song “La rosa enflorece” (aka “Los bilbilicos cantan”) by the children and the congregation up to “ve-hagen be’adenu” (15:46). At this point, the cantor takes over in a fast recitation that includes the final blessing (“Shomer et ‘amo Yisrael laʿad”) that leads into next section. The use of the Judeo-Spanish tune is most probably a recent musical acquisition that took place in Israel. The song was made famous by its recording by the popular singer Yehoram Gaon, whose family is of Eastern Sephardi origin. It is also widely sung, following a Jerusalemite custom, with the text of “Tzur mishelo akhalnu,” a piyyut for the Grace of the Meal. I have heard the tune of another famous song from Gaon’s repertoire, “Morenica” (“Sheḥarḥoret” in its Hebrew version also by Gaon), in Moroccan synagogues in Israel.


The original melody of Hashkivenu, "La rosa enflorece/Los bilbilicos cantan" sung by Manes Malka.

16:06 Yirʿu eineinu, a special third blessing after the Shema in the evening service. Rabbinical authorities contest its inclusion in the liturgy. Most, but not all, Sephardim include it in their services. Moroccans have a special version of it. The entire congregation recites this prayer with a short melody associated with the Judeo-Arabic song “Arsl yah l’wahad wahdani” a poem by the early seventeenth-century Moroccan born Rabbi Fradgi Shawat from Béja in Tunisia. This song is included in Shir Yedidot in chapter belonging to the Shabbat of the pericope Beshalah. It has a characteristic hemiola metric pattern of twelve beats divided in 2+3+3+2+2.

17:11 Half qaddish, with a different melody by the cantor in the same musical mode as the previous piece. The melody is adapted from the qasidah “El shokhen shamaiyma” from the bakkashot repertoire for Sabbath Va’era, a pericope that usually coincides with the beginning of the Hebrew month of Shvat. Such a selection “paints” the qaddish with a specific sound associated with that particular season of the year while clinging the liturgical with the paraliturgical repertoire.

At 18:14 the congregational response “Yehe sheme rabba mevorakh” generates another “sonic cloud effect,” that is followed by the last verse of the text recited by the cantor with a slight slowing down in the tempo.  The qaddish is followed up by a public delivery of a short verse recited at this specific point only on the eve of the New Moon, “Rosh hodesh li-vrakha le-hayyim tovim ul-shalom” (“A New Moon for blessing, good life and peace”). This verse echoes in its wording the Blessing of the Month, to be recited in full during the following morning prayer.

The original melody of the Hatzi Qaddish, "El shokhen shamayyima" sung by Rabbi David Edrey and students. Source: Mo'etzet ha-piyyut ha-maroqa'I (Moroccan Piyyut Council in Whatsapp).

At this point the Silent 'Amidah (lasting about three to four minutes, not recorded) took place.


Ending section of the service (Total duration 2:34)

18:56 “Yehi shem adonay mevorakh meʿata veʿad ‘olam” (Psalm 113, 2-4; Psalm 8, 2) cantor sings this psalm verses solo in fast recitation, without prosodic pauses between verses, unlike the psalm at the opening of this service.

19:05 Qaddish titkabal, extremely fast recitation by cantor and congregation without clear beat.

19:43 Shir hamaʿalot essa ʿenay el he-harim (Psalm 121), very fast recitation by whole community. Unlike the festive opening Psalm and other psalm verses throughout the service, in the present recitation there are no pauses between verses (let alone between the two halves of the verses) or a defined rhythmic pattern. Thus, the performance generates a “sonic cloud effect.”

19:54 Qaddish yatom, another extremely fast recitation by cantor and the mourners of the congregation replicating the sound effect of “Qaddish titkabal.”

20:35 Barekhu, Aleinu le-shabeaḥ, closing prayer of the service. The cantor opens but only the first two words of the prayer are heard as the whole congregation joins the fast recitation of this text, creating a “sonic cloud effect.”

21:03 Half qaddish, extremely fast recitation by cantor and congregation, fading out (or rather dissolving) into the end (21:30).


Analysis of the Service

Let us now summarize the detailed analysis of this service, starting with the human composition of the performers and moving into the sounds they generate. The performers were immigrants removed from their original habitat in Morocco where they were born and educated. Originating in diverse settlements in Morocco, many of them had experienced a transitional period in their lives when they moved to Casablanca and, to a lesser extent, other cities such as Rabat-Sale, Meknes, Fez, Marrakesh, Essaouira, Tetuan and Tangier on their way to immigration to Israel.  They were joined in the performance by a new generation of Israeli-born and Israeli-educated Moroccan Jews, the offspring of a still largely endogamous community. 
Thus, one can assume that the performing style and content of this service still reflect practices firmly rooted in the mother community in Morocco. My long-term fieldwork experience in Moroccan synagogues shows that this service is a clear instance of an extremely consistent pattern of prayer performance that goes back to the synagogues of large cities in the 1940s and early 1950s. The historical recording by Paul Bowles of an evening service in Meknes, 1959, attest to this remarkable consistency. Only the use of a tune from an Eastern Mediterranean Judeo-Spanish song discloses the specific spatial context and time of the service’s performance, Israel, 1981.
This consistent pattern of Moroccan Jewish prayer includes five distinctive genres of non-verbal sound patterning: 

1)    extremely fast non-metrical recitations on a very narrow range; 
2)    biblical cantillation with unclear beat and great contrast between fast recitation on a narrow range and lyrical isles of ornamented sound on a wider range; 
3)    psalmody with clear beat but no steady meter in a narrow range with intermittent heterophony; 
4)    metric melodies; 
5)    the “sonic cloud” effect. 

Performance includes solo singing alternating between two leading cantors, responsorial singing, and choral singing (of adults and children). This variety of sonic genres and manners of performance creates a dynamic stream of sound characterized by abrupt shifts in all the parameters, i.e. range, presence or absence of clear beat, density of texture, dynamics and performing voices. 

While most of these sonic shifts mark the boundaries of the literary sections of the service, i.e. different sound structures correspond to defined liturgical units, others reflect either older kabbalistic practices or modernizing trends. Kavvanot (“intentions”) embedded in certain passages or in specific words may receive a special musical treatment. On the other hand, the choral responses by the congregation in our recording, in a suspiciously Ashkenazi-tinged tonal structure, most probably derive as we have seen from much more modern traditions imported to Casablanca through the agency of renovating educators from Wahran. The absence of this distinctive feature from the Paul Bowles recording of Meknes, 1959, reveals how recent such musical innovations took roots in the liturgical practices of Moroccan Jews relocated in Israel.

This detailed analysis reveals the striking amount of historical, social and aesthetical information one can gather from a short, daily twenty-two minute service. Starting with the text of the service, one can observe unique textual additions (and omissions) that singles out the Moroccan practice not only musically but also textually. Put differently, the sense of liturgical belonging is reflected also in the use of specific editions of the prayer book that are still maintained in Israel. In this sense, there is a liturgical resistance to the hegemony of a state-sponsored Sephardic rabbinical establishment overwhelmingly dominated by rabbis of Iraqi and Syrian origin, most notably by the late Rabbi Ovadiya Yossef. Thus, the use of a particular liturgical formula is already a form of Moroccan (and more broadly Maghrebi) resistance to centripetal forces affecting the religious culture of the Jews from Islamic countries in Israel.

However, it is in the performance practice of this service, which includes its musical component, but not only it, that its Moroccaness is located. This sonic distinction, expressed in the specific sonic genres described above as much as in a particular selection of melodies, reproduces venerable practices from “the old country” as much as the taste of the specific performing community and its leaders whose voices were caught in a specific recording. 

The singing of the opening festive Psalm 104, for example, is quintessential to the Moroccan Jewish soundscape but also links it to a broader Sephardic custom. This liturgical music genre has no parallel in the Moroccan Arab repertoire and thus it is a marker of sonic Jewish difference. Yet, the choral performance with its intermittent split of voices in fortuitous polyphonies recalls the performance practice of choral singing in the Al-Ala repertoire. The performance of this festive Psalm, including the additional verses from other Psalms that follow it lasts for more than five and a half minutes, or a quarter of the entire service. Thus, this is a musical moment when the Moroccan Jewish synagogue reveals itself as a singular singing community. This participatory aspect is a staple of the Moroccan liturgy in comparison to other traditional Jewish communities where a soloist ḥazzan stands out.

Singing of specific texts to melodies adopted mainly from the non-liturgical repertoire comprise another notable feature of the Moroccan Jewish liturgy. I have called this slippage of the Moroccan bakkashot repertoire into the liturgy “Andalusianation.” This deeply rooted practice is audible in the service analyzed here in the application of the melody of the piyyut “Lindodkha” to the opening qaddish of the service. The musicalization of this specific qaddish, unlike most other qaddishim in the service that are pronounced in a very fast recitation formula (as in the ending session of this service), is customary in Holidays and other special occasions. In this entrenched Sephardic practice, melodies of piyyutim related to the occasion are recruited for the opening qaddish as if announcing musically the themes of the Holiday. For example, the melody of the piyyut “Aḥot qetannah” that opens the evening service of Rosh Hashanah is also applied to the opening qaddish of this Holiday. 

The reservoir of melodies borrowed for liturgical purposes, be it the qaddish or any other text that is traditionally musicalized (that is set to a distinct musical unit) such as “Ra’u banim” in the evening service, is not limited to those of liturgical or paraliturgical piyyutim. Arabic melodies such as “Qum tara,” a popular inqilab from nawba dhil of the Algerian ghrnati repertoire popularized among Jews by the recording of Reinette l’Orainesse, with the khruj (“exit”, fast ending) from the insiraf (fast section) of the Moroccan mizan quddam (rhythmic phase) of the nawba Istihilal, is a favorite piece of Moroccan cantors. It is adopted to liturgical texts such as the aforementioned “Ra’u banim” or “El adon ʿal kol hamaʿasim” from the morning service. We have also seen how in our service the Moroccan liturgy is also “painted” with pan-Sephardic sounds by applying a melody of a song from the Judeo-Spanish repertoire of the Eastern Mediterranean that became familiar to Moroccan Jews through the mass media. 

It is worth noting that the Moroccan practice of musicalizing liturgical “stations” can be attested to in written sources since at least the late eighteenth century. The important manuscript Oxford, Bodleian, Heb. F. 49 is exemplary in this regard. Dated at the earliest in 1796 and compiled by Maimon ben Yaakov Kadosh from Demnate with later additions by his disciple Rabbi Shlomo Cohen of Marrakesh, a master of the Andalusian Hebrew repertoire. Fundamental to our argument is the index of noʿamim (“melodies”) appearing at the beginning of this manuscript. Here the compiler assigns names of melodies of piyyutim included in his compendium to liturgical texts. This rich cross-reference between liturgy and the paraliturgical piyyut indicates the extent to which the Moroccan Jewish liturgy was musicalized already in the “old country” more than two centuries ago.

Finally, the analysis of the service shows that there is an “economy of liturgical time” in this performance. It means that there has to be a balance between musicalized and non-musicalized sections in order not to extend the services to extremes. The “price” paid for the extended musicalized sections in our service, such as the opening Psalm, the opening “Qaddish” or “Ra’u banim,” is compensated by the flying speed of the closing section, in which three qaddishim and two substantial texts, an entire Psalm and the important prayer “’Aleinu leshabeah” are all squeezed in two and a half minutes (!). This balancing of richly musicalized sections of the service with “non-musical” sections recited in a fast pace is yet another staple of the Moroccan Jewish liturgy.