The pellizco (pinch)

There are these two young fish swimming along and they happen to meet an older fish swimming the other way, who nods at them and says ‘Morning, boys. How’s the water?’ And the two young fish swim on for a bit, and then eventually one of them looks over at the other and goes, ‘What the hell is water?’

David Foster Wallace (2005)

Flamenco performance is a commercialized spectacle, but its language is that of communal ritual. Footnote 1 All flamenco practice centres around the singer and the repertoire that they interpret, passed down from generation to generation within specific communities, for whom these lyrics are important markers of identity (Chinoy 2022). The flamenco singer must essay a soaring and improvisational vocal journey, must pelear con el cante (fight with the song), a high-stakes sortie which entails wresting an ancient song into the vivid present. Embroidering a melodic line with modulations ranging from whispers to shouts to coughs to sobs, keeping to complex rules of rhyme, rhythm, melody, and verse structure, the singer must also sound the verse’s provenance by citing past interpreters. Footnote 2 Even in theatrical performance there is an openness of form, an unfolding dialogue between the artists that for flamencos is essential. Flamenco must decir algo, must say something; to perform rote shapes without this dimension of jointly making meaning is seen as ersatz, devoid of aesthetic value. Flamenco singers are thus venerated as voices of the community.

As someone who has devoted fifty years to its study and practice, I have observed a continual evolution in the forms of flamenco, in tandem with far-reaching changes in performance context and medium. The artistry and creativity of individual reinterpretations and reinventions of the canons continually drive flamenco into the future. Notwithstanding, with the passage of time, I have also noted that flamenco’s embodied semantics, what we might consider its aesthetic or epistemic frame of reference, along with the basic syntactical structures within which these semantics operate, have remained relatively constant and stable. Indeed, it is fair to say that anyone performing flamenco anywhere in the world today is ultimately referencing these codes—even if only to bend or break them.

And the codes which structure all flamenco performance revolve around the letra or sung verse. Far beyond merely determining the unfolding of the singer’s interpretation, the lyric shapes an emotional passage shared by all participants—singers, guitarists, dancers, palmeros (hand-clapping percussionists), and listeners alike—to rematar el cante, or finish off the verse. For flamencos, this musical, poetic, and gestural system, like language, like the repertoire, is a cultural legacy, something that they have grown up with. And the ¡ole! which non-natives often conjure when imagining flamenco is indeed central to this process. It is an exclamation which marks the contours of the lyric journey, and a celebration of arrival at its conclusion.

The deep description presented here is of a musical and corporeal gesture that I call the pellizco. Flamencos often use this term in a generalized way, to denote an aesthetic value. Pellizco means ‘pinch,’ and flamencos think of it as an expressive quality. Tener pellizco, to have pellizco, is the ability to grab ahold of the song (or the dance) in such a way that it is felt instantaneously and intensely by the entire flamenco circle. The pellizco is a shock wave of energy and emotion, an embodied reverberation of feeling that lights up the song and gives it specificity and power. Sometimes the pellizco can illuminate the doubled or ambiguous intention of a verse and sometimes, as poet Federico García Lorca writes, it is the ‘scorched throat’ able to ‘kill all the scaffolding of the song,’ leaving ‘not forms but the marrow of forms, pure music with a body lean enough to stay in the air’ (Lorca 1998, 61–62).

If flamencos use the term ‘pellizco’ in a generalized way, to describe a basic aesthetic value, I have, over many decades of teaching improvisation in the U.S., come to use this term to describe a specific melodic, corporeal, and rhythmic gesture. Footnote 3 At this level, what I call the ‘pellizco’ is a rhythmic signal, a key component of improvisational syntax. Footnote 4 Flamenco’s improvisational play often involves extending the phrase; the pellizco signals this by pulling upward and outward in order to keep the phrase open, down to the ground in order to rematar or close the phrase.

My proposal here is to use this flamenco dance gesture as a hook with which to approach medieval dance. First, I situate the pellizco as an embodiment of the sung flamenco verse, showing how it is inseparable not only from musical and verse syntax, but also from the dimensions of meaning and memory whose emotional import conditions this gesture’s transmission. Then, I cast back over a thousand years, to Islamic Iberia, and describe the emergence of the ancestors of flamenco’s strophic forms as citations of the (Christian) vernacular by the ruling (Muslim) court. Finally, I trace the transmission of these syncretic verse structures, along with the process of citation or contrafacta itself, across the gradual reversal in hegemonic polarity between Muslim and Christian domination of the Iberian Peninsula. From the lyrics of Ibn Quzmán (Córdoba, 1078–1160) into the ca. 1280 Cantigas de Santa María, strophic structures are retained—despite the representational flip in whom these verses cast as subaltern.

As Manuel Pedro Ferreira has described (2014), the strophic structures to which the flamenco pellizco relates are artefacts of the transmission of contrafacta as a lyric technique. These strophic structures are also artefacts, I argue here, of pastoralism itself (in the broadest sense of courtly representations of the vernacular) as a political trope. That is, in flamenco’s gestural language I read the splendidly facetious pastoralism of medieval Iberian lyric. In this regard, this essay continues the project I began in my first book, Sonidos Negros: On the Blackness of Flamenco, aiming to ‘celebrate and foreground flamenco’s hip-hop equivocations, its soulful intent, its sacred communal dimensions, [and] the satiric arrows of its virtuosic spectacles’ (Goldberg 2019, 3). To consider living gestures, such as the flamenco pellizco, ‘capable,’ as performance theorist Rebecca Schneider suggests, ‘of registering in the annals of history’ allows us to pierce the supposedly impermeable barriers separating Christian and Islamic, elite and vernacular, literary and oral worlds (2011, 39). The pellizco, I argue, allows us to study, as the David Foster Wallace parable that opens this essay trenchantly illustrates, the ways we breathe in the past, and the past breathes within us.

The pellizco in relation to strophic structure

The ‘grain’ of the voice is ‘the materiality of the body speaking its mother tongue.’

Roland Barthes (1988, 182)

The pellizco describes a fundamental aesthetic, emotional, and epistemological orientation toward the sung verse. As such, it is a deep structure that toggles open what Schneider calls a ‘cross-temporal space’ (2011, 39). When flamenco finds its groove, the pellizco’s vivid present is not only a ‘reiterative response’ to the past, ‘as a copy appears to cite a so-called original,’ but also calls forward in time, to ‘the past’s future reply.’ The pellizco, applying Schneider’s theorization, ‘dislocates the linear order of presents.’ Stitching past into future, it vibrates with ‘transgenerational memory.’ Much more than a corporeal or musical ‘gesture’ in the narrow sense of the term, it is a shock of recognition in which ‘“now” speaks or calls forward to (and through) “then”’ (Schneider 2011, 123). Simply in existing as a fundamental code, therefore, the pellizco manifests the past in the embodied present.

The syntax of the pellizco, of when to open the phrase by going up and when to close it by going down, always relates to rhythm and melody, but also to the strophic structure of the verse, adding a layer of considerable complexity and virtuosity. Flamencos know the oral songbook already; it is part of their culture. And yet the process of listening to the interpretation of a traditional verse is never predictable or rote, as listening to ‘Jingle Bells’ is. It is always a dramatic, emotional, anticipatory, and very active process, requiring not only deep ‘local’ knowledge, but also rapt attention in order to follow the intersecting twists and turns of verse, melody, and rhythm, and to say ¡ole! at precisely the right moment.

The flamenco verse’s strophic structure echoes that of medieval Iberian lyric. Flamenco letras, or verses, vary greatly, but their structure can usually be thought of as having three tercios (‘thirds’: lines, or sections), which I call ‘A,’ ‘B,’ and ‘C.’ The first part of the verse (‘A’) is where the singer has the most freedom to improvise—to elaborate, invent, and set up the joke or the tragic tale. It lines up with what some medievalists call the mudanza (literally, movement), the solo and virtuosic section of medieval Iberian lyric (Beltrán 1984). Footnote 5

The flamenco ‘B’ and ‘C’ lines usually function together as a unit, sometimes called the coletilla (tag), or estribillo (refrain). The ‘B’/’C’ lines are the punchline, as it were. They are much more rhythmically and melodically predictable than the ‘A’ line, and so they lend themselves to louder and more emphatic participation and commentary from the flamenco circle. Melodically, if the first part of the flamenco letra generally begins by arching up and outward, this part of the verse signals the all-important ending by stepping downward toward the ‘home tone.’ Some flamencos call the ‘B’ line’s initial descent in melody the cambio (change), because the guitar accompaniment changes chord here. And this melodic turn lines up with what some medievalists call the vuelta (return), which similarly signals the coming ending of the verse, and the recapitulation of the refrain (Beltrán 1984).

The flamenco estribillo is not technically a refrain; neither the words nor the melody are necessarily repeated exactly, as they are in ‘Jingle Bells.’ But this ‘B’/’C’ part of the verse functions semantically as a refrain—it signals the all-important remate of the letra, the conclusion of the verse. What I call the pellizco is located here. If at the microcosmic, rhythmic level the pellizco signals the turn of a phrase upward to continue or downward to conclude, at the macrocosmic level of the verse’s structure, the pellizco signals a transition between the upward flight of the verse’s opening mudanza and the remate, its landing.

Flamencos signal their comprehension of the junctures where the singer chooses whether to continue her improvisatory ascent or to turn downward toward the conclusion of the verse by commenting rhythmically (with palmas), melodically (with the guitar), and verbally with jaleos (rhythmic vocalizations such as ¡ole!). I read an echo of Islamic Iberia, al-Andalus, in this improvisational practice. ‘When a man listens to the notes of a melody, plays them on an instrument as an accompaniment to poetry, and makes the effort to understand them,’ writes philosopher and lyric poet Ibn Bājja, or ‘Avempace’ (ca. 1090–1138), in an epistle on melody called the Risalat al-alham, ‘the music penetrates to the deepest level of his being and purifies him’ (cited from Cortés García 1996, 15; Goldberg 2023c).Footnote 6 In other words, in al-Andalus, as in flamenco today, the drama, pleasure, and virtue of listening to the interpretation of a given verse was wrapped up with the task of following—that is, comprehending—its trajectory.

¡Ole los obreros flamencos! (¡Ole to flamenco construction workers!)

In a very short clip that I randomly came across on Facebook, five construction workers, gathered in the bright sun around a wooden table, sing a single verse. Footnote 7 There is no guitar. Instead, as is customary in flamenco spaces, we are oriented to the rhythm by palmas, by rapping on the table and stamping of the feet, and by jaleos. One of the men sings a letra (see Fig. 1):

Figure 1
figure 1

Basic verse structure

Full size image

We see in this case that the ‘B,’ the penultimate line, contains a change in rhyme. That is, ‘madre’ (‘A’) and ‘grande’ (‘C’) rhyme, but ‘juramento’ (‘B’) does not. Footnote 8 As mentioned above, this change in rhyme, which some medievalists call the vuelta or return, maps onto the cambio, the melodic change signalling that the soloist has begun his descent toward the home tone and the conclusion of the verse. In other words, if in the ‘A’ the soloist tends melodically upward, the ‘B’ turns toward the home tone, priming the listener/participants to anticipate the ‘C’ line and the verse’s imminent conclusion.

The last line of this verse (‘C’) is ‘merece un castigo grande’ (deserves a terrible punishment). The rhythmic juncture that I call the pellizco falls at the mid-point of this line, on the syllables ‘casti’—this is where the remate or conclusion actually happens. All the men punch this ‘ti’ upward: the melody tugs upward, and the men in the circle stamp this beat with accents and counter-accents. One man, seated in the right foreground, opens his hands and releases his arms outward, as if letting the pellizco fly away.

The next and final word, grande, the conclusion of the verse, is left unstated—everyone knows what it is. In its place, everyone in the circle marks this beat downward with an exuberant ¡ole! And the man who had released the pellizco outward on the previous accent now snatches it back, clapping his hands together and pulling the beat into his torso with a joyful ¡ah!

Snatching the unstated final phrase of the verse into his hands and concave chest, the man fluidly turns and lifts, releasing the gathered tension of the ending outward again. His gesture finishes on the beat where, had it been stated, the ending of the line would have been. The singer stands to leave the table, and the men turn to smile at the camera, happily commenting ‘¡qué bonito!’ (how lovely). That is, the singer doesn’t conclude the verse—the group does. I read here too echoes of the Andalusí (from al-Andalus) lyric: the entire poem is calqued upon it, yet the xarcha, the ending refrain, is sometimes left unstated.

As seen in the analysis presented here, the gestures structuring the flamenco verse are not only corporeal but also rhythmic and melodic. They celebrate the verse as an embodied episteme, a transgenerational and transtemporal shock of recognition in which, to use Schneider’s beautiful phrase, ‘“now” speaks or calls forward to (and through) “then”’ (2011, 123). Like the building these workers are constructing, their present references their past and reaches intentionally toward their future, toward the continuity of a tradition, identity, and a sense of belonging.

The shape of memory: the pellizco as vuelta in the Andalusí strophe

The xarcha is the pepper of the moaxaja; it is its salt, its sugar, its musk, its amber.

Ibn Sanā’ al-Mulk (Egypt, ca. 1155–1211)

(Cited from García Gómez 1962a, 48) Footnote 9

Medieval Arabic accounts document that lyric poetry was valuable ‘currency’ in what historian Brian Catlos calls the ‘economy of wisdom’ that shaped courtly culture in ninth-century al-Andalus (Catlos 2018, 89–90). And lyric influence flowed in multiple directions. Islamic influence travelled westward, in part via artists educated in the Middle East, such as the famous Ziryab (d. ca. 857). Conversely, already by the ninth century there is a suggestion that distinctive Andalusí lyric forms had emerged and begun filtering eastward (Zwartjes 1997, 46, 66; Reynolds 2000, 183–85; Gómez Muntané 2001, 326–27). Thus, two of the most important extant sources on the Andalusí lyric are from the East. Ibn Sanā’ al-Mulk is the Egyptian author of the famous quote cited above on the xarcha, the ending refrain in a longer lyric form called the moaxaja (García Gómez 1962a).Footnote 10 Tunisian poet and chronicler Ahmad al-Tīfāshī (1184-1253) wrote about the moaxaja’s sister-form, the zéjel (García Gómez 1962b; Liu and Monroe 1989).Footnote 11

Lyric poetry was of capital importance in al-Andalus, both in sacred spheres and also at court and among those aspiring to climb the social ladder. As medievalist and musicologist Reynaldo Fernández Manzano observes, although this lyric poetry was less rigidly structured than its classical Arabic models, it was governed by the standards of good taste and an educated ear (Fernández Manzano 2015, 84–85). As Avempace writes in his Risalat, the practice of listening to and comprehending lyric poetry was invested with values that cultivated intellectual, spiritual, and we might even say moral development in the practitioner/listener—values that seem very familiar to flamencos (Cortés García 1996, 15).

The verse forms of al-Andalus were unique and perhaps influential throughout Europe in two important and intertwined ways: their rhyme system and their strophic structure. The Arabs among whom Islam arose and whose culture would be dominant in Islamic lands such as al-Andalus had an oral tradition whose lyric poems, Dwight Reynolds explains in The Musical Heritage of al-Andalus (2021, 29), possessed a unique feature: the last syllable of every line rhymed. If it did not, it was not considered poetry.

Like the long mono-rhymed qasida in classical Arabic, Andalusí lyrics were sung. Footnote 12 But the Andalusí lyrics were structured, unlike their Arabic models, into strophes, or stanzas, plus an estribillo, or refrain. As Benjamin Liu and James Monroe describe in their essential study of Ten Hispano-Arabic Strophic Songs:

A soloist would sing the refrain to the audience, who would then repeat it in a chorus, thereby learning it by heart … Then the soloist would sing the first strophe. Its vuelta’s rhyme, which coincided with the refrain, would serve as a cue to the audience that it was their turn to repeat the refrain in chorus. (Liu and Monroe 1989, 16)

This strophic structure may itself be an indication of appropriation of the vernacular. That is, the elite audiences who would have heard these lyrics would have associated their stanza+refrain organization with popular culture. Samuel Armistead, Monroe, and Liu postulate that, as part of the earliest consolidation of Islamic hegemony in Iberia, ‘illiterate Arab poets, possibly bilingual minstrels who knew the tunes and words’ of the indigenous Christian folksongs, might have begun to compose new songs ‘in colloquial Arabic or in a mixture of Arabic and Romance, to the old Romance tunes’ (Liu and Monroe 1989, 8; Armistead and Monroe 1985, 207). These indigenous songs, as Reynolds explains, would not have contained the complex rhymes of the Arabic qasida. However, they may well have been strophic, with refrains, perhaps in the manner of Jewish and Christian responsorial liturgy.Footnote 13

Eventually, scholars postulate, these originally vernacular Christian songs would have been adapted by court poets to the language and poetic conventions of classical Arabic. In other words, the syncretism of these lyrics is evidenced by their application of complex, Arabic-influenced rhyme-codes to what is thought to have been an indigenous and vernacular Christian strophic structure.Footnote 14 In the Andalusí lyric, the first part of the strophe has a separate rhyme, while the second part of the strophe rhymes with the refrain, functioning as a signal and a transition to this all-important conclusion of the verse (Stern 1974, 15–16; see Fig. 2).

Figure 2
figure 2

Detail from the moaxaja ‘Ramal al-Máya,’ from the Kunnas al-Ha'ik (1789), the songbook by Muhammad al-Ha'ik al-Titwani al-Andalusi. Signatura: 398-POP-hai — Nº de registro: 3311, fol. 26. Centro de Documentación Musical de Andalucía. Copia digital: Biblioteca Virtual de Andalucía

Full size image

As described above, this rhyming cue, which scholars call the vuelta or return, maps onto the flamenco cambio, the melodic change signalling that the soloist has begun their descent toward the home tone and the verse’s ending. That is, in the Andalusí lyric, as in flamenco, the penultimate part of the verse contains a signal, like the flamenco pellizco, priming the listener/participants to anticipate taking part in the verse’s imminent conclusion. This strophic structure, Rip Cohen writes, is a ‘container of memory’ (2018, 44).Footnote 15

Games of allusion and the politics of pastoralism

The strophic poetry of the muwashshah [moaxaja] was invented with the purpose of having the kharja [xarcha] inserted into it.

Samuel Miklos Stern (1974, 225)

Aside from rhyme and strophic structure, there is one more important distinction between the lyric forms which arose in al-Andalus and the sung poetry that arrived with the Islamic cultural elite following the year 711: their language. And it is in the braiding together of multiple languages in these lyrics that we find the most direct evidence of their citation of the vernacular. The moaxaja, like the qasida, is in classical Arabic, the language of the court. But the moaxaja’s xarcha, its ending refrain, could be in a different—vernacular—language: colloquial Arabic or even Andalusí Romance (Liu and Monroe 1989, 6–7, 42; Corriente 2009, 110). And the zéjel, al-Tīfāshī recounts, combined the folksongs of Arab camel-drivers, in colloquial Arabic, with those of indigenous Christians (Liu and Monroe 1989, 6–7, 42). The meters and rhyme schemes of these Iberian hybrids function, remarkably, across multiple languages (Andalusí Romance, Andalusí Arabic, Hebrew, and classical Arabic) in a single song, and sometimes even in a single verse.Footnote 16

María Rosa Menocal views these lyric forms as becoming seminally important for Europe, comparing them to ‘Ibn Rushd’s (Averroes) commentaries on Aristotle,’ which ‘revolutionized (and radicalized) philosophical thought’ in the thirteenth century (Menocal 1993, 24–25). They ‘invent new Romance and Arabic and Hebrew poetics in one swoop, all in the same poem’; they are ‘poetry of and about the creation of a new language for poetry.’ Heated controversy, in terms of nationalist, identarian, and also academic politics surrounds the analysis of these poems. But there is a certain amount of consensus that: 1) the Andalusí lyrics are new, sui generis forms; 2) the Andalusí lyric is a mestizaje, a syncretism of the multiple languages, religions, and cultures of medieval Iberia; and 3) the Andalusí lyric is a pastoral, a citation and/or representation of the bilingual illiterate minstrel in the proverbial marketplace, singing Andalusí Arabic words to Mozarabic (Andalusí Christian) tunes (Armistead 2003, 3).Footnote 17 And this last point is where the politics get really interesting.

From the outset, transmission from vernacular to courtly culture would have logically been part and parcel of the Islamization of the Peninsula. Matthew P. Loar, Carolyn MacDonald, and Dan-el Padilla Peralta’s incisive analysis of The Dynamics of Cultural Appropriation holds true from Rome, to the medieval Islamic world, to the Americas: ‘literary practices ranging from plagiarism to quotation, and material practices ranging from spoliation to commercial import’ mediated the discourse of imperial domination (2017, 3–4). This mestizaje would have involved not only the folksongs of Arab camel drivers and Iberian Christians, as al-Tīfāshī writes, but African folk musics as well (Liu and Monroe 1989). Many of the Islamic forces invading in 711 were North African Amazigh (so-called ‘Berbers’). But further, and crucially, from the founding of the Umayyad Emirate in 756 Córdoba to the fall of Nazarí Granada in 1492, dynastic links to Africa were key to Islamic rule in Iberia. The Islamization of Africa, Amnon Shiloah observes, was ‘greatly aided by the Africanization of Islam’ (1995, 86).

In a 1912 lecture delivered on the occasion of his being inducted into the Real Academia Española, the eminent Arabist Julián Ribera y Tarragó (1858–1954) describes a blind Andalusí singer known as El Qabri (ca. 888–912), a court poet from Cabra (Murcia), as an early trace of the courtly adoption and reinterpretation of vernacular Andalusí song (1928, 56). El Qabri, according to the Treasury Concerning the Merits of the People of Iberia by Ibn Bassam of Santarém (1058–1147), used ‘non-Arabic’ phrases as the basis for a new kind of courtly poetry, the moaxaja. In 1948, Samuel Miklos Stern significantly advanced our understanding of these ‘non-Arabic’ phrases, discovering that some xarchas in Hebrew moaxajas, previously thought to have been ‘gibberish,’ were actually Hebrew transliterations of Andalusí Romance (Stern 1974, 123–60). That is, the Hebrew letters could only be read once it was understood that they were spelling words of a different language from the rest of the poem.

The poetics of the moaxaja require that the xarcha, Stern explains, be ‘written in some vernacular dialect…i.e. the everyday speech of the characters taking part.’ And we know from al-Mulk’s twelfth-century treatise on this poetry that the moaxaja’s xarcha not only closes the poem but, as Stern says in the epigraph above, actually determines its rhyme scheme and metric structure (Stern 1974, 225; Zwartjes 1997, 120–23; Monroe 1989). The poet must invent the ending of the moaxaja first, al-Mulk writes, ‘and from there he constructs the rest of the poem’ (Zwartjes 1997, 31–32; Beltrán 1984, 241–42; Stevens 1986, 191).Footnote 18 Even though it is placed at the end, al-Mulk writes, ‘it must be the beginning’—it is the ‘tail’ which the poet seizes to place at the head of the poem (Heller-Roazen 2002, 108; García Gómez 1962a).

The xarcha, Ribera’s student Emilio García Gómez adds, should explicitly be in the voice of another: ‘The xarcha, in whatever language, had to be in a direct style, and presented in the voice of someone introduced in the scene, and beginning with a verb (“he said,” “he sang,” “he recited”…)’ (García Gómez 1972, 3: 247). We learn from al-Tīfāshī’s account of Andalusí music that Avempace had ‘invented’ the zéjel as a courtly form by secluding himself for several years to work with ‘skilled singing girls’ to ‘improve’ and ‘combine’ Christian songs with those of the East (Liu and Monroe 1989, 42).Footnote 19 But many scholars think that the zéjel had existed as a vernacular form long before the twelfth century. Armistead and Monroe document ‘a surprisingly large number’ of xarchas which can be identified as being borrowed from earlier extant zéjels (Armistead and Monroe 1985, 207, 229; Corriente 2009, 118; Stern 1974, 54; Zwartjes 1997, 121–23; Monroe 1986, 240; Monroe 1989).Footnote 20 In the voice of an Other, the xarcha is, they write, ‘quotational in nature’—it employs contrafacta.

Melody and rhythm were intrinsic to this process of citation. Describing the ‘games of allusion’ that shape the zéjels of one of the masters of this lyric, Ibn Quzmán, García Gómez notes that the ending lines of Zéjel #71 (an ironic panegyric, written in the voice of a dissolute and aging poet)—‘que, si a un poeta | ves saludar, |*| algo hay ahí’ (‘when a poet offers his greetings, there’s usually a reason why’)—contain a reference that escaped notice for eight hundred years (García Gómez 1972, 3: 252–53, 1: 352–55; Quzmán and Monroe 2017, 428–31).Footnote 21 Because we find a variation on this concluding line in two other zéjels (#60 and #130), García Gómez reasons, we can deduce that ‘this ending phrase is a modification and allusion to the tune of a popular love song which everyone in twelfth-century Córdoba would know and could sing along with’ (García Gómez 1972, 1: 302–305, 2: 648–51; Quzmán and Monroe 2017, 364–69, 776–79). ‘By ending his poem with a quotation from the refrain of an earlier poem,’ Monroe writes, the ‘poet is indicating that his new creation is not only a metrical emulation, but also a musical contrafact,’ intended to be sung to the tune of the poem cited (Monroe 1987, 280).Footnote 22 The poems themselves thus contain evidence not only of linguistic but also of melodic citation. As I will briefly discuss, there is some evidence that this contrafacta included rhythm—perhaps hinting at the presence of dance—as well.

The moaxaja, with its ‘non-Arabic’ phrases in the words of an Other, ‘invented’ as a courtly form by El Qabri around the turn of the tenth century, presents the exceedingly thorny problem of reconciling meter, rhyme, and strophic structure across languages and systems. If we consider the primacy and specificity of rhyme as used in the Arabic poetic tradition that would have arrived in al-Andalus in 711, the problem of reconciling rhyme across languages—not to mention lyric traditions—would have meant that only rhymes that work exclusively in Arabic or in both Arabic and Andalusí Romance would have been legible to the Arabic-speaking elite, which was the intended audience for this courtly poetry. Rhymes that work only in Andalusí Romance would not have been intelligible in this context. Indeed, Zwartjes’s analysis of the bilingual xarcha texts concludes that most Romance xarchas are in fact ‘compatible with Arabic rules’; they have ‘been written in agreement with the Arabic system’ (1997, 178).

It would be a mistake, Zwartjes alerts us in his extended discussion of ‘code-switching,’ to interpret the xarchas’ citation of the vernacular as a straightforward ‘reflection of primitive Castilian popular lyricism’ (1997, 178, 276–87, 295–304). Neither, as evidenced by their strophic and rhyme structure, are they classical Arabic. As Federico Corriente disparagingly writes, the prosody of the Andalusí lyric could only have been the product of a desperate attempt ‘to imitate Eastern Arab culture’ by those who, ‘lacking the means to master Classical Arabic and the elaborate rules of its poetry, contented themselves with mimicking it in their own vulgar dialect’ (2009, 118). The modifications and manipulations required to accommodate the xarchas’ Romance language to classical Arabic prosody simply don’t work. They create ‘incoherence,’ García Gómez writes— a ‘pseudo’ or ‘false classical prosody’ (1972, 3: 33, 247–48). And this misappropriation is essential to the genre’s poetics. As al-Mulk writes, the xarcha is the pepper, salt, sugar, and amber of the moaxaja precisely because it is—and must be—written in a ‘debased, dialectal, wretched and gypsy-like’ style (Corriente 2009, 113).

Ibn Quzmán was a courtly poet and member of the intellectual elite, yet he writes in colloquial Hispano-Arabic rather than classical Arabic, and uses popular rather than classical meters, conventions, and themes (Monroe 1979, 78). El Qabri composed his moaxajas in tenth-century Córdoba, when Umayyad al-Andalus was at its imperial zenith. But Ibn Quzmán wrote two centuries later, in an era when political instability and flux in al-Andalus both inflamed religious hostilities and also accelerated cultural and intellectual exchange with Christian Europe. Thus, in the Quzmani zéjels we note a representational turn.

Quzmán’s ‘vulgarity,’ García Gómez explains, must be understood ‘cum mica salis’—with a grain of salt (1972, 2: 875–76). The Quzmani zéjels’ seeming failure to master classical prosody, he surmises, is ‘reflexive and conscious’—an ‘aesthetic if not stylistic’ choice. Ibn Quzmán’s ‘constitutive linguistic position’ is thus ‘ironic.’ That is, in the Quzmani zéjels, the ending xarcha’s ‘incoherence’—its very failure to capture classical prosody—is essential to the dramatic tension it builds within the poem (1972, 3: 247). Like the punchline of a joke, the xarchas’ meaning is apprehended comparatively, by reading it against its erstwhile classical models. As Monroe writes, the Quzmani zéjel’s construction of ‘conflict between form and content’ prompts the reader/listener to seek a deeper level of meaning (2017, 1382, 1386).

In their failure to conform to classical norms and, by extension, courtly culture, the Quzmani zéjels’ misappropriations implicitly critique not the vernacular culture from which they are drawn, but the hegemonic standards themselves. Systematically espousing ‘the opposite of noble sentiments, noble rhetoric, noble metaphors, and noble language for deliberate stylistic purposes,’ Monroe explains, Ibn Quzmán’s pastoralism actually declaims ‘some sort of unpanegyric,’ a satire of the panegyric itself, and of the elite world it represents (Monroe 1979, 120, 125–27; see also Monroe 1973, 91–93). That is, in Ibn Quzmán’s poetry, we find a member of the elite couching a critique of his own caste in terms of an Other, and this register of meaning is revealed through an apprehension of the poem’s ‘incoherence’ and ‘false’ or ‘pseudo-classical’ prosody (García Gómez, 1972, 3: 33, 247–48). The object of this poetry’s barbs (the Andalusí elite) is thus revealed in its very failure to reconcile the clashing ‘constitutive linguistic positions’ of those whose encounters it depicts (García Gómez, 1972, 3: 33). Ibn Quzmán builds an ‘inverted poetics,’ Monroe writes, in order to ‘criticize the social and literary practices of his age’ (Quzmán and Monroe 2017, 1382).

The ironic subtext of these syncretic, exquisitely detailed, and difficult poems, perhaps even the critique implicit in their pastoralism, is inherent to the word zéjel itself. As Corriente explains, the xarcha’s unique linguistic register ‘obviously served no other purpose than what is meant by the very word’: ‘to make an ‘outing’ from the rigid literary and linguistic conventions of classical poetry’ and ‘towards the lighter and livelier style of folk poetry’ (2009, 110). This type of poetry is hazl, meaning ‘facetiousness.’ The word hazali, Corriente adds, related to hazl, is synonymous with zéjel for Ibn Quzmán. That is, one gloss for the word zéjel is a facetious invocation of folk style.

Appropriation and misappropriation across the hegemonic divide

The single most striking characteristic of refrains [is] their cross-contextuality.

Ardis Butterfield (1991, 4)

As the pellizco illustrates, both the vocal and gestural citation of past interpreters is essential to flamenco’s construction of its own identity. This is a millennial practice: in medieval Europe, poetic citation or contrafacta facilitated transmission between vernacular and elite cultural spheres as between Muslim and Christian courts. In Poesía juglaresca y juglares, Ramón Menéndez Pidal illustrates this, explaining that itinerant juglares (street minstrels) contributed to the standardization of national dialects by translating songs from one language into another as they travelled (1924, 11, 142). According to Cohen, the leading authority on the cantigas d’amigo, this Gallego-Portugués lyric is the product of a similar process: ‘these verses may be read as an allegory for what happens to folksong when it dies in the arms of writing’ (2012, 637). The cantigas d’amigo are a ‘tradition of female voiced love‑lyric,’ Cohen continues, first written down in the 1190s and 1200s, which, Cohen explains, migrated by the 1240s from the courts of ‘aristocrats and magnates in Galicia and the North of Portugal to the court of Alfonso X’ (2012, 639).

Alfonso X, known as El Sabio or ‘the Wise,’ inherited an aggrandized Castilian kingdom in the wake of his father Fernando III’s decisive conquest of Almohad Sevilla in 1248. Under Alfonso’s rule, four magnificent codices of Cantigas de Santa María (CSM), the earliest extant notations of song and music for medieval Iberia, were compiled.Footnote 23 Recording how widespread and popular this lyric form was in the Iberian South, the Andalusí zéjel, Ferreira writes, was the ‘musical schemata for the vast majority’ of the CSM (2016, 296, 299).

The zejelesque schemata of strophe+estribillo is recorded in the 1789 Kunnas al-Ha’ik, the first songbook to record the Andalusí repertoire conserved in diaspora among Tetuani families whose ancestors were exiled Iberian Muslims (al-Ha'ik 2003, 13).Footnote 24 In the detail from the moaxaja ‘Ramal al-Máya,’ the ‘girdle’ shape of the verse surrounded by two estribillos is clearly discernible, along with their rhymes. The endings of the first and last two lines (read right to left) are identical. These are the estribillos bracketing the strophe, the three lines in the centre, whose ending rhyme is distinct (see Fig. 2).

The same structure of strophe bracketed by two refrains is visible in the detail below of stanza XI of CSM 419, a zéjel titled, in translation, ‘The Vigil of the Feast of the Assumption (14 August).’ The estribillo’s incipit, ‘Des quando Déus sa Madre | aos céos levou’ (Of when God took his Mother to heaven), in red, brackets this as it does every verseFootnote 25 (Fig. 3).

Figure 3
figure 3

Stanza XI of CSM 419 Cantigas de Santa María, Códice de los músicos Patrimonio Nacional. Real Biblioteca del Monasterio de El Escorial, B-1-2_009v, https://rbme.patrimonionacional.es/s/rbme/item/11338#?c=&m=&s=&cv=29&xywh=-3119%2C-1%2C9980%2C5616

Full size image

The adoption of the Andalusí zéjel as a model, Ferreira observes, was ‘based on previous experiments in satirical song by Alfonso and his circle of troubadours’ (2016, 296). Like the Andalusí poets before them, the erudite poets at Alfonso’s court strived to artfully incorporate contrafacta, whether of learned poets or of the popular songbook, into their work (Filgueira Valverde 1975, 435). This technique was called ‘fillar’ (to affiliate oneself), ‘seguir’ (to follow, to trail), or ‘seguida’ (pursuit)—evoking the early modern dance-songs called seguidillas (Clotelle Clarke 1944). Lexicographer Gonzalo Correas described the seguidillas as being ‘very old’ in 1626; they are a popular vernacular form whose origins, he suggests, may be found in their ‘cabezas de cantares’—their estribillos, or refrains (Correas 1903, 272–73).

The ‘experiments in satirical song’ at Alfonso’s court were often citations of the vernacular: villancicos, or songs of villanos (villagers), composed using the mudanza+vuelta+estribillo structures of the Andalusí lyric. In fact, aside from the Andalusí xarchas, the earliest documented estribillo in an Ibero-Romance dialect is a villancico, first recorded in the Chronicon Mundi by Lucas de Tuy (1236), written for Alfonso’s grandmother, Berenguela de Castilla (Monroe 1988–89, 25). And the term estribillo (little stirrup) is itself thought to derive from the Arabic márkaz (stirrup), said to have been coined by El Qabri in the early-tenth century (Monroe 1987, 296; Pope 2001). This term designates the vuelta, the penultimate part of the verse that, like the flamenco cambio or ‘B’ line, prompts the audience to participate in the verse’s ending.

In addition to villagers’ songs, the poets at Alfonso X’s court also contrafacted songs from elite Andalusí circles. Lopo Lías’s ‘son de negrada’ is literally a ‘song of blackening,’ but Ferreira, noting its racialization of religious difference, translates the title as ‘song of infidels’—that is, of Muslims and Jews (2002, 109).Footnote 26 This cantiga, Ferreira writes, proves the lines of contact, even at this relatively late-reconquest date, between Christian troubadours and Muslim courtly culture—it is thought to be a seguida of a song by Ibn Quzmán himself (2002, 108–109; Filgueira 1975, 442).

The cantigas, Ferreira explains, were ‘linked to a program of religious and dynastic propaganda,’ promoting an imperial Christian restoration in the Muslim South (2016, 296). Muslims are represented in ten percent of the cantigas, while seven percent of the CSM deal with Jews, Albert Bagby explains, who are depicted as the ‘archenemy of Christianity,’ as the ‘devil's disciples, usurers, and traitors’ (1970, 164; 1971, 675, 687). CSM 419 thus begins: ‘E com' éu hei oído | estes maos judéus, que mataron méu Fillo | como falsos encréus’ (As I have heard these Jewish hands who as false believers killed my Son …).Footnote 27

Consider the irony: in furtherance of their colonizing project, the erudite poets at Alfonso X’s court satirized Andalusí song forms, using the Andalusí zéjel, itself a citation of the vernacular—of the Christian vernacular—as their model. As Benjamin Liu writes of the ‘experiments in satirical song’ at Alfonso’s court, the Cantigas d’Escarnho e de Mal Dizer (CEM), these texts situate themselves within a literary code as a ‘counter-text,’ a deliberate juxtaposition that employs the code’s procedures but turns the poem away from its referential content (2004, 11).

In the Andalusí xarchas, Christians are represented, in the words of al-Mulk, as ‘debased, dialectal, wretched and gypsy-like in style’; as Corriente writes, they mimic the elaborate rules of Classical Arabic poetry poorly (2009, 113, 118). And in the flux of twelfth-century Almoravid al-Andalus, Ibn Quzmán purposefully misappropriates courtly culture—in the voice of vernacular oral tradition—in order to critique it (Monroe 1973, 96). In the Alfonsine cantigas, this relation of power is reversed; it is the moros (‘Moors,’ referring to Muslims) who, along with the Jews and ‘pagans,’ are subaltern. Reflecting the tipping of hegemonic polarity following Fernando III’s 1248 conquest of Sevilla, the imperial Christian program of these lyrics is mirrored in their appropriation of the Andalusí zéjel as a ‘musical schemata.’ That is, the zéjel rhetorically structures this crucial reversal in representation—it provides the language, if you will, for this parody of itself.

Reflecting the flip in both hegemony and representation, the Andalusí zéjel, a courtly Muslim appropriation of the Christian vernacular, is preserved within the Spanish songbook as a pastoral, as a villancico. Marisa Galvez traces this forward in time, into Juan Ruiz’s Libro de buen amor (Book of Good Love) (1330–1343) (Galvez 2011, 39–40). For Ruiz, as for later Spanish poets (recalling the 1236 Chronicon Mundi), the term villancico designates an estribillo (Pope and Laird 2001, 1). Thus, the zéjelesque ‘musical schemata’ of mudanza+vuelta+estribillo structures the villancico cortés, the courtly pastorals of the fourteenth and fifteenth centuries (Beltrán 1984, 258). (One beautiful example is the well-known ‘Tres morillas m’enamoran,’ in the ca. 1500 Cancionero Musical de Palacio).Footnote 28 Even the sixteenth- and seventeenth-century villancicos de negro, cultivated in the service of Spain’s ruinous program of enslavement and colonization in the Americas, retain this schema. (See, for example, Gaspar Fernandes’s ‘Guineo a 5: Eso Rigo e Repente’ [Puebla, México, 1609–1616], in which the singing Others are now enslaved Africans.)Footnote 29

And the word zéjel, from the Arabic hazl (a facetious pastoral), is retained in Spanish, with a similar meaning. Thus, the earliest Spanish-Arabic dictionary, the Vocabulista arauigo en letra castellana, published in 1505 by Pedro de Alcalá, defines ‘romance cantar’ (song in Romance) as ‘zéjel’ (Alcalá 1505, 458; Zayas 1995, 59). And this gloss carries into the first monolingual dictionary of Castilian Spanish, Sebastián de Covarrubias’s Tesoro de la lengua castellana o española (1611), which defines ‘çagal,’ an alternate transliteration for ‘zagal,’ ‘zajal,’ or ‘zéjel,’ as an ‘Arabic word’ for brave, daring, and strapping village youths and maidens (Covarrubias 1611, 261).

Thus far, I have been tracking the transmission of strophic structure and also of facetious pastoralism across the boundary separating Muslim from Christian hegemony, and pre- from early modernity. The Iberian poetics valuing learned citation or contrafacta, along with listener/participants’ efforts to understand these references, also seem to have crossed the hegemonic membrane. As Galvez perceptively demonstrates, the ‘jingle and tune’ of the CSM’s devotional songs sung to familiar Andalusí refrains served a didactic, and indeed evangelical purpose (2011, 50). Now bent to a different dominant class, the CSM’s verse+refrain structures would not only have prompted the audience to sing along but, further, might have invoked the Andalusí practice of critical and comparative listening. The estribillo became a didactic device, Galvez argues, employed broadly across medieval European songbooks (2011, 38–52). The ‘tune and jingle’ of the repeated refrain produces ‘echoing effects…requiring stanza-by-stanza reflection’; each recapitulation can thus tap new layers of meaning (2011, 47).Footnote 30

This poetry values ‘the secret rather than the explicit,’ Galvez writes; it cultivates ‘the pleasure of intertextuality’ (2011, 196–97). As what Cohen terms a ‘container of memory’ (2018, 44), the estribillo can thus be understood to transmit, beyond mere strophic structure, the cultivation, detailed in Avempace’s Risala, of listening to and comprehending lyric poetry as an intellectual, spiritual, and we might even say moral practice. As we have seen, these values are deeply embodied in the flamenco episteme.

A body of knowledge: toward a gestural methodology

The truth is that in the air of Córdoba and Granada one still finds gestures and lines of remote Arabia, and remembrances of lost cities still arise from the murky palimpsest of the Albaicín.

Federico García Lorca (1998, 18)

Like the flamenco estribillo, the Andalusí xarcha, as mentioned above, is not technically a refrain; its words and melody not repeated exactly after every verse. Even though the ending xarcha recapitulates the ‘rhyme-sound’ of the refrain that has been heard throughout the song, the xarcha, Daniel Heller-Roazen explains, is written in a language ‘of which there has until then been no sign in the work’ (2002, 106–107). This is why the Romance xarchas, though published and studied for centuries, were unintelligible until Stern’s 1948 discovery.

The third volume of García Gómez’s magisterial Todo Ben Quzmán deals with meter, and it includes a section on Ibn Quzmán’s use of the xarcha (1972, 3, 227–304). The moaxajas are ‘quite strange,’ he writes (1972, 3, 247). They are short—just five to seven strophes—and within this truncated structure they must accommodate the conventions of the classical Arabic qasida, including an ‘erotic, bacchic, or descriptive prologue,’ and the body of the poem itself, along with the introduction of a foreign element, the xarcha, and finally the xarcha itself. García Gómez poses the question: ‘why is the xarcha there?’ (1972, 3, 247). In the courtly Andalusí context, he concludes, the ‘essential reason’ for the xarcha would have been as ‘an indication of the rhythm’ (1972, 3: 254).

According to the conventions of classical Arabic poetry, García Gómez continues, the poet might end by lauding his own artistry, inserting literary and technical details to prove his point—including a ‘determination and specification of the rhythm’ (1972, 3: 254–55). An indication of rhythm was necessary in the courtly zéjel precisely because, as we have seen, this ‘exquisitely detailed’ poetry did not work at all with classical Arabic meter (1972, 3: 256). Eventually, the quotation of the xarcha (as in Ibn Quzmán’s zéjel #71: ‘when a poet offers his greetings…’) comes to name the rhythm itself. That is, the xarcha may be only implicit. Nevertheless, even if the xarcha is missing, García Gómez concludes, ‘it leaves the reminder of its function’ (1972, 3: 256). ‘The xarcha simultaneously exists and evaporates.’ ‘The simple allusion to the xarcha (and no longer the xarcha itself)’—as we saw in the flamenco verse, finished not by the singer but by the group—fills in the ending of the poem.

The practice of critical and comparative listening, making the effort to understand, which we have seen crosses the boundary separating Muslim from Christian hegemony, is thus deeply embodied in the very rhythms of the song. Multilingual Andalusí audiences, Dorothy Clotelle Clark notes, would have found ‘a certain piquancy in the artistry of the rhythmic enmeshing and the euphonic blending’ in mixed-language xarchas (1988, 55). But for audiences who did not understand, she adds, who heard only what Stern calls ‘gibberish’ in the poem’s sudden revelation of an Other language, the xarcha’s stress-beat patterns could ‘fill in’ the gaps ‘otherwise left vacant save for meaningless sound’ (Clotelle Clark 1988, 55). The stress-beat patterns of the xarcha would have invited a ‘choral’ audience response of ‘handclapping,’ ‘foot-tapping or stomping, or even dancing.’ Flamenco invites this response too, as I have described here. Further, flamenco’s rhythmic/strophic codes, embodied in the pellizco, describe its deepest and most important ways of knowing and remembering.

The concluding xarcha, Heller-Roazen beautifully writes, is the poem’s ‘essential aim, that from which it emerges and to which it continuously moves’ (2002, 108). As in flamenco today, it is the event for which the poem’s ‘entire rhetorical, metrical, and semantic architecture’ has prepared its audience.Footnote 31 It is this concluding verse which suddenly reveals that we have been rhyming all along with the voice, and language, of an Other. Like the flamenco pellizco, the xarcha is a shock of sensation and emotion—it tastes, as al-Mulk writes, of pepper, salt, sugar, musk, and amber.Footnote 32 In the xarcha, the poem’s language is dissolved into something unknown, Heller-Roazen writes, and the poet is transformed into a ‘barbarian…speaking in tongues’ (2002, 108). It is a dislocation of identity that, in its incoherence, transcends and critiques the codes to which it fails to conform. But, more than that, it is a shock that suddenly reveals new dimensions of meaning.

I have argued in Sonidos Negros: On the Blackness of Flamenco that this dislocation of identity haunts Spain to this day (Goldberg 2019). As in the representational turn of the thirteenth-century Alfonsine cantigas, with Peninsular conquest completed and eyeing an empire upon which ‘the sun never set,’ Catholic imperialism sought to both subjugate and appropriate the cultural legacy of al-Andalus. The Spanish Church-Crown apparatus would found its modern European and Christian identity, its whiteness, on its pureza de sangre, the absolute purity of its religious and soon-to-be-racialized lineage. But the unspoken secret of Spain’s identity (and of course that of the West more broadly) is that this was a fiction, a failed project. Neither had Iberian Jews and Muslims been eliminated, nor had the Andalusí cultural legacy been extirpated. On the contrary, Catholic Iberia was manifestly impregnated with Islamic Iberia, as evidenced by, as we have seen, the vernacular’s many Arabisms (‘embaraços,’ meaning ‘impediment’ or ‘encumbrance,’ but also ‘pregnancy’…)—such as the word ‘zéjel’.Footnote 33

Long before its emergence as a named genre in the mid-nineteenth century, the Jewish, Muslim, Afro-descended, and Roma people who made flamenco were despised, in al-Mulk’s words, as ‘debased, dialectal, wretched and gypsy-like’ (Corriente 2009, 113).Footnote 34 Flamenco inherits the Blackness of the reversal in hegemonic polarity following Christian conquest—the keystone of the West’s foundational racism. And flamenco’s performance of this Blackness derives from the performance of the villano, the villager (Goldberg 2019, 31–49). As Ibn Quzmán’s legacy mouldered into the Iberian cultural substrate, it is precisely the villano’s indeterminacy, its instability and off-centredness, its scepticism of its own surface and of what García Gómez terms its ‘constitutive linguistic position’ (1972, 3, 33), that allows flamenco, as is true of so many Africanist forms, to open ‘a space for enacting the most profound imperatives of resistance to slavery and genocide’ (Goldberg 2019, 9).

Perhaps it is here, in the practice of making the ‘effort to understand’ not words but glossolalia, ‘not forms but the marrow of forms,’ not what is seen but what is felt, that we find the pellizco—as a shock of transgenerational memory, an embodied synapse connecting Avempace and Lorca, ‘now’ and ‘then,’ the incomprehensible ecstasies of God-intoxicated Sufis and the rapturous poetry of San Juan de la Cruz (1542–1591) (Cortés García 1996, 15; Lorca 1998, 61–62; Schneider 2011, 123). Perhaps, ‘precisely and less paradoxically than it seems,’ it is because, as Barthes writes about the ‘Grain of the Voice,’ flamenco ‘was already marginal, mandarin, that it was able to bear traces of signifiance, to escape the tyranny of meaning’ (1988, 185). ‘Isn’t it the truth of the voice to be hallucinated?’ Barthes asks. ‘Isn’t the entire space of the voice an infinite one?’ (1988, 184). Despite flamenco’s ironic duplicity and facetious performance of the racialized national self, its tormenting spirit, its inner life—its duende—remains. A small gesture like the pellizco vibrates within the body as a soundbox, where the sonic arrows of what flamencos call ‘intention’ resonate across time and memory.