Girl in the dawn: textual criticism and poetics.

Abstract. D. Dinis' celebrated cantiga d'amigo Levantou-s' a velida presents problems for both textual critic and interpreter, but these turn out to be linked, so that the resolution of syntactical and metrical problems provides a sound basis for the interpretation of symbolic logic. The girl gets angry at the wind, which would seem to represent the boy, but this reading conflicts with the joy that usually results from the meeting of boy and girl near fresh water. Two acrostics, one in Latin and one in Hebrew, lead to a startling conclusion: that this is no ordinary girl, but the Virgin Mary, and that her 'friend' is the Holy Spirit. Both Mary's presence eno alto (at the river) and her less than enthusiastic reception of the Holy Spirit are related to traditions in the pseudo-gospels and medieval art. The song thus provides a daring portrayal of the annunciation, and is placed strategically at the centre of an organized sequence of cantigas d'amigo.

Keywords. Galician-Portuguese lyric; King Dinis of Portugal; textual criticism; symbolism; organized sequences of cantigas d'amigo.

Resumo. A celebre cantiga d'amigo Levantou-s' a velida de D. Dinis suscita problemas tanto para o critico textual como para o interprete, mas estes problemas estao interligados, de modo que a resolucao dos problemas sintacticos e metricos oferece uma base solida para a interpretacao da logica simbolica. A amiga zanga-se com o vento, o qual parece representar o amigo, mas esta leitura nao se conforma com a alegria que normalmente caracteriza o encontro de amigo e amiga ao pe de agua doce. Dois acrosticos, um em latim e o outro em hebraico levam a uma conclusao surpreendente: que esta amiga e a Virgem Maria e o amigo e o Espirito Santo. A presenca de Maria eno alto (ao pe do rio) e a sua atitude para como o Espirito Santo relacionam-se com tradicoes dos evangelhos apocrifos e com a arte medieval. Assim sendo, a cantiga oferece um retrato um tanto audaz da Anunciacao, e situa-se estrategicamente no centro de uma sequencia organizada de cantigas d'amigo.

Palavras-Chave. lirica galego-portuguesa; D. Dinis; critica textual; simbolismo; sequencias organizadas de cantigas d'amigo.

**********

Anyone who edits medieval poetry must transcribe and collate manuscripts, divide words, correct letters, supply missing syllables and phrases, and transpose words, verses and strophes. During the years I spent preparing a critical edition of the 500 cantigas d'amigo I corrected the text in many places and did my best to supply what was missing. Another challenge was to analyse metre and strophic form. After studying the problem in all its detail, I concluded that, contrary to standard opinion, the cantigas d'amigo are metrically regular, and that the strophic form of many poems had been wrongly analysed. I found that correcting errors in morphology and syntax often led to metrically perfect verses where in the manuscripts there is neither metre nor grammar. Sometimes the matter ends there, but at other times there are repercussions.

There are about a thousand philological notes in the edition, on topics from etymology to syntax, but questions of interpretation are not discussed unless they are relevant to a textual problem. Now I am preparing a commentary on all 500 texts. In doing so, I have not doffed the cap of textual critic to don that of interpreter. One can hardly edit texts without understanding them. (1) To show how tightly textual criticism and interpretation can be interwoven, I shall examine a controversial cantiga (see Appendix) of King Dinis of Portugal (1261-1325).

   Levantou s' a velida
   Levantou s' <aa> alva
   E vai lavar camisas
   Eno alto
   Vai las lavar <a> alva.       5

   Levantou s' a loucana
   Levantou s' <aa> alva
   E vai lavar delgadas
   Eno alto
   Vai las lavar <a alva.>      10

   <E> vai lavar camisas
   Levantou s' <aa> alva
   O vento lhas desvia
   Eno alto
   Vai las lavar <a> alva.      15

   E vai lavar delgadas
   Levantou s' <aa> alva
   O vento lhas levava
   Eno alto
   Vai las lavar <a alva.>      20

   O vento lhas desvia
   Levantou s' <aa> alva
   Meteu s' <a> alva en ira
   Eno alto
   Vai las lavar <a alva.>      25

   O vento lhas levava
   Levantou s' <aa> alva
   Meteu s' <a> alva en sanha
   Eno alto
   Vai las lavar <a> alva.      30

If we did not know who the author of this poem was, we might at first glance think it rather close to folksong. But D. Dinis was no ordinary poet, nor even an ordinary king. As son of the French-educated Afonso III, grandson of Alfonso X of Castile and Leon, and husband of the great-grand-daughter of Federico II of Sicily, Dinis brought together in his person, by birth or marriage, the cultural inheritance of the most advanced courts in Europe. And European culture was astir: Greek texts had reached Spain in Arabic translations; treatises were being written on medicine and music, mathematics and grammar; Arabic, Hebrew and Romance poetry were thriving, and so were Christian, Islamic and Jewish theology and mysticism. From architecture to philosophy, a renaissance was in the air. It is in such a world that we should imagine these words being sung and heard, written and read.

Scholars agree that Dinis modelled his song on one by Pero Meogo, a Galician jograr. (2) In Meogo's cantiga, a girl gets up, goes to wash her hair in the fountain, her boyfriend comes along and, at the same time, a stag stirs the water (see Appendix). (3) In Dinis' poem, it is her clothes that the girl washes at dawn, and the wind takes them away. This suggests a symbolic equivalence between the stag stirring the water and the wind bearing off the camisas. But Meogo's girl is happy, while the other is not.

Stephen Reckert and Helder Macedo compare the songs in detail, basing their interpretations of Dinis 17 partly on the assumption that alva is ambiguous throughout and refers to both dawn and girl. (4) Their text is mainly faithful to the reading of both codices. (5) But manuscripts can be wrong. An analysis of syntax, metre and strophic form led me to restore <aa> alva 'in the dawn' and <a> alva 'the white one', meaning the girl. (6) These supplements clarify what had appeared a series of ambiguities. The strophic form (aBaCB) is unique in any event, but the use of four different verse-lengths would have made it strange indeed. (7) Still, that would have been acceptable, if there had been external responsion, that is, if each verse had corresponded metrically to the verses in the same position in the other strophes.8 The lack of both external responsion and syntax signals an error. We need a preposition and an article for alva meaning 'dawn' and an article for alva referring to the girl. Restoring those words for the sake of syntax adds syllables which not coincidentally yield a credible scansion for the verses (6' 6' 6' 3' 6'), and this in turn gives us a strophic form that conforms to standard practice.

These minute emendations had several repercussions. First, the restoration of <aa> alva in Dinis 17 led me to conjecture that the missing half-verse in the opening of Meogo 5 should be <Levou s' aa alva>. (9) This would then be the direct source of Dinis' verse levantou s' <aa> alva. Dinis would have substituted levantar, a later verb, for levar, split Meogo's long verse more or less in half (6' in Dinis against 11' in Meogo), and inverted the order of the phrases. If my conjectural supplement in the first verse of Meogo 5 is correct, we have one more example of the Galician-Portuguese dawn-song, representing a meeting between lovers at dawn (not to be confused with the Provencal or French alba, where lovers part at daybreak). (10)

Another repercussion involves the symbolism in Dinis 17, which has been the subject of dispute. Once we have distinguished the girl from the dawn, alva from alva, we can proceed from syntax to symbolic logic on the basis of a sound text.

Eugenio Asensio (1970: 66) wrote: 'La lozana que lava camisas--rito que simboliza una magica intimidad con el amado--pierde su misterio y se amohina sanuda por nada, por el viento'. Reckert and Macedo (1996: 67, 226-27, following Asensio 1970: 198-99) call attention to similarities between Dinis 17 and the Nausicaa episode in Homer's Odyssey. Athena takes the form of one of Nausicaa's friends and appears to her in a dream, saying it is time for her to be married. Here we find a striking parallel: 'Let's go wash (your clothes) at break of day', Athena tells her. (11) So the nubile girl who goes at sunrise to wash her clothes appears at the beginning of European poetry.

But what about the wind? The association between wind and sexuality occurs in archaic Greek lyric, in Sappho:

   Love shakes up my heart
   Like a mountain wind smashing into oaks. (12)

There is another lusty breeze in an epigram from the Hellenistic period:

   I wish I were the wind and you, walking along the shore,
   Would bare your breasts and take me as I blew. (13)

And scholars have cited Hispanic parallels such as these: (14)

   Levantose un viento
   Que de la mar salia
   Y alcome las faldas
   De la mi camisa

   Un mal ventecillo,
   Loquillo con mis faldas:
   Tira alla, mal viento!
   Que me las alcas!

Reckert and Macedo say that in Dinis 17 the wind represents the boy and that the girl is angry at what he does. This is disputed by Rosario Ferreira (1999: 129-37; 2002: 160-70), who sees no reason why the girl should be upset if her friend is there, and thinks he has failed to appear and that is why she gets angry. Reckert takes the washing of clothes as a prenuptial rite and cites many parallels, including this Sephardic cantar de boda: (15)

   Debajo del limon
   dormia la nina
   y sus pies en el agua fria,
   Su amor por ahi vendria:
   --'Que hases, mi novia garrida?'
   --'Asperando a vos, mi vida:
   lavando vuestra camisa ...'

Yet here, as in similar scenes in the cantigas d'amigo where girl meets boy near fresh water, she is happy, as in Meogo, not upset, as in Dinis, so Ferreira's puzzlement is valid. Assuming the friend is symbolically present, why does the girl get angry?

In the cantigas the causes of female fury vary. (16) The boy's failure to keep his word is a common cause, but not the only one. One girl is annoyed because her boyfriend loves her and does what she says (Calheiros 5, vv. 1-4). Another threatens to get angry if her boy won't tell her exactly what he wants: assanhar m' ei un dia / se m' el non diz qual ben de min querria (Servando 12, fiinda). So why should this girl not show anger at the male wind for taking the blossom of her youth, even if at another level she has come to wash her clothes precisely to that end?

We could argue that she would be conforming to the literary construct of a girl eager to make love who nevertheless says 'No'.17 Ovid describes what happened when Corinna came to see him one afternoon (Amores 1.6.13-16):

   Deripui tunicam; nec multum rara nocebat,
   pugnabat tunica sed tamen illa tegi;
   quae, cum pugnaret tamquam quae uincere nollet,
   uicta est non aegre proditione sua.

   I took off the gown; it was thin and didn't really matter,
   But still she kept struggling to keep it on.
   But since she struggled like a girl who doesn't want to win
   She was easily undone by her own betrayal. (18)

According to this logic, the girl in Dinis 17 gets angry when the boy-wind takes away her clothes. But the symbolic puzzle may not be so easily explained.

Some scholars (such as Macedo and Mercedes Brea) think the song describes a violent sexual initiation. (19) But in the cantigas d'amigo, where there is not a single kiss, and where eroticism is alluded to by expressions like falar and fazer ben or comes encoded in images that portray the girl as willing, it is unlikely that any poet would refer to sexual violence. (20)

This riddling wind may be part of a bigger mystery. Before trying to solve that enigma, I would like to mention something that has escaped notice: Dinis 17, like Meogo 5, falls in the centre of a set of songs.

Whether there are organized sequences of cantigas d'amigo is a question that has been simmering for a century. (21) If they exist, they would be the earliest sequences of love songs in medieval Romance literature. One set that scholars agree is a sequence are the nine cantigas of Pero Meogo, and there <Levou s' aa alva,> levou s' a velida comes fifth, in the centre. And if Dinis composed a set of 32 songs, as I have argued, (22) the midway point falls between nos. 16 and 17, so Levantou s' a velida is the second half of the centre. In both Meogo and Dinis, dawn marks the middle.

These two songs are distinguished in another way too. In this genre, the speaker is normally the girl, her mother, the girl's girl friend or the boy. In 500 cantigas d' amigo there are only a dozen exceptions--poems that begin with a narrative voice outside the action--and Meogo 5 and Dinis 17 are the only ones where no girl speaks.

Formally, Levantou s' a velida is marked in three ways: its strophic form is unique, it is the longest song in Dinis' set of 32, (23) and the only one with an intercalated refrain containing a repeated rhyme-word. (24) So numerically, pragmatically, formally and symbolically this song is suitably marked for a centrepiece.

The larger enigma I detect in Dinis 17 begins with two acrostics: reading the initial letters of the strophes we get LL<E>EOO, that is, LEO, the lion being an ancient symbol of kingship; and in the two central strophes, reading the initial letters of the verses, we find: ELOEnV, which means 'Our God' in Hebrew. (25) As a written performance inaccessible to an audience, acrostics are extremely rare in Galician-Portuguese lyric. Until recently only two had been recognized, both in the Cantigas de Santa Maria and both spelling MARIA (CSM 70 and 410). Would Dinis have known the word Elohenu? Alfonso X uses Adony (Adonay, 'My Lord' CSM 270.35), so to find Elohenu in the work of his grandson should not be so surprising. By the thirteenth century, Christian scholars knew some of the names of God in Hebrew. (26) And the combination Adonay Elohenu occurs regularly in blessings, as well as in the central prayer of Judaism, the Shemah. (In all such cases, it should be remembered, Adonay is pronounced instead of the Tetragrammaton, which is what is actually in the text.)

Appearing in the middle of the central poem of a sequence, inside the acrostic LEO, Elohenu would reflect Christian use of Jewish traditions in thirteenth century Iberia. Johan Garcia de Guilhade has an acrostic with the word ZION, (27) the name of the ancient acropolis of Jerusalem, and a key term in post-exilic--messianic--theology.

If the acrostics in Dinis 17 are not a mirage, the impregnating wind that Reckert and Macedo see may be the Holy Spirit. Hebrew ruah, forerunner of the Christian spiritus sanctus, means 'wind' and 'spirit', including 'holy spirit'. (28) And if that is right, LEO may refer to the kingship not of Dinis but of Jesus, since 'Judah is a lion's whelp', as we know from Genesis (49. 9), and the lion in Ezekiel's vision (Ezek. 19) was understood to be Jesus, 'the anointed one whom the Most High has kept until the end of days, who will arise from the seed of David' (4 Ezra [= 2 Esdras] 12. 32).

The sequence of Johan de Guilhade consists of 22 poems, and I have argued that this corresponds to the number of letters in the Hebrew alphabet. Given that, we might look again at the number 32, the number of songs in Dinis' sequence and, strangely enough, the number of syllables in each strophe of Levantou s' a velida. At the beginning of the Sepher Yetzirah (or Book of Creation), a key source for medieval Kabbalism, we learn that God 'engraved His Name in 32 paths'. So 32 had profound significance in Jewish mysticism, representing the sum of 22 and 10--where 22 stands for the number of letters in the alphabet, and 10 the integers from 1 to 10, as well as the ten sefirot of the Kabbalah. It is possible, then, that Dinis 17 contains religious symbolism and is the centrepiece of a sequence based on a mystical numerology that Christianity borrowed from Judaism. (29)

But how can Mary balk at the Annunciation? And what on earth is she doing washing her clothes in the river? There seem to be two interrelated traditions here. One is that in some of the pseudo-gospels that were popular in the Middle Ages, Mary receives the Annunciation when she goes to draw water from a fountain. She gets frightened ([TEXT NOT REPRODUCIBLE IN ASCII.]) and runs home. (30) The other, found in medieval art, portrays the Virgin as hesitant, if not downright displeased. Perhaps the most famous example is the Annunciation of Simone Martini and Lipo Memmi, painted in 1333 (Figure 1). There Mary is drawing her

body back, her fingers on the collar of her cape are pulling the garment shut, as if to protect herself, and her head is tilted at an angle, as though against an oncoming gust. From the angel, a wind seems to emanate. The Virgin appears half blown away by the news, not at all happy.

[FIGURE 1 OMITTED]

If it is this attitude of the Virgin that explains the girl's sanha in Dinis 17, the poem would be laden with much more than erotic symbolism. It would be a deft and daring portrayal of the Annunciation, something nobody would have expected to find in a cantiga d'amigo. (31) But in all Galician-Portuguese lyric the only being who is ever called alva is Santa Maria. At the beginning of six successive strophes of a Cantiga de Santa Maria, Alfonso X says to the Virgin: Tu es alva, and all seven end with alva (either as adjective or as noun or both). (32) Here is the second strophe:

   Tu es alva dos alvores,
   que faze-los peccadores
   que vejan os seus errores
   e connoscan sa folia,
   que desvia
   d'aver om' o que devia,
   que perdeu por sa loucura
   Eva, que tu, Virgen pura,
   cobraste, porque es alva.

In this light, it seems plausible that a alva in Dinis 17 symbolizes the Virgen pura, that 'woman clothed with the sun' mentioned in Revelations (Rev. 12). If so, this would be an indirect allusion to the Cantigas de Santa Maria and a stunning centrepiece in a sequence of 32 love songs. If I am correct to any appreciable degree, the poem gains in symbolic and referential scope by stepping clear of its fabled ambiguities.

And the first step to solving the enigma was the restoration, for syntactic and metrical reasons, of a few a's. So textual criticism is not a craft practised on the margins of reading; it is smack in the middle, directly linked not just to metre, grammar and pragmatics, but to larger questions of interpretation, from erotic symbolism to the history of religions--which, in this song, seem not to be very far apart.

Appendix

B = Biblioteca Nacional, Lisbon, cod. 10991
V = Biblioteca Apostolica Vaticana, cod. lat. 4803

Dinis 17

Levantou s' a velida              The lovely girl arose
Levantou s' <aa> alva             She arose at dawn
E vai lavar camisas               And goes to wash her clothes
Eno alto                          In the stream,
Vai las lavar <a> alva        5   Goes to wash them, white as dawn.

Levantou s' a loucana             The beautiful girl arose
Levantou s' <aa> alva             She arose at dawn
E vai lavar delgadas              And goes to wash her garments
Eno alto                          In the stream,
Vai las lavar <a alva.>      10   Goes to wash them, white as dawn.

<E> vai lavar camisas             And goes to wash her clothes--
Levantou s' <aa> alva             She arose at dawn--
O vento lhas desvia               The wind takes them from her
Eno alto                          In the stream--
Vai las lavar <a> alva.      15   Goes to wash them, white as dawn.

E vai lavar delgadas              And goes to wash her garments--
Levantou s' <aa> alva             She arose at dawn--
O vento lhas levava               The wind lifted them from her
Eno alto                          In the stream--
Vai las lavar <a alva.>      20   Goes to wash them, white as dawn.

O vento lhas desvia               The wind takes them from her--
Levantou s' <aa> alva             She arose at dawn--
Meteu s' <a> alva en ira          The radiant girl grew angry
Eno alto                          In the stream--
Vai las lavar <a alva.>      25   Goes to wash them, white as dawn.

O vento lhas levava               The wind lifted them from her--
Levantou s' <aa> alva             She arose at dawn--
Meteu s' <a> alva en sanha        The radiant girl grew grim
Eno alto                          In the stream--
Vai las lavar <a> alva.      30   Goes to wash them, white as dawn.

B 569 f. 127r V 172 f. 23r-v

2, 7, 12, 17, 22, 27 <aa> supplevi

5, 15, 30 <a> supplevi

8 delgadis V

10 las] la B

11 <E> Lang; cf. v. 16 Vay B : Voy V

20 Vay B : uya V

21 desuya V : deuya B

23, 28 <a> supplevi

Pero Meogo 5

<Levou s' aa alva>, levou s' a velida,
Vai lavar cabelos na fontana fria
Leda dos amores, dos amores leda.          3

<Levou s' aa alva>, levou s' a loucana,
Vai lavar cabelos na fria fontana
Leda dos amores, dos amores leda.          6

Vai lavar cabelos na fontana fria,
Passou seu amigo, que lhi ben queria
Leda dos <amores, dos amores leda.>        9

Vai lavar cabelos na fria fontana,
Passa seu amigo, que <a> muit' a<ma>va
Leda dos a<mores, dos amores leda.>       12

Passa seu amigo, que lhi ben queria,
O cervo do monte a augua volvia
Leda dos a<mores, dos amores leda.>       15

Passa seu amigo, que a muit' amava,
O cervo do monte volvia <a> augua
Leda <dos amores, dos amores leda.>       18

B 1188 f. 152v V 793 ff. 124v-125r

1, 4 <Levou saa alva> supplevi
(ex B 569 / V 172 [Dinis], v. 2)

1 <Levou-se mui cedo> Micha: <Levou- s'a
loucana> Nunes : <Leda dos amores>
Bell uelida B : uenda V

4 <Levou-svelida> Nunes : <Dos amores
leda> Bell

8 passa Nunes; cf. vv. 11, 13, 16

10-12 copied at the end of the text in BV

11 <a> Nunes muita<ma>va Monaci
(cf. v. 16) : muytaua B : muyau9 V

15 dos a] d9 a V : do a B

16-18 om. V

16 amana B

17 noluya B <a> Nunes; cf. v. 14

<She arose at dawn>, the lovely girl arose
and goes to wash her hair in the cool fountain,
Happy in love, in love and happy.                   3

<She arose at dawn>, the beautiful girl arose
and goes to wash her hair in the fountain's cool
Happy in love, in love and happy.                   6

She goes to wash her hair in the cool fountain
And along came her boy, who really loved her
Happy in love, in love and happy.                   9

She goes to wash her hair in the fountain's cool
And along comes her boy, who truly desired her
Happy in love, in love and happy.                  12

And along comes her boy, who really loved her
The stag of the hills was stirring the water
Happy in love, in love and happy.                  15

And along comes her boy, who truly desired her
The stag of the hills set the water astir
Happy in love, in love and happy.                  18

King's College London & The Johns Hopkins University 

Note: This is an ever so slightly revised version of a paper given at King's College London in February 2005. I am pleased to acknowledge a heavenly host of friends and scholars, too numerous to name, who read and commented on earlier drafts. Thanks in particular to Peter Bing, Rosario Ferreira, Tony Grafton, Kirstin Kennedy, David Nirenberg, Stephen Parkinson, Stephen Reckert, and Flora Ward. I thank the Fundacao para a Ciencia e Tecnologia for a research grant that made this work possible.

Works Cited

Aldrich, Keith. 1975. The Library of Greek Mythology by Apollodorus, translated, with notes and indices (Lawrence, KS: Coronado Press)

Alvar, Manuel. 1971. Cantos de boda judeo-espanoles (Madrid: Instituto Arias Montano)

Asensio, Eugenio. 1970. Poetica y realidad en el cancionero peninsular de la Edad Media, 2nd edn (Madrid: Gredos)

Azevedo Filho, Leodegario A. de. 1974. As Cantigas de Pero Meogo (Rio de Janeiro: Gernasa e Artes Graficas)

Beltran Pepio, Vicente. 1984. 'O vento lh'as levava. Don Dinis y la tradicion lirica peninsular', Bulletin Hispanique, 86.1-2: 5-25

Bing, Peter and Rip Cohen. 1991. Games of Venus: An Anthology of Greek and Roman Erotic Verse from Sappho to Ovid (New York/London: Routledge)

Brea, Mercedes. 2000. 'Levantou-s'a velida, un exemplo de sincretismo harmonico', in Estudos a Ricardo Carvalho Calero, ed. by Jose Luis Rodriguez, vol II: Literatura Miscelanea (Santiago de Compostela: Parlamento de Galicia/Universidade de Santiago de Compostela), 139-51

Campbell, D. A. 1982-93. Greek Lyric, 5 vols (Cambridge, MA: Harvard University Press; London: Heinemann)

Cohen, Rip. 1987. Thirty-two Cantigas d' amigo of Dom Dinis: Typology of a Portuguese Renunciation (Madison: Hispanic Seminar of Medieval Studies)

--. 1996. 'Danca Juridica. I. A poetica da Sanhuda nas cantigas d' amigo; II. 22 cantigas d' amigo de Johan Garcia de Guilhade', Coloquio-Letras, 142: 5-50

--and Federico Corriente. 2002. 'Lelia doura revisited', La coronica, 31.1: 19-40

--. 2003. 500 Cantigas d'Amigo: Edicao Critica / Critical Edition (Porto: Campo das Letras)

CSM = Mettmann, Walter. Afonso X, O Sabio: Cantigas de Santa Maria, 4 vols (Coimbra: Por Ordem da Universidade, 1959-72)

DDD = Dictionary of Deities and Demons in the Bible, ed. by Karel van der Toorn, Bob Becking and Pieter W. Van der Horst (Leiden/New York/Koln: E. J. Brill, 1995)

Deyermond, Alan. 1979. 'Pero Meogo's Stags and Fountains: Symbol and Anecdote in the Traditional Lyric', Romance Philology, 33.2: 265-83

Ferreira, Maria do Rosario. 1999. Aguas Doces, Aguas Salgadas: da funcionalidade dos motivos aquaticos nas cantigas de amigo (Porto: Granito)

--. 2002. 'A sombra de Tristao: Do potencial estruturante da Materia de Bretanha na mundivisao aristocratica do Portugal medieval', in Materia de Bretanha em Portugal (Lisbon: Colibri), 159-75

Frenk, Margit. 1987. Corpus de la antigua lirica popular hispanica (siglos XV- XVII) (Madrid: Castalia)

HALOT = The Hebrew and Aramaic Lexicon of the Old Testament, 5 vols. (Leiden/ New York/Koln: E. J. Brill, 1994-2000)

Lang, Henry R. 1894. Das Liederbuch des Konigs Denis von Portugal (Halle: Max Niemeyer)

Lewis, Charles Bertram. 1922. 'The Origin of the Weaving Songs and the Theme of the Girl at the Fountain', Publications of the Modern Language Association of America, 37.2: 141-81

L-P = E. Lobel and D. L. Page, Poetarum Lesbiorum Fragmenta (Oxford: Clarendon Press, 1955)

Maas, Paul. 1962. Greek Metre, trans. by Hugh Lloyd-Jones (Oxford: Clarendon Press)

Macedo, Helder. 1996. 'Uma Cantiga de Dom Dinis', in Reckert and Macedo 1996: 59-70

Mendez Ferrin, Xose Luis. 1966. O Cancioneiro de Pero Meogo (Vigo: Galaxia)

Michaelis de Vasconcellos, Carolina. 1895. 'Zum Liederbuch des Konigs Denis von Portugal', Zeitschrift fur Romanische Philologie, 19: 513-41

--. 1904. Cancioneiro da Ajuda, 2 vols (Halle: Max Niemeyer)

Nunes, Jose Joaquim. 1926-28. Cantigas d'Amigo dos Trovadores Galego-Portugueses, 3 vols (Coimbra: Por Ordem da Universidade)

Page, D. L. 1975. Epigrammata Graeca (Oxford: Clarendon Press)

Parkinson, Stephen. (forthcoming, 2006). 'Concurrent patterns of versedesign in the Galician-Portuguese lyric', in Papers from the 13th Colloquium of the Medieval Hispanic Research Seminar, ed. by Jane Whetnall and Alan Deyermond

PMG = Poetae Melici Graeci, ed. by D. L. Page (Oxford: Clarendon Press, 1975)

Reckert, Stephen and Helder Macedo. 1996. Do Cancioneiro de Amigo, 3rd edn (Lisbon: Assirio e Alvim)

Riquer, Martin de. 1983. Los trovadores: Historia literaria y textos, 3 vols, 2nd edn (Barcelona: Ariel)

Stegagno Picchio, Luciana. 1982. La Methode Philologique: Ecrits sur la litterature portuguaise, 2 vols (Paris: Fundacao Calouste Gulbenkian, Centro Cultural Portugues)

Tavani, Giuseppe. 1967. Repertorio metrico della lirica galego-portoghese (Rome: Ateneo)

--. 1986. 'Filologia e critica textual na edicao das cantigas medievais', in Critique textuelle portugugaise: Actes du Colloque (Paris: Fundacao Calouste Gulbenkian), 29-39

Vallin, Gema. 1997. 'Variaciones sobre el alba: a proposito de algunas cantigas gallego-portuguesas', in O Mar das Cantigas (Santiago de Compostela: Xunta de Galicia), 329-44

Weiss, Julian. 1988. 'Lyric Sequences in the Cantigas d'amigo', Bulletin of Hispanic Studies, 65: 21-37

RIP COHEN

Einen Satz verstehen, heisst, eine Sprache verstehen.

Wittgenstein

(1) See Stegagno Picchio 1982: ii, 317-19.

(2) The relationship was first noted by Lang 1894: 134: 'Sowol bezuglich der form als des ausdrucks stimmt mit diesem liede eines von Pero Meogo, V 793, uberein'.

(3) On Meogo, see Mendez Ferrin 1966, Azevedo Filho 1974, Reckert & Macedo 1996: 108-31, Deyermond 1979, Ferreira 1999: 83-120.

(4) Reckert & Macedo 1996: 48, 65. Nunes (1926-28: iii, 27) takes alva to be an adjective in the second verse of each strophe. Lang had understood it as an adverb (in der fruhe) throughout, as is clear from his glossary (s.v. alva). Cf. Michaelis 1895. Throughout this article the texts and numbering of the cantigas d' amigo refer to Cohen 2003.

(5) Reckert & Macedo (1996: 219-20) write alva throughout (in the second and fifth verse of all strophes and in the third verse of the last two), follow Nunes in printing loucaa for the manuscripts' loucana, and decline to restore the initial E in v. 11. Other changes affect only spelling and punctuation.

(6) Parkinson (forthcoming) reaches conclusions almost identical to mine, observing that although the poem seems 'polymetric to excess, with four different line lengths in the five lines of the strophe', 'if [...] we restore most of the presumably elided vowels, the result is an equally surprising metrical regularity'. He writes that 'editors who use forms like alva (Reckert & Macedo 1996: 46, 219) are anachronistically trying to shuffle the problem off into morphology, by presupposing the morphologised contractions of later forms of Portuguese'. Yet Parkinson still holds that 'the poem is clearly an exercise in ambiguity at several levels--metrical, phonological, and textual'. Arguing for an accentual pattern in this poem, he says (per litteras electronicas), 'It is not self-evident that a performer of this poetry could produce a phonetic sequence ambiguous between alva, a alva and aa alva (particularly if there were no possibility of elision), except by relaxing the constraints on strict syllable counting (easier in singing than in recitation)'.

(7) 6' 4' 6' 3' 5'--see Tavani 1967: 314 and schema 119:1.

(8) Maas 1962: 23, 45-51. Strictly speaking, it is only in vv. 23 and 28 that external responsion requires <a> alva. In v. 11 Lang restores <E> for metre and parallelism.

(9) Tavani (1986: 33-34) says it is impossible to restore the missing hemistich.

(10) My conjecture for the first hemistich (first published in Cohen 1996: 45, note 41) is defended and explained in Vallin 1997. Independently, Ferreira (1999: 105-06) argues that the missing hemistich in Meogo must have contained a reference to the dawn. Michaelis (1904: II, 60) had already thought as much, proposing <Levou-se mui cedo>.

(11) Odyssey 6.31. See also Anacreon PMG 385: 'I come up from the river bringing the washing all bright' (trans. by Campbell 1988: ii, 74).

(12) Sappho 49 L-P; trans. by Bing & Cohen 1991: 76, slightly modified. This fragment is cited by Deyermond 1979: 278.

(13) Anthologia Palatina 5.83 = Anon. LXXII Page; trans. by Bing & Cohen 1991: 188.

(14) Frenk 1987, nos. 972 and 973. Both are cited by Reckert (Reckert & Macedo 1996: 225, 227), the second by Deyermond 1979: 277 and Beltran 1984: 22.

(15) Alvar 1971: 240.

(16) Cohen 1996: 12-27.

(17) Ovid writes in the Ars Amatoria (1.665-66): pugnabit primo fortassis et 'improbe' dicet; / pugnando uinci sed tamen illa uolet (Maybe she'll fight back at first and say 'fresh!'/ But she will want to be won over in the fight). The ensuing passage has sometimes been misinterpreted as suggesting rape, but what Ovid is recommending is gentle pressure. At AA 2.197 he says to back off if you know you're not wanted. At Metamorphoses 1.525-30, Ovid combines the attractiveness of the woman who runs away with the sexuality of the wind: Daphne's clothes are blown off as she flees Apollo, enhancing her beauty.

(18) Bing & Cohen 1991: 264, slightly modified.

(19) As though this wind were Boreas, who in the Metamorphoses of Ovid decides to abandon sweet words and take Orithyia by force (6.682-721). In Apollodorus' Library of Greek Mythology, where the story is told very briefly, there is an element of interest here: 'As Oreithuia was playing by the river Ilissus, Boreas kidnapped her and had sex with her' (Aldrich 1975: 85). Plato (Phaedrus, 229a-b) already knew a version in which the river Illisus was the site of the snatching.

(20) We do find a reference to torn clothes in Meogo 8 Fostes, filha, eno bailar / e rompestes i o brial, but that symbolizes the loss of virginity and does not imply violence.

(21) See Weiss 1988, Cohen 1987, Cohen 1996: 27-36.

(22) Cohen 1987.

(23) Only one other cantiga d' amigo, Charinho 1, has 30 verses, and only two, Solaz 2 and Berdia 4, have more, with 32 each.

(24) In the corpus of cantigas d'amigo there are only five such songs, the others being Reimondo 2, Solaz 2, Zorro 8, and Afonso Sanchez 2. Of these, the only one that was definitely earlier than Dinis is Solaz 2, which (coincidentally?) contains three words in Andalusi Arabic (see Cohen & Corriente 2002) and is one of only two cantiga d' amigo with 32 verses.

(25) There is no /h/ in Galician-Portuguese, so its absence is insignificant.

(26) For example, Moshe Ha-sefaradi (baptized as Pedro Alfonso in 1106), Ramon Marti (who died about 1286), and Ramon Llull (1235-1326). I am grateful to Harvey Hames, of Ben Gurion University, for clarifying this point.

(27) Guilhade 19; see Cohen 1996: 34-36. And, most curiously, in Johan Airas 23, the first poem of the second group of 22 (and so corresponding to Dinis 17), the initial letters of the three strophes and the fiinda, read vertically from top to bottom, spell AVBE, so that we find dawn, or the French word for it, in the centre of Johan Airas' set.

(28) See HALOT, s.v. c a o ('breeze' 'breath' 'wind 'sense' 'mind', etc.) and DDD, s.v. Holy Spirit.

(29) We should remember that the Zohar (Book of Splendour), one of the most important texts in Jewish mysticism, is thought by secular scholars to have been composed in Castile (mainly in Avila and Guadalajara) from around 1280. Orthodox tradition holds that it was written by Rabbi Shimon Bar Yochai in Eretz Yisrael/Palestine in the second century.

(30) The fountain is implied in the Protevangelium of St. James the Less, and appears in Pseudo-Matthew. See Lewis 1922: 150-51.

(31) There was a tradition that Dinis composed songs in honour of the Virgin. See Michaelis 1904: II, 112-13. Michaelis (1904: ii, 235, note 5) herself ventures: 'Nao seria de admirar, se D. Denis tivesse imitado o avo tambem como trovador da Virgem'.

(32) CSM 340, where alva is a palavra rima. This is taken from Cadenet, S'anc fui belha ni prezada (Riquer 1983: III, 1231). In Dinis 17, alva is a mot equivoc, since it means both 'dawn' and 'girl'.