Analysis of manuscripts: A 178v (Peirols . et Amors), C100v (Peyrols), D 147r (Peirols), G 48v (staves above stanza I completed with musical notation), I 158r (Peirols . et amors), K 144r (Peirols), L 30r (Peirols), M 183r (Peirol daluergna), N 282v, O 34, R 76v (tenso), S 83 (Peirol), T 159v (Peirols), a 180 (en peirols). − Several factors point to CMOTa’s arrangement of material being the original authorial arrangement (see ‘Order and amount of material’ below). Within this group, O shows intervention to supply consistent ‘b’ rhymes in -i and it also has many other isolated readings and individual errors (see 9, 12, 19 (+1), 25, 28 (+1), 29 (+1), 30, 45 and so on). T and a similarly have many isolated readings and individual errors (see respectively 3, 9, 11, 16, 21, 25, 26, 28, 30, 38 and so on, and 3, 12, 16, 30, 33, 39, 53). In 31 Ta share a reading (·m), in 38 Ra are joined in error, and in 35 Oa share an indifferent alternative reading with the ADGS group; it preserves no superior readings which would recommend its adoption as base. C and M require the fewest number of interventions: C has minor errors in 41, 49 and 52 and isolated readings in 11, 12, 30 (facilior patching following eyeskip?), 46 and 53. M on the other hand has only minor slips in 27, 30, 33 and 44, although on two further occasions (32 and 45) its individual readings are probably the product of later, non-authorial reworking which replaces historically specific ‘kings’ with a blander ricx (32) and a criticism (45) with a courtly platitude. Because of the slightly smaller number of errors, I have based the text on M and corrected its innovative individual readings in 32 and 45 from the majority readings common to all MS groups. The result does not differ very greatly from a text based on C.

Order and amount of material:

    I II III IV V VI VII   y*
  CMOTa 1 2 3 4 5 6 7 -
  ADGS 1 2 3 6 5 7 - 4
  IK 1 2 3 4 5 - - 5
     L** 1 2 3 4 5 6 - 4a
  N 1 2 3 4 5 7 - 6
  R 1 2 3 4 5 - - -

* Supernumerary stanza y, composed of 5 lines not found elsewhere (x1-x5) + lines 50-53 (that is, the second tornada, stanza VII).

** In L stanza ‘y’ is added in the margin and its intended position as 4th stanza is indicated by a little ‘d’ written beside its first line, the letters ‘a’, ‘b’ and ‘c’ appearing next to stanzas I, II and III respectively. A small ‘f’ and ‘e’ have been added beside stanzas IV and V respectively, reversing their order. The table above reflects the order in which the scribe of L first transcribed the song. − The scribe of R has left enough space for the two tornadas to be added.

As Crescini observed, the mss. transmit essentially two versions of the song, one with stanza ‘y’ (ADGS + IKN + L), the other without (CMOTa + R). The mss. which have stanza ‘y’ all lack VII, the second tornada; RIK additionally lack the first tornada VI. While it is not impossible that a dialogue, however fictive it may be, should conclude with only one surviving tornada (compare L’s version), it is much more common for an exchange with any surviving tornadas to have one for each of the participants (as is the case here for CMOTa): see Ruth Harvey and Linda Paterson, The Troubadour Tensos and Partimens: A Critical Edition, 3 voll., Cambridge 2010. In the ADGNS and revised L arrangements, only Peirol is accorded a tornada which follows directly on from his last complete stanza. Not only does this give him an unbalanced lion’s share of lines in the dialogue but, as Crescini observed (p. 499), this sequence destroys the close link between lines 42 and 46, where Peirol’s tornada echoes closely and forms a reply to the words of Love in the second part of stanza V. (In addition to the notion of the kings going in 42 and 46, the idea of war in 43 is picked up in 48, and that of Dalfi’s departure in 47 and 49 is used to counter Love’s cynicism about the barons in 44-45.) − L’s relationship to the two main groups suggests that its first source was a C-type version, but that was then collated with an A-type version from which L derived not only stanza ‘y’ but also a different stanza order which reestablished a regular alternation of speakers, this revised order being indicated in L by the added letters. It is possible that the addition of stanza ‘y’ also explains the omission of VII from L (see Crescini, p. 498). Such scribal changes in stanza order in dialogue pieces which have lost stanzas or have otherwise become damaged are not uncommon: see for example BdT 194.2; 201.5; 350.1; 344.3a; 366.30. Moreover, the role of collation of sources by the scribes of L to supply missing material is clearly demonstrated by the case of BdT 238.2 (see the notes to these pieces in Harvey and Paterson, Troubadour tensos and partimens). − Crescini (p. 497) argued that the version with the tornadas was the original text on the grounds that, firstly, the tornada is necessary but stanza ‘y’ is not; and secondly that, were stanza ‘y’ authentic, it would mean that both interlocutors were advancing the same argument - which I understand to be that the demands of love are superior to those of the crusade [since lovers are lamenting their separation from their ladies because of Saladin]. This is Love’s position in III and V; it amounts to the opposite of Peirol’s. If stanza ‘y’ had belonged to Peirol (which is what its opening address to Amors indicates), he would have been noting that emotion made lovers repent for having taken the cross and would thus substantially be agreeing with Love’s view. As Crescini suggested (p. 499), pero (stanza ‘y’ line 6) is readily understandable as a mistake for Peirol (compare A’s reading in stanza I) and this may have been at the root of the reworked stanza ‘y’. − The different factors point to CMOTa’s arrangement of material being the original authorial arrangement.

Critical apparatus:

Missing and / or misplaced material is indicated in the table and this information is not repeated in the variants below.

I.  1 amor T; trobet] ac tot R; parti O    2 cors R; del ADGLS; son] sieu ADGLOS    3 raizon GT, renson a; mais sasailit IK, maisailit N, masagi O, masaglic T, massaillit RSa    5 Amic G; perols A, peirol GLOSa, perrol N    6 amatz N; loignaz N    7 emi emon D    8 es v.] er u. CIKOa, auretz ADGL, auez S; er uostrem | tensios N    9 pueis] om. O, plus T; quel O; q. ualrz u. G, q. ualgreszuos L, ce uareuos T; diatz q. u. pueis vos R.

II.  10 serui O    11 e pietatz C, e nul pecat IKNT, e chauzimen R, e n. pechat a; nous e. p. CNRa, nouosen p. T; e nul pechatz de mi nous p. O, Que negus tortz no men p. ADGLS    12 Cum uos C, Euo D; qan] cum ADS, qun L, tan RT, cani a; e s. qe tan peci O    13 auut CIKR, aut Da, aguz GL; chauzimen DO    14 demen AL, di ni | ent N    15 tengatz ADGLOS; dor enan GO    16 quals CIKNO, plus ADGLS; no LS, nouos T; demanz a    17 car R; nul IKLOS, lunhs R; autre DGIKLS; guizandos G    18 Nom pogra O, nom poiri a.

III.  19 Peirols CDIKORTa; uos m. O, metres IKNRTa, metenz G; obli O, oblitç T    20 bella ADGLOS; auinen L, uailen O    21 qui CDO; lautrier ADGLS; acolli O,  acuglic T    22 om. IKLN; Aitan AG    23 pel ADG; mieu ADGLOS    24 torp A, truep a; anez G; legier O; talen LNOT    25 e nous (nos S) e. g. ADGIKLNS, e non era ies ies O, e no faziatz R, enouos e. g. T; seblan T, semblam a    26 gars I, gai RT; ni t. O, eran N, etat T    27 en nostra L, en uostras en uostras M; tensos expunctuated, followed by cansos D.

IV.  28 anc an O; mais om. L; no CDOa, nous GL, nos S, nouos T; faili O, faglic T; A. may | nous ay f. R    29 Er o fauc A, Eto faiz DS, Er o faz (fach L) GLR, m. aral fas O, m. er fagli T    30 i. crist qem guit AGLS, ihesu qem guit D, d. iesu q. g. MR, d. yhesu qen gait T, d. qen guit a; e p. dieus quem sia g. C, e preo nai ezu qem gai O    31 O S; qel O, cem Ta; tramata T; breu metet N    32 dels (del S, densz L) dos (dous L) reis ADGLS, e. rics M, etrelrei T; acor damant N, acordamens R    33 uan L, ua OS, irai changed from vai a; lo | inhan M, tarian S    34 auriay R; mesters LRS, me steir N; grair N    35 margqes the ‘q’ added above G; ualens ACDGIKLNOSa; e bos ADGOSa    36 n’ages] Agues O; mis N; Lai a. mains (mais GLS) c. ADGLS.

V.  37 Peirols CDIKOTRa; turcs O; ni] et T, qui expuncuated, followed by ni I; arabi O, rabit T; P. t. marabit N    38 Ja ADGLOSa; pel ADGS; u. enui | amen A, u. iu | iamen D, u. mandamen G, u. uançiment N, uostres uazimen Ra, u. ciausimen T    39 no om. T; laisseron C, lasseran S, laissaram a; tort D, torn IK, cor NT; daui O, dauit changed from daiut a    40 uos or nos G    41 chassat A, quantatz C, cachaz D    42 eriseuos T; e·l] sil ADT, on L, els R; reis NOR; no CDLNR; uian N; unouan S    43 ueser T; latz las g. IK; qenfan Oa, qen | [illegible] R    44 es garatz IK, egardas M, es gueratç N; des L; barols I    45 comsi qeron ADGLS, Con sistrobam O; ozasos S; con fan bellas messios M, casi t. ociaisons T.

VI.  46 qan ADGLNS, sil T; los reis O; no u. C, iran ADGLNS    47 nos G    48 Ja ADGLNOS; gerra ACDGLSTa; nos N    49 n. remanha C, nollaisera O, n. remara T.

VII.  50 Peirols CT, pero ADGIKLNS; mains ADGNS, molt C, mait T; amics ADNO; partan S    51 lor ADGIKLNOSTa; amigas corrected from amicas G; ploiran O    52 Qi NO; sen IKLS, se DGNO; saladins AO, sa | lidis C    53 remansera LS, remandria O, remasegran T, remazeron a; ab nos C.

Supernumerary stanza ‘y’ in ADGIKLNS (base A):


Amors, midonz, pois la vit,


ai amada longamen;


enquer l’am, tant m’abellit


e·m plac al comenssamen,


mas foillia no·i enten.


Pero mains amics partran


de llor amigas ploran

    que, si Saladins non fos,

sai remaseran ioios.

1 pos midonç N, pois anc eu IK    2 lai a. IK; finamen G    3 Qenqer G, Em quer IK; tamnabilit N    5 Qe G    The variants to 6-9 are given under those for lines 50-53 above.

(‘Love, since I saw my lady, I have loved her for a long time; I love her still, she attracts and pleases me as much as (?) in the beginning, but (?) I don’t think there is any folly in this (?). However, many lovers will depart, weeping, from their ladies who would stay here and be joyful if it were not for Saladin.’)


Dating and historical circumstances:

The poem was composed following the fall of Jerusalem in October 1187 (37-39) to the forces of Saladin (52) and while conflict (30-32) and fighting (43) between the kings (32) of England and France (who had taken the cross in the period late September 1187 to January 1188) were delaying their departure for the Holy Land. This also delayed the arrival of crucial help (33) to the marqes, Conrad of Montferrat, who was holding out in Tyre with the surviving remnants of the nobility and military forces of the kingdom of Jerusalem (34-36: see Steven Runciman, A History of the Crusades, 3 voll., Harmondsworth 1971, first published Cambridge 1951-1954, vol. II, pp. 471-472). The kings are Philip Augustus on the one hand and most probably Henry II (according to Crescini and Aston) or Richard I (as proposed by De Bartholomaeis). Both the Plantagenet kings were in conflict with Philip in the period 1188-1190, but in 1188 this strife also involved rebellion by the barons in Aquitaine and Richard’s attack on Toulouse and campaign in Quercy (compare 42-45): see John Gillingham, Richard I, New Haven and London 1999, pp. 88-94. These factors more strongly support an identification of Henry II as one of the kings and hence a dating of the piece to before Henry’s death in July 1189. After Henry II’s death, it is not clear that Richard engaged Philip in any significant warfare and De Bartholomaeis’s suggestion seems to concentrate on the discussions and lack of agreement between King Richard and Philip and not to take account of the poem’s references to their gerras. − Peirol’s song is then contemporary with BdT 80.17, in which Bertran de Born similarly evokes the kings’ delay and Conrad’s valiance. − Peirol seems to have been composing songs for some time already (see lines 25-27). Since 366.28 can be dated to 1221-1222, he appears to have enjoyed a long career.


Textual notes:

3. Crescini and Aston note that asailhit 3 p. sg. pret. is a form used for the sake of the rhyme (Crescini calls this ‘alla francese’) and that the same applies to acuilhit in 21. Paden (p. 192), however, gives as possible forms partit and partic as well as the expected outcome parti (the first two being analogical formations). Together with comen (4), Riquer saw this as creating a ‘French’ colouring.

11. no·s (IKM) = no·us (CNORa): see Crescini, Manuale, p. 81 and n. 3, and note to BdT 12b.1, 36 in Harvey and Paterson, Troubadour tensos and partimens.

20. Crescini and Aston identify the donna as Sail de Claustra, sister of Dalfi d’Alvernhe, who according to Peirol’s vida was the poet’s lady. I see no reason to be so precise.

21. ADGLS l’autrier may suggest some specificity is being attached to the lady, but this usage, which recalls the narrative frame of a pastorela, does not work syntactically with 22 e tan (compare AG’s intervention there).

25. Strict grammar would require semblans.

28. falhit: the ‘t’, necessary for the rhyme, has been extended by analogy also to the 1 p. sg.

30. If we leave aside O, which is garbled, Da suggest that the divergent readings here probably arose from eyeskip or confusion between (d)ieus and iesu(s): compare C’s isolated, 2-syllable ‘filler’ sia, and the monosyllabic crist supplied by AGLS. Strict grammar would require Ihesus (IKN), rather than Iesu (MRT), and I correct for clarity.

33. I correct M’s isolated loinhan (repeated from 6).

35. onratz M(R)T: correction is not imperative.

39. David’s Tower: the main citadel in Jerusalem, captured when Saladin took the city.

46. se concessive, whence the translation. Peirol’s words highlight the worth of Dalfi, although he did not participate in the crusade.

[RH, lb]

BdT    Peirol

Songs referring to the crusades