3. poiri’ieu. The reading of C is supported by the slightly different readings of ORST. The other MSS do not offer a consistent alternative reading. The a of poiria needs, however, to be elided in order to satisfy the metre requirements; cf. II, l. 7 (note).
8. tot ai. CRS give ai tot(z), the order of words being supported by T (es tut). The order of words offered by ADIK is preferred, however, particularly as it is supported by Mc and O, which is normally closely related to CS (see ‘Classification’).
16. amor, C contains a scribal error (amo). All the other MSS have amor.
18. sauputz. Although accepted as an alternative form of the more normal past participle, saubut, the word in fact breaks the rules of phonetics: the p between two vowels vocalizes into b (cf. Altprov. Elem., p. 42; Anglade, p. 137). It is probably here a question of a Latin influence.
19. saupi. Cf. note to l. 18. The normal first person preterite form of saber is saup < sapui (cf. Altprov. Elem., p. 99; Anglade, p. 308) but saubi, saupi exist as rare forms (cf. Anglade, p. 308; Grandgent, p. 144).
·l mielhs. This is an example of the neuter mielhs used adjectivally with reference to a person. Cf. R. de Vaqueiras, Epic Letter, III, ll. 30-1:
E vos mandetz cinq escudiers muntar,
de tot lo mielhs que vos saupes triar,
...
See also Literaturblatt, XV (1894), 191, in which A. Stimming cites other examples, and Altprov. Elem., p. 123.
21. ben leu. The reading of the ADIK group is chosen. In this stanza, which Appel considers may have originally been missing in MRc (see ‘Classification’), we frod that M has been added according to the ADIK tradition, whereas Rc generally offer independent readings or are individually linked in turn with both traditions. It is significant, therefore, that in this line Rc both join ADIKM in opposition to COST.
23. For this line the reading of ADIK is again chosen in preference to that of CST. The line is missing in O. M retains its relationship with ADIK. Rc, although independent, appear to be nearer to ADIK than to CST (cf. note to l. 21). Appel (p. 76) suggests that the omission of the line in O might indicate a mutilation in the common original of the COST group.
24. me say ... s’es. Appel chooses the reading of ADIKM (sai ben ... es). For the first half of the line (sai ben) c joins this group, whereas R joins COST. However, c is perhaps more closely related to COST than is at first apparent in that the use of me vay in the second half of the line in both c and R may well be due to the influence of the earlier me sai of the COST tradition. This strong suggestion of support on the part of Rc justifies the retention of the reading of COST.
25. Greus ... maltraitz. Their spelling greu indicates that the scribes of CDIKST were influenced by the impersonal expression greu m’es, in which the neuter form of the adjective is often employed; cf. Altprov. Elem., p. 71; Grandgent, p. 96. (According to Grandgent, the normal nominative form ending in s is also used in these constructions.) By offering the accusative form maltrag and thereby providing a direct object for sufrir CT are, however, the only MSS to follow through the impersonal use for the rest of the line. The other MSS give variations of the normal form of the nominative singular (mals traitz—maltraz). Moreover, in order to justify the use of the impersonal construction, dolors in the following line would need to have the normal accusative form, as corresponding to maltrag, and offered by RT only.
We have followed Appel in preferring for this line the version of AO, representatives of both traditions. The nominative form greus is thus used to serve both subjects, maltraitz and dolors (in l. 26): ‘Suffering and grief are hard for me to bear.’ (Cf. note to III, l. 9.).
In not adding s after mal the scribes of AOS have regarded maltraitz as forming one word. This view is supported by Levy (Pet. Dict., p. 234).
Lavaud (Troub. Cant., p. 70) erroneously groups Mc with DIK as the MSS which he states Appel follows for his reading mals tragz. Both MSS, in fact, omit the whole of this stanza.
27. no·m pot lo cors. Appel prefers the order of words given by ADIK (lo cors no·m pot). T joins them here while being firmly linked with COS in the rest of the stanza. It has been noted under ‘Classification’, however, that of the four MSS of the second group T is the most distantly related.
It seems reasonable to retain here the version of COS, particularly as R, although having a slightly different reading, supports them in the order of words, whereas elsewhere in this stanza it is generally attached to the ADIK group. The stanza is missing in Mc.
28. autr’amistatz. See note to IV, l. 41.
31. s’ieu. C has si ieu. It is necessary to follow all the other MSS in eliding the i for reasons of metre. Cf. II, l. 7 (note).
33. soven. C is the only MS with the reading per vos. The version clearly favoured by all traditions (ADKMSTc), and which also produces the alliteration soven sospir, is preferred here. O is independent, the stanza is missing in R, and the relevant part of the line is omitted in I.
34. gran pen’e. The reading of AMOTc is preferred to that of CS (greu), shared by DIK. Where the two main traditions are in opposition earlier in this line (e—en) and in l. 33, Mc identify themselves with the COST group. It is notable, therefore, that in this instance they depart from CS in order to follow the other two MSS of this group, OT. The support of A provides further justification for this reading. The appearance of greu later in the line in OT might be the result of the influence of the earlier greu in CS.
The a of pena is superfluous from the point of view of the metre and is therefore elided. Cf. II, l. 7 (note).
37. mas. The reading of COSc, the meaning of which is quite acceptable, is retained in preference to that of ADIKMT (e), chosen by Appel. It is significant that c supports COS at this point while offering an entirely independent reading for the line as a whole, as elsewhere in the stanza (cf. ll. 35, 36, 40). See ‘Classification’.
38. vos. C contains a scribal error (vo).
39. Tu. In addressing his lady elsewhere in the stanza Peire employs the normal second person plural. A change to the second person singular from the plural does sometimes occur in Old Provençal, but normally only in a narrative poem (cf. Altprov. Elem., p. 129). Tu is therefore taken here to denote a different person in the form of Peire’s interlocutor.
40. Appel appears to have been justified in not including this line. He considers (p. 76) that one would expect it to read something like Lo sens qu’ieu ai totz es ab liey, and points out that most of the MSS merely offer a remodelling of l. 8. The independent version of M seems, in fact, to be the only one approaching this meaning.
Lavaud (Troub. Cant., pp. 70-1) does not follow any of the MSS but goes as far as including (in brackets) mas at the beginning of the line, in order to express the contrast he says one would expect after l. 39, and de liey at the end, on the grounds that they are the words which constitute the refrain in all the other stanzas. Suchier (op. cit., 1883, 1344), in fact, proposes the following reading, having regard to the refrain: Mas luenh de ti e pres de liey.
ADIK form one group and OST another. M is fairly close to OST, while c is more independent. The scribe of C appears to have followed the ADIK tradition for the first half of the line and the OST tradition for the second half.
It is interesting to note that the plural form els offered by DIK may suggest an attempt on the part of the scribes of these MSS to correct the grammar of A (el bes) by making bes accusative plural.
41. que. C is the only MS with the reading que·t. The reading offered by all the other MSS except T, which has ll. 41-61 missing, is therefore preferred. Planher is used transitively here, with que as the direct object. CMR are the only MSS with the correct form of the second person singular (plangz) (cf. Anglade, p. 339).
ia tem. The reading of ADIK is chosen in view of the fact that they are joined by O and S (which has a scribal error, la tem). Independent versions are offered by C (laissim), Mc (qe tems—ge tem), and R (temi).
46. sofre. The normal form of the imperative singular found in most of the MSS (cf. Anglade, p. 293) is preferable here to either of the readings offered by the COS tradition. C has sofri and OS have soffra.
48. No? — Qu’o fas. The reading of DIKM has been adopted in preference to the less satisfactory one of CORS (no quas fai (fa) (faz)). The respective readings of A and c (noquo fai—noqua o fas) show a tendency towards both groups. The MSS of the Breviari, on the other hand, incline towards the version of DIKM. They have either no quo or no que, and all except three offer fas.
49. n’ai. The grouping of the MSS follows a pattern similar to that of the previous line. The reading of the first group is preferred to that of CORS (ai) in view of the support given by Mc.
55. s’amar. C has si amar. For the elision of the i cf. l. 31 and II, l. 7 (note).
56. aissi·t p. We adopt the reading offered by the majority of the MSS. ADIK are joined by all three MSS MRc as well as by S, while C (aissi p.) is supported only by O which is independent for the rest of the line.
It is worth noting, however, that both readings are equally acceptable, as jauzir de may be used with or without the reflexive form in this context (cf. B. von Vent., 9, l. 41; 25, l. 22 (j. de); 18, l. 22 (se j. de)).
57. mant. C has prec. The reading given by AS is chosen, as it seems significant that S should support A after being linked with C for most of the remainder of the poem. In the tornada as a whole S plays an important part in being often the deciding factor in the choice of readings.
58. s’el es bos. The reading of S, supported by A (s’il es bon), is preferred to that of C (si bos es) in order to meet the demands of the rhyme scheme. The e of se needs to be elided, however, for reasons of metre. Appel’s version (s’il es bos), which Suchier (loc. cit.) prefers to express as si l’es bos, is not given by any of the MSS but appears to be a corrected form of A.
59. e puois vol. S again joins A in opposition to C (e si vol). Their reading is therefore again preferred.
60. dreit-n’avetz. The version of AS is once more chosen in preference to that of C (dreit na lieys). It was probably intended that dreit-n’avetz should correspond to tort-n’avetz in l. 57. No clue as to the identity of dreit-n’avetz has been found.
en Saves. The reading of C is retained here, as it is supported by S in opposition to A (on ill es).
Savès was probably a small district in the present day Departments of Gers and Haute-Garonne between Samatan and the Isle-Jourdain. Although it originally fell under the jurisdiction of the Counts of Toulouse and, in fact, gave its name to one of the archdeaconries of the church of Toulouse, it appears to have belonged, as early as the twelfth century, to the Counts of Comminges. (Cf. Hist. gen. Lang., III, pp. 70, 336, 375; XII, pp. 161, 200, 241.)
61. sal. Appel has chosen the reading of AS (salv). The spelling of C is, however, quite acceptable. After l and r the v hardens to f or is dropped: salve(t) > salf (or sal). (Cf. Altprov. Elem., p. 45; Roncaglia, p. 76.) |
