Forty years after Raymond Leppard astonished Glyndebourne audiences with reconstructions of Monteverdi’s two late operas‚ the debate he provoked about ‘authenticity’ remains undiminished. In the liner­notes for these new releases‚ Gabriel Garrido unambiguously aligns himself with René Jacob’s assertion that ‘Preparing a performance of Il ritorno d’Ulisse is an undertaking which requires the courage to take decisions‚ lots of decisions’. One man’s courage may turn out to be another’s self­indulgence; certainly lots of decisions have been taken here for which the sources do not provide authority. Just as Leppard’s heavily doctored versions provoked considerable controversy‚ in part because it was not made clear what had been written by the composer and what added by the interpreter‚ so too‚ Garrido’s choices stir up debate‚ in this instance in relation to structural questions (such as adding extra ritornellos) and matters of orchestration‚ where the approach is full­blooded. (In addition to a generous string group and a large continuo one there are also winds and brass in both performances). There is also a degree of re­composition‚ as in scene 8 of Poppea. It’s clear the sources continue to intrigue and puzzle as much as ever‚ though the appearance of Alan Curtis’s masterly edition of Poppea has established a definitive version of the text‚ even while robbing Monteverdi of responsibility for some cherished moments including the celebrated final duet. As Curtis and others have emphasised‚ the texts of 17th­century Italian operas for the public stage (as opposed to court operas) were unstable‚ and changed as they were taken from place to place. The case of Ulisse is less complicated since only one manuscript score survives‚ but the lack of a decent modern edition usually drives performers back to the old Malipiero edition. Gabriel Garrido has taken Malipiero as his basic text‚ but has also added music from elsewhere‚ partly to fill the places in the libretto for which no music survives‚ and partly to insert ‘instrumental pieces which were to accompany the movement of the actor­singers’. These additions include pieces by Francesco Cavalli‚ Biagio Marini and Sigismondo d’India as well as Monteverdi himself. It could be argued that all this reflects both the spirit and realities of 17th­century performance‚ though purists may find the intrusions irritating. Editions apart‚ there remains the thorny question of instrumentation. It is clear that the Venetian theatre orchestras were small‚ partly since the houses were too. But it was also an economic question. For the impresario‚ the priority was to attract an audience by engaging well­known singers who could entertain; all other costs were kept to a minimum. Singers and the libretto were paramount; the orchestra had a purely supportive role. The skeletal nature of the original scores‚ with the ritornellos scored for two violins and bass only and the continuo grouping made up of one or two harpsichords and a couple of lutes‚ has been taken at face value by some conductors. Curtis‚ for example‚ believes that ‘historical evidence does not support the addition of instruments other than continuo to perform simultaneously with voices’‚ and his lead has been followed by others‚ notably Richard Hickox in his 1990 recording with Arleen Auger in the title­role (Virgin‚ 5/90). Yet most conductors prefer to believe that the surviving material is no more than a 17th­century vocal score and that the instrumental parts were lost and need to be recomposed; René Jacobs (Harmonia Mundi‚ 4/91) is a keen proponent of this view. Harnoncourt‚ whose landmark performances of Monteverdi’s two late operas (Ulisse and Poppea‚ both Teldec) were the first major attempt at ‘authenticity’‚ advocates an intermediate position which allows discreet additions while rejecting the ‘inappropriate “flesh” of a much later age’. Garrido‚ as I said earlier‚ sides with Jacobs. For both these new recordings his casts are strong‚ and there is a degree of overlap between the two. Ivan Garcia’s voice is somewhat lighter in tone than that of the traditional Seneca‚ but his performance in the suicide scene is moving. There is some fine singing too from Maria Cristina Kiehr‚ who has made this repertory something of a speciality. But undoubtedly the star performances in both works are given by Gloria Banditelli‚ whose wide dynamic and emotional range‚ combined with a genuinely felt sense of the text‚ makes for impressive readings of both Penelope and Octavia. These recordings complete the Teatro Massimo’s Monteverdi opera project‚ which has already produced a reading of L’Orfeo (K617‚ 10/98) with much ornamentation added to instrumental lines‚ even when accompanying. In my review I was critical of these distractions‚ and I am concerned about the effect in these new recordings too. That and the orchestration question may in the end count as much with many listeners as the performances of the singers‚ a curious reversal of 17th­century priorities.