I had heard these recordings a few times before in the 1980s, when they were first released, but then they dropped out of my sight and I went back to Kempff and Brendel and on to Uchida and Perahia when it came to these sonatas. I love these works so much and I also appreciate DG's remastered re-releases of their most highly regarded recordings, so I decided to go for yet another copy of Schubert's "last three." What a fortunate re-release this is. Coming back to Pollini's interpretations after listening to many others, I'm struck by how Pollini stands alone for the supreme clarity of his passage articulation and his masterful grasp of extended, complex and potentially formless works (think of his Schumann, his late Beethoven, the Liszt and Chopin sonatas, and his recording of Schubert's D. 845 sonata) over which lesser pianists often lose control.
A brief comparison with other interpretations of these sonatas. Brendel, Perahia and Pollini are all quite similar in that they take a lean, spare and unembellished approach to the D. 960, as opposed to Kempff, Richter and Uchida, who bring out the sonata's more spacious and sonorous qualities and make so much of the silences between the sonata's many mood changes. For a while I was completely smitten with Uchida's recording, but lately I've gone cold on it. It has the tendency to wax a little too poetic and to be somewhat self-conscious. I have the impression that Uchida feels burdened by the effort to make every note and every passage ring with profundity; eventually the first two movements collapse under the accumulated weight she heaps upon them. Richter to my mind is even more culpable in this respect (his first movement clocks in at about 26 minutes if my memory serves, while Uchida comes in at about 23 and change. Perahia, Pollini and Kempff manage it in about 18-20, while Brendel foregoes the repeat and his first movement is about 12-13 minutes long). Brendel does a magnificent job with the D. 960, but I've come to appreciate the first movement repeat that almost everyone else who records this sonata observes. Pollini therefore supplants him here, but what really stands out is the emphasis Pollini gives to what Schubert was doing with the bass clef. The recording really brings out the left hand complexities of the composition (beyond the foreboding trills) in the first and second movements, while the clean purity of his playing gives the second movement all the transcendent, dreamy qualities that are so abundant in the Kempff recording, albeit in a different manner. As for the third and fourth movements, Pollini's control is just a wonder. It's always eay to overlook these movements because of the overwhelming nature of the first two, but Pollini made me sit up and pay attention in a way I never have with the other interpretations. Perahia is excellent too, but strangely enough (since Pollini is often the one accused of remoteness and over-objectivity), there seems to be something absent from Perahia's recording of a piece that seems to rest on its many emotional intangibles. It's fine--flawless actually--in respect to its architecture and pacing, but there seems to be nothing remarkable about it regarding the emotions it evokes. So here Pollini and Kempff take the honors.
Pollini is also magnificent with the D. 959. Pollini goes here for a lean, mean, stripped-down-to-essentials Schubert, taking a Beethovenian approach that successfully highlights the many dramatic contrasts without any over-exaggeration or over-emphasis. To my mind no one surpasses Brendel's perfectly modulated and more measured interpretation; suffice to say though that Pollini is a clear and close second, and that many might prefer his approach to that of Brendel. I like both and will always own both. Although excellent and flawlessly executed, Perahia takes this tricky sonata just a little too briskly for my taste; Kempff takes it a little too easily and there's too much of a drop off in the level of tension that threads its way through the first two movements; and Uchida makes a complete hash of the slow movement.
If Brendel is the D. 959 sine non qua, he has for me always sounded a tad too brittle in bringing off D. 958, and here both Perahia and Pollini are more successful. The Perahia recording is more thick-textured than that of Pollini, and again I just have to go with the fact that Pollini manages the diffcult task of coupling absolute technical perfection with a deep understanding of these sonatas' other-worldly, emotionally ambiguous qualities.
There's definitely some Pollini out there that I am not fond of; his interpretations of early and middle-period Beethoven and the Diabelli Variations immediately come to mind. Anything from the '70s and '80s is golden though, and taken overall, this is the best 2-CD set of Schubert's late sonatas out there (it also includes the Allegretto and the late Klavierstucke). It will bring endless satisfactions and joys.