The first recording of Lutoslawski's most recent large-scale work, the Piano Concerto of 1988, is an important event, but I find the work itself an enigma. Is it an imaginative, resourceful adaptation of the textural and formal conventions of the romantic and early modern concerto to the very personal features of Lutoslawski's own musical language? Or is it an uneasy compromise between old and new—perhaps even a failed attempt to escape into the new from the old? Undoubtedly there are passages which not only recall Bartok and Prokofiev, but even the kind of music which the unsympathetic might dub ''atonal Rachmaninov''. Nevertheless, Lutoslawski—the old Lutoslawski—is never wholly absent, not least when one appreciates the skilful proportioning of four fairly extended movements, and the presence of some lively orchestral textures. The performance itself is exemplary, with Krystian Zimerman immaculate in his gradations of touch and tone: he is indeed a commanding soloist, completely at home with the ambiguous idiom.
The two purely orchestral works go well together. Novelette (1979) is rarely heard, but it is one of Lutoslawski's most attractive and original orchestral compositions, worthy to stand alongside Livre and Mi-Parti. It has the lightness of an extended scherzo, beginning and ending with a flourish that seems to invoke the finale of Beethoven's Seventh, but its dancing rhythms and crisply contrasted sonorities don't preclude more sustained materials, especially in the final section. Chain 3 (1986) is more soberly elegant and shapely, with many ravishing instrumental effects and a tendency to focus on cantabile lines that anticipates the Piano Concerto. These performances of Novelette and Chain 3 are expertly played, but perhaps a little too cautious. Certainly, greater interpretative boldness in Novelette might bring out even more of its abundant character. The sound throughout is spacious and well-balanced.'