What if, for one reason or another, Liszt had never completed his Second Piano Concerto and it was suddenly rediscovered? Its first recording would probably be carefully prepared but, given the brief acquaintance, not the most inspiring imaginable; the critical response might well speak of banal themes, effects without cause, and unconvincing structure (plenty of musicians who should know better say those things anyway).
I say this because on first hearing the so-called Concerto No. 3 struck me as a fairly ramshackle assembly of not very distinguished ideas which might well have come from the reject bin for the two concertos we all know and love (or hate). But then going back to it the ideas start to stick, and to sound less trite. The main lyrical theme, unfolded in a piano soliloquy from 5'43'', seemed at first something like a kitsch heavenly paean to welcome fallen heroes—with its eventual up-tempo transformation at 13'10'' a comic celestial knees-up. But now I just hear it as a rather noble, almost distinguished idea. And the superior smile provoked first time around by the full orchestra march at 4'30'' has given way to enjoyment of its healthy vulgarity.
Well, I haven't reached the stage where I can calmly accept all the two-by-two sequences in the later stages, and there are one or two distinctly awkward-sounding joins (one in particular at 8'46'' bothers me). The concerto is in a single 15-minute movement with plentiful thematic transformation, which is partly why comparisons with the Second Concerto arise. And with the best will in the world I can't imagine that Liszt would consider the piece as it stands worthy of its companions.
What are we listening to in fact? The insert-note, credited jointly to the soloist Steven Mayer and scholar Jay Rosenblatt who reconstructed the piece, speculates that the work may have been prepared for Liszt's tours from 1839 onwards (by which time the other two concertos were well advanced, though they would take 20 or more years to reach their definitive shape). Unforgivably, however, the essay fails to clarify the nature of Rosenblatt's reconstruction. We learn that sketches for the work were discovered in libraries in Germany and the Soviet Union, where they were mistakenly catalogued along with the E flat Concerto (No. 1). And that is all. Of the state of completion of the sketches (continuous drafts, orchestrated?) and of the extent and nature of Rosenblatt's work we are told nothing.
That is a deplorable blot on an otherwise outstandingly valuable issue. The value is, I should stress, primarily musicological rather than musical. The ''Instrumental Psalm'' De profundis was composed in the winter of 1834-5 and is the work of an enthusiastic 22-year-old apparently devoid of all self-critical awareness. At 34 minutes it is Liszt's longest concertante work, and it certainly feels like it. Similarly I cannot imagine many listeners finding the 1853 version of Totentanz an improvement on the familiar and much tauter 1859 version—its inclusion of the ''De profundis'' plainchant sentimentalizes the mood and makes the whole structure bloated.
Still, the interest value of all this is obvious, and Steven Mayer has to be admired for his dedication to not always very thankful tasks—he plays with exemplary brilliance and refinement, and is given full-blooded (at times maybe too full-blooded) support from the LSO. Recording quality is rich, but a trifle overblown, with a tendency to saturate too easily.'