Peter O'Hagan, Wigmore Hall, London

University of Surrey-based Peter O'Hagan is a formidable scholar-pianist whose recent researches have focused on that most mysterious of post-war conceptions, Pierre Boulez's Sonata No 3 (1955-63) – mysterious, not only because its published sections offer the performer some degree of freedom in the ordering of sections, but because other bits of it remain unpublished, unrevised and even, apparently, uncomposed.

This effectively planned recital, presented by the Park Lane Group, offered a rare chance to hear all the extant sections together, framed by music of two of the French masters that have meant most to Boulez. The programme opened and closed with works of his teacher Messiaen: the uncompromisingly schematic Quatre Etudes de Rhythme (1949) and two movements from Vingt Regards sur l'Enfant-Jésus (1944), in the latter of which O'Hagan's differentiation of textural layers – the clangorous bell sounds and sumptuous harmonic litanies – was especially impressive.

Also heard was a group of Debussy rarities, including the only recently rediscovered Les Soirs Illuminés par l'ardeur du Charbon (1917), a very late thank-you piece scribbled to his coal merchant and full of nostalgic self-quotations. These were complemented by some late Boulez; his short Incises (1994/2001) alternating battering note tirades with funereal bell sounds.

So what of Sonata No 3? O'Hagan had ordered the sections in a three-movement sequence, and his attention to the varieties of touch and pedalling demanded by Boulez's fanatically precise notation were all that could be desired. Yet it was hard for the ear to get a more than approximate grip on the music's 25 minutes of textural turbulence, rhythmic flux, and unremitting dissonance.

Over and over again one discerned tiny patches of invention which, if extracted and treated as little Webern-like forms on their own, might prove striking indeed – only to hear them swept away in the surge. If the true comprehension of complexity depends on not being confronted with too much of it at a time, then Sonata No 3 seems to flout that principle. Maybe that is why Boulez has found it so hard to get into final form.