Once the ears of the world were ready to return to the solo guitar innovations of John Fahey, roughly around the time Rhino released the two-CD retrospective Return of the Repressed, there were a number of players who lined up with guitars in hand to revive his particular style of fingerpicking fusion. Some merely replicated his country blues ragas, others tried to update the sound with laptops or placing a choice phrase within a bigger modern rock puzzle, but few have been able to truly add to the language that he laid out so fully from the late '50s to the early '80s. Three Improvisations on Modified Banjo is one of those few recordings that burns away the brush and forges ahead into new territory. That's not to say that it isn't without precedent. Paul Metzger's banjo modes include a set of sitar-like sympathetic strings on the body similar to what John McLaughlin did with his guitar in the '70s, and he dips into Eastern raga patterns like Fahey, Robbie Basho, and many other guitarists have too. What does set him apart, other than his chosen banjo, is a patient, expansive sense of phrasing that has few Western peers. The album is made up of three long untitled pieces, the opener and closer each passing 20 minutes with the centerpiece ending just short of ten minutes. The first improvisation opens with soft drawn-out tones, and if it weren't for the occasional string buzz, it would be easy to mistake it for the sound of a wood flute. The other two pieces are more easily recognizable as products of a banjo, with the second drawing from American roots for a spare and relaxed front-porch meditation. The final improvisation draws on Eastern influences with an overlaying raga structure but also displays some moves that would not be out of place on a Andrés Segovia album, or a Derek Bailey recording, for that matter. Masterful and sublime through and through, this is a record that will have more than one day in the sun as it passes from one impressed set of ears to another.
Share this page
AllMusic Review by Wade Kergan