The Aranjuez cadenza, and why it’s (even) harder than you think…

The adagio from Rodrigo’s Concierto de Aranjuez: there is surely no more widely known and popular movement in the guitar repertoire, and deservedly so.

The cadenza famously works itself up to an anguished climax. Swirling arpeggios lead to a tremendous strummed outburst, punctuated with vicious off-beat pizzicatos in the tutti strings, before the iconic theme’s fortissimo reappearance…

If you don’t know the score, or can put yourself back in the mind of someone who doesn’t, I’m willing to bet that the description above is what you hear. Yet that’s not what’s written. Here’s a reduction:

Concierto de Aranjuez II — cadenza (fig. 11)

The string pizzicatos are on the first, strong beat of the measure, with the upper voice of the chords emphasising the descending semitone a-g#. In particular, the last two quarters of the measure repeat the a-g# in 8th notes, and to my ears need an emphasis on the a, not the g#:

As it might be…

If you sing the passage with this articulation, it all makes perfect sense.

A quick listen to twenty-odd recordings on Spotify suggests that no-one really makes an effort to play it this way (or at least not to audible effect). All too often this is the result:

…but as it usually turns out.

Two exceptions I found were John Williams’ recording from 1984, with Louis Frémaux and the Philharmonia Orchestra (barely, but I think I detect a fuller rasgueado on the a at the mid-point of the first bar of the passage), and Göran Söllscher’s 2003 version with the Orpheus Chamber Orchestra (again, just about detectable). Sadly both of these fine players rule themselves out of my desert-island selection for the work by playing the chords down an octave, pulling the punch of the climax.

Some other recent recordings (Barrueco, Milos) play the chords extraordinarily evenly: surely it’s not a strong reach to think that one can shape the climax more expressively. (And yes, I know Rodrigo asks for fff, but that’s part of the challenge). In particular, not thumping the last eight-note resolution to g# will make the string pizzicato sound more like the strong beat that it is, and playing with the timing of that beat after the strumming, and then the timing of the next strummed chords, can help. Conductor and soloist both need to be sensitive to this (the New York Philharmonic’s remarkable online archive has a copy of the score, marked up by conductor André Kostelanetz, that shows that one conductor, at least, was alive to the problem).

The confusion actually sets in earlier, in the arpeggio ascent to the climactic chords. So characteristic of Rodrigo, it is hard to feel the true meter when it’s played (as commonly) with a strong accent on the open E string at the bottom of the pattern. These off-beats can easily sound like the main beats of the meter, so it’s worth taking seriously Rodrigo’s accents on the upper line. As players if we can really start to hear this line in meter, and project that through our playing, then there’s a much better chance that what follows will make sense.