For a short piece of music, like a three minute song, context is all important. Back in rock-group days I always had the strongest feelings about running order, whether for a live set or for the albums -- it makes a huge difference to the effectiveness of each number, and the cumulative effect of the whole.

Scarlatti's sonatas need just such attention to their scheduling, whether for concert or recording. His 30 "Essercizi per Gravicembalo", first advertised as published in early 1739, are considered to be the only 'official edition' of his keyboard music issued during his lifetime, although his long-time friend Roseingrave published another version of the same 30 sonatas (with 12 added) at the same time. (In fact, his publisher Benjamin Cooke's royal privilege is actually dated just four days before the Essercizi's announcement, making the question of priority a moot one. In any case, his edition was the one that popularized Scarlatti in his own time, and beyond)

I had always assumed that Scarlatti himself was responsible for the ordering of the Essercizi as they appear in the original edition (now K1-30). But now, forced to realize how much I dislike the context in which each piece finds itself (to the point where I've always rejected out of hand, on musical grounds, the desirability of playing the complete set), I must consider the possibility that he simply left the choice to an eager publisher who claimed he "knew the market best", and therefore considered himself best placed to take certain "artistic decisions". Meanwhile, Jane Clark refers to Roseingrave's publication as "pirated, and much inferior" as well as "creatively edited" (Farinelli as Queen of the Night, Eighteenth-Century Music 2,2, 2005), although the effusive dedication to Joao V itself makes it unlikely George II would have granted his privilege to such a venture.

Playing through the Roseingrave version, I was immediately struck by the effectiveness of the pieces' ordering. The edition's title is "42 Suites", and although this is clearly nonsense, the presence of the word itself suggests some kind of grouping. But in fact, he has only loosely assembled two key groupings at the start (more, I'm sure as examples of the possibility of effective sequencing than as fixed suites). It is the musicality of the overall ordering that impresses most, each piece finding a home regardless of its key relationships with its neighbours.

As the only sure way to test this pudding will be to record the complete set, here's the first 'suite'...

"At the Request of several Subscribers, Mr. Roseingrave has been prevail'd upon, to add a piece of his own", which proves to be a worthy and suitably poignant (given the circumstance) addition:
Introduction -- Andante Moderato download mp3
After this there's an early version of K8 (a great boon to musicologists, once they get around to noticing it), then:
K8 in g -- [L'istesso] Allegro [Differente] download mp3
K4 in g -- Allemanda download mp3(NB clearly slower than the Essercizi's Allegro marking)
K31 in g -- Allegro download mp3
K30 in g -- Fuga (Allegro) [aka "The Cat's Fugue"] download mp3
K2 in G -- Vivace download mp3

If you're familiar with the music, you'll notice differences in some of these. This last one K2 provides a useful comparison between the two editions (the matter of context hardly arising, as in its 'usual setting' this sonata comes across as a mere bauble, its momentary 'minor key diversion' in each half giving the merest dilettante - and quite predictable - hint of 'light and shade'). The textual differences are as follows:
1st half: bb19-20 missing; bb26-27 missing; bb23 & 24 (previously 25-26) are now in the major, repeating new 19 - 20 verbatim three times in all.
2nd half: bb50-53 now in A major (F#s and C#s), producing a highly effective false relation at b54; b61/2 LH adds a single quaver d.
(There are also some differences in ornamentation, chiefest being a mordant in place of a trill on the second bar of each entry of the initial subject.)

What are we to make of these (in the context of a short piece, quite substantial) differences? Which one is 'correct' (in the sense of conveying the composer's original intention)? In brief, I find this version far more musically satisfying -- the repetition of bb17-18 coupled with the four-square excursion to the minor of the Essercizi version becomes already predictable in the first half, let alone the second. But with the omission of bb19-20, we are plunged as it were unexpectedly into the second thematic motif (d'' a'' c#'' d'') - the *possibility* of whose expression in the minor can already be anticipated (this anticipation being deliberately if momentarily heightened by the single extra repetition). It is as if the whole first half becomes a preparation for the second, when the sonority of compass as well as sense of anticipation (intensified even, I would say by the readings in bb54 & 61) lend real weight to the minor key shift when it 'finally' occurs. Which is most likely - that the far more subtle and musical version comes from the composer, or the 'superior understanding' of a creative editor? :) Personally I would suggest that editors are more likely (here anticipating Longo by nearly two centuries) to cut their cloth to suit their perceived market...

Now, the second 'suite': first of all, K32 & 33 (in d and D), which are in a single file on the previous page. Then, K9 in d -- Presto (ie. faster than the better-known Allegro version)download mp3
K34 in d -- Larghetto download mp3
K1 in d -- Allegro download mp3

K3 in a -- Presto download mp3

Then comes K35, the 'Handelian' one on the previous page
K29 in D -- Presto download mp3 (this is the one that's easier on a left-handed keyboard.)

(Not in Roseingrave, a few favourites -- in K# order for now, sorry Rosy):
K84 in c download mp3
K147 in e download mp3
K204b in f -- Allegro download mp3
K205 in F -- Vivo download mp3
K218 in a -- Vivo download mp3
K234 in g -- Andante download mp3
K235 in G -- Allegro download mp3
K519 in f -- Allegro assay download mp3

***NB No Scarlatti page would be complete without a link to Christopher Hail's most exceptional Scarlatti site. The downloadable pdf catalogue itself is a work of enormous labour and huge benefit.***

30/10/09 I decided to do the thankfully minimal amount of work needed to get the 'Scarlatti' harpsichord (a Sandy Rogers 61-note 2x8' - loosely based on the V&A Baffo - which can be heard to great effect on "Westway" from Sky1) working. To demonstrate, here's another version of my current favourite
K205 in F download mp3

as an afterthought, Padre Antonio Soler's amazing Fandango. Totally unique in the literature, it nevertheless represents a style of improvisation with which I doubt Scarlatti was unfamiliar (Soler was his pupil, possibly main copyist). There were several inventions from the 1750s onwards for recording keyboard actions, and it seems to me more likely this piece was 'captured' by some such means, rather than being written down in non-realtime. I often used to play it for my solo spot with Sky - sounds great through a 7kW system! Also had the 'life-privilege' of playing it at the QEH (on this instrument) as a guest of Juan Martin, who provided an excellent extempore flamenco guitar accompaniment - sadly the occasion wasn't recorded. I hope this might be *some* compensation... :)
Fandango download mp3 NB at over 13 mins, this file is 19MB (worth it though)

02/11/09 After the Fandango, the instrument seems to be played-in - just as good if not better than I remember it. And, this version of K205 begins to get the story-telling feel I'm after (for example, Handel S7 Gigue, or the Dieupart Allemande). For those who know the music, the a's in m34 are not marked as flat (despite 33 & 35 having such) and on the basis set out on the Handel page I planned to play the repeat with the usual reading - but as I thought I might, end up liking the 'modally-rich' original, whether an error or not. So I played the repeat the same - you'll have to hear an older version for the difference :)
K205 in F [the good one]download mp3

Of course, it would happen that I read Kirkpatrick's words "the Scarlatti sonatas tell no story", just as I become convinced -- rather suddenly and unexpectedly -- that they, or at least a proportion of them, do just that. "The lessons of life" I call K205 (from the refrain near the end of each half), and just now I realized there's more to K84 than I suspected. Without tempo marking in its single source, it easily invites the usual 'virtuosic' approach - the figuration itself seems to invite brilliance. But the staccato markings over the quarter notes surely require a smoother background in the parallel 6ths (far easier played detached)... at which point "the clock is ticking -- just think, just think, are you really ready?" it seems to ask, the dominant pedal in the 2nd half now sounding eerily cataclysmic.
K84 in c download mp3

This next is one of Scarlatti's most uncompromising. I often used to play it in my teens (without the difficult ornaments), much too fast -- 2/4 Allegro needn't be any faster than tempo giusto, a pleasant walking pace. Handfuls of notes, which on Sandy's instrument sound amazing - I'm now using facs (at last) and thanks to Chris Hail for pointing out several 'missing accidentals' in two primary sources, which if played as written, carry the weirdness level through from the flamenco bits to the 'major key' bits (which I always felt missed some). Anyone who has ever heard a Spanish village band will know that all kinds of things can happen, despite a generally-prevailing major key :)
K175 in a -- Allegro download mp3

And at last, this pair of old favorites (K545 has one of the best riffs of all time :) )
K544 in Bb -- Cantabile download mp3
K545 in Bb -- Prestissimo download mp3

17/11/09 Another possible pair, together in one source (V42), but generally overlooked as K17 (in a rather different version) already appeared as one of the Essercizi
K69 [V42:32] in f download mp3
K17 [V42:33] in F [Presto in Ess. & Ros.] download mp3

18/11/09 [updated version of]
K176 in d/D -- Cantabile Andante/Allegrissimo download mp3

21/11/09 Although perhaps not so well-known today, this next compelling piece is one of 26 sonatas 'arranged for piano' by Granados in the 19th century
K48 [V42:6] in c -- Presto download mp3

29/11/09 Some more favorites -- the first seems to presage the 19th century..
K451 in a -- Allegro download mp3
.. and the second of this pair (at least) everybody knows - for obvious reasons :)
K426 in g -- Andante download mp3
K437 in G -- Presto, quanto sia possibbile (aka 'need for speed') download mp3

08/12/09 completed the following pair
K454 in G -- Andante Spiritoso download mp3
K455 in G -- Allegro download mp3

K46 in E -- Allegro [or Presto] download mp3

K27 in b -- Allegro [Allegretto in one important source] download mp3

K10a (V 42:22 version) in d -- Muipresto [sic] download mp3

29/03/10 new version (with both repeats) of K27 in b -- Allegro/Allegretto download mp3

K443 in D -- Allegro download mp3
K444 in d-- Allegrissimo download mp3

K318 in F# -- [Allegro] Andantedownload mp3
K319 in F# -- Allegro download mp3

K347 in g -- Moderato è Cantabbile, & K348 in G -- Prestissimo download mp3

K263 in e -- Andantedownload mp3
K264 in E -- Vivo download mp3
[this is is an alternate reading from the Münster MS (which may have a greater claim to authenticity than has been allowed) - those familiar with this wonderful piece will be surprised to hear mm97-108 in the minor key]

And to round off this page, probably Scarlatti's best-known sonata with its partner:
K380 & 381 in E -- Andante comodo, Allegro download mp3

I'll be starting another Scarlatti page, put down some ideas along with more recordings...

But meanwhile, here's Scarlatti at his most irrepressible, in
K127 in Ab -- Allegro download mp3
but as always subtlety personified -- consider the jaunty little tune that follows the opening arpeggios. Based on rising apoggiaturas and falling triplets, it's heard nowhere else in the piece. But when the '2nd subject' arrives (around 50"), with its curious, almost obsessive insistancy, it's based on *falling* apoggiaturas and *rising* triplets (along with repeated quarters), and somehow pervades the whole like some kind of introverted psychic backdrop, behind the outer show of bombastic confidence.

[Since writing, I saw it another way: an army on the march (and of course one could make a great arrangement for a military band), all show and flash at the front, with the corps trudging along behind ("Tommy this and Tommy that"), but while it sags downwards a tone in the first half, it opens the second with a step up, and then a further (also the repeated note element ends with an upturn), so I think my previous idea was unnecessarily pessimistic! :) ]

.. And, while we're on the subject of irrepressible, here's
K13 in G -- Presto download mp3
the "Neapolitan street urchin" as my ma dubbed it some fifty years ago.

I'm much indebted to Cubus for his exceptionally moving YouTube performance of K466/L118. I remember, in the early 70s when I got my French Double, a friend told me this was her favorite sonata -- but I'm ashamed to say I couldn't see it back then. But Cubus has made the leap of imagination that can connect this piece with the world of the Romantics. Coincidentally, it's also the very first sonata in the Münster collection, which was owned by Fortunato Santini, who showed it to Chopin and Liszt. So, apart from Clementi's worthy but slim edition (and possibly Roseingrave's), this sonata *may* have formed their introduction to Scarlatti's world. As for the tempo, Andante Moderato is slower than Andante, and we must remember that Scarlatti's tempo indications are Baroque not Classical. (An example of a Baroque Andante, from Handel's 7th Suite). So here's my interim reading of
K466 in f -- Andante Moderato download mp3
And, I decided to pair it with K463 which I've been learning recently. I hope you agree, they suit each other very well!
K466 & K463 (also in f -- Allegro Molto) download mp3

A couple of days ago I tuned the Rogers up from a'=415 to 422.5 Hz. This is the pitch of Handel's tuning fork (and quite possible is more 'authentic' than 415, which is just modern pitch a semitone flat (mostly for reasons of - admittedly great - convenience). The instrumemt is already tuned according to Handel's system -- I wonder if anyone else has tried this combination? :) Anyway, enough to say, it sounds completely convincing to me -- see what you think:
K466 in f -- Andante Moderato (pitched at a'=422.5 Hz) download mp3

On the subject of baroque vs classical tempi, remember what Mozart said about Clementi? "He writes Presto and plays Allegro". Yes of course, because Clementi's ideas of tempo denomination were formed from those of baroque-era composers, largely Scarlatti (who had probably pretty-much fixed his by 1705 or so). Mozart designates the Figaro overture Presto, although it's considerably faster than most of Scarlatti's Prestissimi (imagine K545 played at a similar tempo, if actually possible :) )