Baldassare Galuppi, a composer of the Italian Baroque, wrote a collection of twelve sonatas for harpsichord. In the first movement (Larghetto) of sonata number 2, the instruction sostenuto (come IVa corda di violino) appears towards the end:

(image from a score on IMSLP, in which the word looks like sosienuto) A keyboard score showing the instruction sostenuto,(come IVa. corda di violino)

My limited 'music Italian' and Google Translate together tell me that "sostenuto (come IVa corda di violino)" means "sustained (like the fourth string of a violin)" - but what that means in terms of playing, I have no idea.

What is special about the fourth string of a violin? How can that property be suggested on a keyboard?

Here is a (lovely) performance on modern piano. At 1:30 we begin the phrase I'm talking about - but I can't tell what's supposed to be happening!

  • 2
    For what it's worth, no baroque composer would ever have written such an instruction. This particular message comes from the editor, as do the slurs, accents, and the decrescendo. (In fact, probably all of the dynamic markings are editorial.)
    – phoog
    Commented Dec 19, 2022 at 12:58
  • 2
    On a violin, sostenuto is quite easy to play. On piano, and in particular harpsichord, very different. I'd guess that this was more written in hope than expectation!
    – Tim
    Commented Dec 19, 2022 at 13:41

2 Answers 2


First of all, you should be aware that this is not Galuppi's marking. I'm guessing you're reading from this IMSLP file (the passage in question appears on p 17). This is a 1920 edition, edited by Giacomo Benvenuti, and should be viewed as a "collaboration" between Galuppi and Benvenuti (like the difference between "Kenneth Branagh's Hamlet" and Shakespeare's). Benvenuti's introduction makes it clear that he is thinking of performance on the piano rather than harpsichord (and, you know, there are pedal markings); he says he has "rigidly preserved the very few original dynamic markings, but introduced more modern ones, suited to our pianoforte." (Though he maintains an awkward "awareness" of its origins, allowing that "Anyone who wants to play this sonata on the harpsichord knows what to do; and so does anyone who wants to reproduce the harpsichord on the pianoforte." He even suggests that "those who perform these sonatas on the pianoforte use the second pedal throughout and a not-too-strong tone.")1

Here is the same section, in the original source2:

enter image description here

If you compare the surrounding pages with the Benvenuti edition you'll see just how much Benvenuti added. There are hardly any dynamics, no indications to speed up or slow down, or to "play like a violin's G string." If you'd like to play something closer to the original, this source is beautifully clean and readable, though Benvenuti mentioned "copyist's errors." There is also what appears to be a modern "Urtext" edition by Schott, which has hopefully done the scholarship to fix any such errors, but refrains from other editorial invasions: it collects 10 sonatas, but one can also buy pdf downloads of one sonata at a time.

Now, even if sostenuto (come IVa corda di violino) is the capricious addition of a heavy-handed editor, and even if today modern scholarship would never throw such whimsical indications in willy-nilly, it's still an interesting slice of performance practice: It's still worth answering what Benvenuti meant and how one is to do it. Spoiler alert, my answer to the second is "beats the hell out of me." But as to what he meant:

On a stringed instrument, you can play a given pitch on multiple strings. For example, you can play A440 on the violin as an open A string, or in a higher position on D, or a still higher position on G. As you shift up on a string, the tone becomes "darker, thicker," that is, lacking in upper partials. In addition, the G string is the physically thickest, which makes it (in a modern performance style) well suited to a big, beefy, muscular tone. The opening of Ravel's Tzigane, written in 1924, four years after this edition, uses only the G string throughout the violin solo introduction until the piano comes in:

Besides "big 'n' beefy," playing in high positions on the G string is also associated with a more soulful, moving quality (especially in the, I'll say it, somewhat maudlin aesthetic of the turn of the century). I would bet good money that Benvenuti was thinking of what had become known as "Bach's Air on the G String":

(Which is itself an attributive chimera: It is in fact August Wilhemj's 1871 arrangement of a piece by Bach that was originally for orchestra and in a different key.)

I think it's reasonable to interpret a period understanding of Benvenuti's direction as "play with a lyrical, soulful, high-romantic sostenuto akin to a violinist playing 'Bach's Air on the G string.'" What does that actually mean on the piano?

Well, I must confess that I find it baffling any time pianists talk of tone. The only options at your disposal are attack and duration; you can't play this on "different strings" as a violin can. The sostenuto part is clear enough; he wants a very lyrical sustain, perhaps especially where one might otherwise add separation, like in the dotted-16th figure. I have a feeling he might be thinking particularly of the second half of the bar, though, especially with the accent he puts on the middle C. But much of it might be less about tone and more about adopting the kind of rubato that might come from a turn-of-the-century violinist's soulful rendering of this measure. Some sense of this could come from this 1914 wax-cylinder recording of Felix Borowski's Adoration in general, and the 1:00 sul G moment in particular; there is much more rubato overall than in many modern interpretations, and the sul G spot in particular is taken as license to stretch the moment out broadly.

It should go without saying that to fully realize this marking (and the others in this edition) is to embrace a recreation of a 1920s work and its aesthetic. There's nothing wrong with that, and indeed it's perhaps a neglected aspect of performance practice (studying and even recreating moments during the reception history of a piece). But if you want to play in a late-18th-century way, you should ditch all these indications, as well as this edition itself, and play from something cleaner.

  1. That's my Google-powered translation from the preface at the start of the file. My impression is that coloriti specifically means musical dynamics, but if it means markings in general, he might be explaining how he added even such directions as you're asking about here.
  2. In case others are curious, the way I found that was: First I found an IMSLP entry for this individual sonata rather than the collection of twelve. This mentions a catalog number of R 50 (apparently this is a system of organizing Galuppi's works, from a 1920s dissertation?), and this line: "Primary Sources: (D-B, Mus.ms. 6997.8)," with a link to RISM, the main database of musical primary sources. The RISM page for this sonata has, at the bottom, "In collection:" and a link to an entry for a holding of 14 sonatas in Berlin, with a handy button saying "See online," where you find a digitized copy.
  • 2
    Great answer. Thanks for posting it. I'm leaving this comment to respond to "a neglected aspect of performance practice" by posting a link to the video Historical Recordings from the Beginning of the 20th Century from earlymusicsources.com -- not their usual subject matter, but a fascinating analysis.
    – phoog
    Commented Dec 19, 2022 at 15:53
  • 1
    @phoog Thanks, that's awesome! Every time I taught someone Borowski's Adoration I made them listen to this 1914 Edison Blue Amberol cylinder; I think I'll add it to the answer as the 1:00 range is not a bad example of the kind of "big 'n' beefy/soulful" sul G I'm talking about. Commented Dec 19, 2022 at 16:15
  • With a dexterous assistant, one perhaps could play it a couple octaves lower on a piano. The assistant would be setting up harmonics on the lower strings - to use those strings to play higher pitches. Not exactly something that you'd see in common performance practice, but weirder things have been done to get extra mileage out of a piano. Done well (that's important), and in presence of an audience of suitably conservative musical tastes, it could cause quite a stir I'm sure. Commented Dec 19, 2022 at 23:34
  • Amazing answer! I don't know how I failed to find that Scott edition but I'm grateful that you did!
    – AakashM
    Commented Dec 20, 2022 at 9:10

That is not sosienuto, but sostenuto, with a bit clipped off during scanning.

So this means you are supposed to hold the notes a bit longer. Playing higher notes on the 4th string or sul G is a technique sometimes called for in scores. It gives a different dark, slightly rough texture, much intensity and a wider vibrato. So in this case the editor wants you to give the notes more "weight", similar to playing the part sul G on a violin.

  • 1
    It's possible that the missing bit of the t arose from a faulty impression rather than a faulty scan.
    – phoog
    Commented Dec 19, 2022 at 12:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.