THE PIANO-HARPSICHORD DEBATE

THE PIANO-HARPSICHORD DEBATE: MY OPINION

I was planning to post about the piano versus harpsichord problem later. But as it is coming up again, and that some kind people expressed the sentiment that my opinion could have some merit (why the hell am I writing in this damn silly way?) here is what I really think on the subject.

The problem is different whether you are considering Bach or not. Couperin does need a late 17^th century French harpsichord, with strong and vibrant basses and a beautifully mellow middle. It is - I feel, unthinkable to play Froberger or Frescobaldi on the piano - and even the recent Hewitt record of Couperin (François I mean) on the piano, even if it is a Faziolli, is nice rather than convincing.

Why? Because this kind on music very strongly relies on two things no piano can do: brilliant chords that generate beautiful difference tones (that is: added brilliance and depth) and a perfectly transparent texture. Moreover, the music up to Fischer and Bach was played on unequal tuning which is quite alien to piano technique and extraordinarily beautiful on the harpsichord. This is because the harpsichord has so many overtones: with equal tuning the harmonics generated by the strings themselves clash with the chords and intervals; this is not a real problem in the piano because all pianos (from the start) were duller sounding than harpsichords and, therefore, one may quite satisfactorily use equal tuning (the octave is divided in 12 exactly equal semitones, which are, of course, different from the harmonic partials generated by each string, whereas in an unequally tempered keyboard you have to chose which intervals are right and which is wrong).

This has a very important consequence in music up to the end of the 17^th Century/ early 18^th Century. Some tonalities sound beautiful (c major, f major, d minor, g minor) whereas others sound upsettingly stressed and distorted (f minor, e minor, or b minor); more alien tonalities (with more sharps/flats) are totally impossible (a# major, for instance).

So, when a piece is written in f minor (as Louis Couperin’s Pavane, for instance) one should expect a kind of harmonic distortion that functions in the same way a very dissonant chord does.

The case of Bach is different. Not only he does not use the harpsichord possibilities to their limit (there are exceptions, and quite a few: the Chromatic Fantasy, several pieces from WTC I and, of course, the partitas and suites) but his music seems quite independent of the medium used. But I still prefer it on the harpsichord.

To begin with, the fast response of the harpsichord generates a lot of energy that allows middle voices to be heard clearly. That makes polyphony very understandable (as what we really listen to except in the upper and lower voices is the note attack). For instance, the Art of Fugue is clearer when played on the harpsichord (even if it has some parts in very long notes) than the organ.

All this seems to be pointing to the advantages of the harpsichord.

But pianists always say that it is quite impossible to make a voice stand out in the harpsichord because of the lack of dynamics. Now there is some dynamics in the harpsichord, but it is so subtle that pianists simply discard it. Moreover, the fact that the attacks are so clear makes it possible to play with the micro agogics: you can delay a note by a tiny fraction of a second (I mean by about 1/10 of a second, or more if you really want to make the difference clearly audible). In the piano this is feasible, but the effect is the one Saint-Sens explored when mimicking the elephant in Animals Carnival: it merely appears that you are not precise and are creating a muddle.

But I cannot deny that you can bring out a fugue theme in the piano - for instance, when the first note appears in the middle of a chord - whereas this is quite impossible to do on the harpsichord. This is true, but that is not the way polyphony works. Polyphony, one must never forget it, is a compositional technique, not a theme catching game. When the same theme appears over and over again (as, for instance, in the much cited Bach Doric fugue) it is completely impossible to bring it out every theme. And, for instance, in the c# minor fugue of WTC I, even with a piano, it is quite impossible to separate the theme and the second countersubject (or the third theme) when they appear in the bass in stretto (I can check where exactly, if requested).

That simply means that the harpsichordist knows it is there; other people will recognise that there is something familiar going on - they will recognize the general rhythmical structure and the harmonic content that results from the theme presence, and nothing more. This is true for every instance of counterpoint I know. The Art of Fugue is a good example, but better ones are to be found in Palestrina, Ockeghem and, of course, Machault (Ma fin est mon commencement).

What I mean is: when listening to counterpoint one doesn’t have to listen to every entry (as a matter of fact, it would be quite unmusical to do so): what is important is the overall thematic unity.

What about dynamic changes - I mean audible and progressive ones? Quite simply that is completely alien to the spirit of the music written for harpsichord. Take, for instance, the Angela Hewitt Well Tempered Clavier: she uses dynamics freely, and the pieces take a huge an Technicolor appearance that I find completely alien to the scores. Of course one may like it that way. But I personally frown a little: Bach is about tension and release, not about loudness. When Bach wants a louder sound he simply doubles the voices. Quoting again from the Doric fugue, there is a mysterious 5^th voice that appears, if I remember correctly in the tenor, doubling it in thirds. I always though that too much fuss was made about that: the reinforcement (in thirds) it is quite frequently used by Bach (Even in the Art of Fugue, where we are talking about rigorous counterpoint): This simply means that Bach wanted that particular bit to be listenable. Another good example, this time not from counterpoint is the WTC I b flat minor prelude: it has a powerful crescendo before a pause. Surely, one must play it on the piano (the clavichord in Bach’s day was only a cheap alternative - a study instrument; you just have to play one to understand why: when you play you listen to the sound; nobody else further than a circle of about two meters does). But a good harpsichordist (and a good harpsichord) can work wonders: just listen to Gilbert’s version, or even to the Leonhardt one (Gilbert’s harpsichord does have a lot of dynamics, whereas Leonhardt’s has only a slight possibility of accenting the attacks).

Also, it is quite evident that the thinning or thickening of harmonic structure is used precisely in that fashion. For instance, the great g minor fugue (from the organ Fantasia and Fugue) is a four voice fugue. Bach increases and decreases the number of voices therefore creating dynamic differences. The same happens with the Passacaglia (admittedly, that is not strictly speaking rigorous counterpoint, but it is counterpoint nonetheless).

Now I surely wished the harpsichord has a little more dynamics - to make a chord stand out, for instance, but you can fake it by means of articulation silences before the chord.

Lastly harpsichord playing (as organ playing) is not strictly speaking legato. That has two reasons. An old harpsichord has a lot of resonance. If you play a chord and then release the notes, you will listen to the reverberance within the box (old harpsichords had closed bottoms, so the case functions as a huge sound reservoir. In an organ, not only a pipe does not stop ringing after the key is released (but for a very short time and chiefly in the larger pipes) but the reverberance makes it imperative that you clear the way to the next note. As far as I know, except for very quick passages, there is no such thing as true legato on the organ. In the piano this does not happen: you damp the string and the sound stops. A détaché style of playing always sounds wrong to me (let alone the bizarre way of playing of Rosalynd Truck or Glenn Gould).

What about pedals, then? In old music one should avoid it at all costs, pianists usually say. I definitely don’t agree. This is because you often use what is commonly called overlegato in the harpsichord. That is, after the melody has passed over the notes, you leave certain of them put in order to strengthen the tone colour. I don’t mean you have to do it. But just play the first prelude of WTC 1 and you will understand the need. Moreover, a piano without the use of pedals sounds too dull - it was never meant to sound that way.

Further still. ‘Piano allows tone colouring whereas harpsichord does not’. That is totally silly. A double harpsichord usually has at least 3 registers. This allows, for instance, to play a sarabande (many from the English suites, the 6^th from the French ones, and the marvellous 6^th from the partitas) with all registers engaged, so as to be piercingly desperate, but without shouting. In the piano you must play them piano - which means they do not sound desperate but merely sad - otherwise you’ll get a very odd effect. That is to say, you can only have real brightness when playing forte (because the felts act more quickly on the strings and therefore they do not muffle the sound so much). This is a very severe limitation when playing period music: if you want brightness you must have loudness; whereas in a harpsichord you may have brightness and not a very loud sound (but beware: a good harpsichord has a lot of sound!).

My last word on harpsichord resources is the fact that you can actually play with the attacks. If you have both unisons engaged, if you press your finger slowly down (a very difficult thing to do, but Leonhardt masters that craft) the sound will be different from the one obtained when pressing fast.

Admittedly, I’m stressing the harpsichord capabilities and not the piano ones. That happens because I know the harpsichord very well and have not set hands on a piano for about a decade. Pianists will defend their instrument with better arguments because they know the instrument better. But I am not completely ignorant about pianos: I used to play a Steinway when I was young. So I know how magnificent an instrument it is.

My position is therefore very clear: I by far prefer a harpsichord version of harpsichord music than a piano one.

Of course some pianists can do wonders. One such pianist is Murray Perhaya (spelling?). I would say his best recording is the English suite series (far better, I think, than the Goldbergs [but then, I really never liked the Goldbergs; as an organist friend of mine doesn’t either, is there a problem for organists to like them?)]. His playing only relies on flow, tension and release. That is, I think, the essence of Bach’s harpsichord music. He achieves it masterfully. But did you all know that he has a harpsichord and practices his Bach touch on it?

Finally. I am definitely not against playing Bach on the piano. I quite like some interpretations. And I wait eagerly for Pollini’s version of the WTC (it has been announced). And also for the Perahya one (it has not been announced, I think). When a great artist really makes music alive I couldn’t care less for organological [=referring to instruments] considerations.

Truth is: no one really knows what kind of harpsichord Bach liked. Yes, he liked the Mietkes, but no one knows how they sounded. Most harpsichordists use a ‘Mietke copy’. But is it a copy of an unadulterated Mietke? Of course not. No one even knows if the original instrument had brass strings (a very mellow sound) or plain iron ones (a strong, brilliant sound). I personally like French instruments. They sound better than German ones. Period, as far as I am concerned.

The ‘Bach-organ’ is better known. It is thought to be illustrated by the Wenzelkirche in Naumburg (Hildebrandt, probably with a specification drawn by Bach himself). It sounds very well (a monster organ: quite outscores a big Arp Schnitger in terms of power and even brilliance). Also the famous Altenbruch Treutmann is thought as a Bach organ. It is a splendid instrument. What about the Silbermanns? Bach said that Gottfried Silbermann deserved his name because the organs sounded like silver. But that his mixtures, in big churches, were too low. I quite agree with that. Nevertheless, they are beautiful. We know Bach liked powerful mixtures and 32’ foundations in the pedals. Does that mean one can only play Bach in such organs? Not at all. I personally like Bach best in smallish organs in a rather dry acoustic. Counterpoint is clearer and all the toucher nuances are understandable.

So, even if I favour the harpsichord on musical reasons, I quite accept and even like piano performances. As far as the spirit of the music is preserved, I feel quite content.

One last comment. Most harpsichords are very badly rendered in most hifi systems. Either they are too bright, too closely miked, too harsh, and so on. A real harpsichord usually is a sweet sounding, crisp and very precise instrument. It is never really loud, but when talking to someone who is turning the pages for you, you definitely have to raise your voice. CDs usually take out some of the brilliance, although this is not crucial. The crucial point is resolution: a harpsichord (as, indeed, an organ) needs great harmonic resolution, otherwise it will sound hard. So you have to actually have listened to some harpsichords live to reach a decision.